THE internet ENCYCLOPEDIA 1 volume 3 phần 9 docx

750 WEB SEARCH TECHNOLOGY document selector is to utilize the fact that most search engines return retrieved results in groups Usually, only the top 10 to 20 results are returned in the first result page but the user can make additional requests for more result pages and more results Hence, a document selector may ask each search engine to return the first few result pages This method tends to return the same number of pages from each selected search engine Since different search engines may contain different numbers of useful pages for a given query, retrieving the same number of pages from each search engine is likely to cause over-retrieval from less useful databases and under-retrieval from highly useful databases More elaborate document selection methods try to tie the number of pages to retrieve from a search engine to the ranking score (or the rank) of the search engine relative to the ranking scores (or ranks) of other search engines This can lead to proportionally more pages to be retrieved from search engines that are ranked higher or have higher ranking scores This type of approach is referred to as a weighted allocation approach in (Meng et al., 2002) For each user query, the database selector of the metasearch engine computes a rank (i.e., 1st, 2nd, ) and a ranking score for each local search engine Both the rank information and the ranking score information can be used to determine the number of pages to retrieve from different local search engines For example, in the D-WISE system (Yuwono & Lee, 1997), the ranking score information is used Suppose for a given query q, ri denotes the ranking score of the local database Di , i = 1, , k, where k is the number of selected local databases for the query, and α = k r j denotes the total ranking score j=1 for all selected local databases D-WISE uses the ratio ri /α to determine how many pages should be retrieved from Di More precisely, if m pages across these k databases are to be retrieved, then D-WISE retrieves m ∗ ri /α pages from database Di An example system that uses the rank information to select documents is CORI Net (Callan et al., 1995) Specifically, if m is the total number of pages to be retrieved from k selected local search engines, then 2(1 + k − i) m∗ k(k + 1) pages are retrieved from the ith ranked local database, i = 1, , k Since 2(1 + k − u) 2(1 + k − v) > k(k + 1) k(k + 1) for u < v, more pages will be retrieved from the uth ranked database than from the vth ranked database Because k i=1 2(1 + k − i) = 1, k(k + 1) exactly m pages will be retrieved from the k top-ranked databases In practice, it may be wise to retrieve slightly more than mpages from local databases in order to reduce the likelihood of missing useful pages It is possible to combine document selection and database selection into a single integrated process In Database Selection, we described a method for ranking databases in descending order of the estimated similarity of the most similar document in each database for a given query A combined database selection and document selection method for finding the m most similar pages based on these ranked databases was proposed in Yu et al (1999) This method is sketched below First, for some small positive integer s (e.g., s can be 2), each of the stop-ranked databases are searched to obtain the actual global similarity of its most similar page This may require some locally top-ranked pages to be retrieved from each of these databases Let sim be the minimum of these s similarities Next, from these s databases, retrieve all pages whose actual global similarities are greater than or equal to sim If m or more pages have been retrieved, then sort them in descending order of similarities, return the top m pages to the user, and terminate this process Otherwise, the next top ranked database (i.e., the (s + 1)th ranked database) is considered and its most similar page is retrieved The actual global similarity of this page is then compared with the current sim and the minimum of these two similarities will be used as the new sim Then retrieve from these s + databases all pages whose actual global similarities are greater than or equal to the new sim This process is repeated until m or more pages are retrieved and the m pages with the largest similarities are returned to the user A seeming problem with this combined method is that the same database may be searched multiple times In practice, this problem can be avoided by retrieving and caching an appropriate number of pages when a database is searched for the first time In this way, all subsequent “interactions” with the database would be carried out using the cached results This method has the following property (Yu et al., 1999) If the databases containing the m desired pages are ranked higher than other databases and the similarity (or desirability) of the mth most similar (desirable) page is distinct, then all of the m desired pages will be retrieved while searching at most one database that does not contain any of the m desired pages Result Merging Ideally, a metasearch engine should provide local system transparency to its users From a user’s point of view, such a transparency means that a metasearch search should behave like a regular search engine That is, when a user submits a query, the user does not need to be aware that multiple search engines may be used to process this query, and when the user receives the search result from the metasearch engine, he/she should be hidden from the fact that the results are retrieved from multiple search engines Result merging is a necessary task in providing the above transparency When merging the results returned from multiple search engines into a single result, pages in the merged result should be ranked in descending order of global similarities (or global desirabilities) However, the heterogeneities that exist among local search engines and between the metasearch engine and local search engine make result merging a challenging problem Usually, pages returned from a local search engine are ranked based on these pages’ local similarities Some local search engines make the local similarities of returned pages available to the METASEARCH ENGINE TECHNOLOGY user (as a result, the metasearch engine can also obtain the local similarities) while other search engines not make them available For example, Google and AltaVista not provide local similarities while Northern Light and FirstGov To make things worse, local similarities returned from different local search engines, even when made available, may be incomparable due to the use of different similarity functions and term-weighting schemes by different local search engines Furthermore, the local similarities and the global similarity of the same page may be quite different still as the metasearch engine may use a similarity function different from those used in local systems In fact, even when the same similarity function were used by all local systems and the metasearch engine, local and global similarities of the same page may still be very different This is because some statistics used to compute term weights, for example the document frequency of a term, are likely to be different in different systems The challenge here is how to merge the pages returned from multiple local search engines into a single ranked list in a reasonable manner in the absence of local similarities and/or in the presence of incomparable similarities An additional complication is that retrieved pages may be returned by different numbers of local search engines For example, one page could be returned by one of the selected local search engines and another may be returned by all of them The question is whether and how this should affect the ranking of these pages Note that when we say that a page is returned by a search engine, we really mean that the URL of the page is returned One simple approach that can solve all of the above problems is to actually fetch/download all returned pages from their local servers and compute their global similarities in the metasearch engine One metasearch engine that employs this approach for result merging is the Inquirus system (http://www.neci.nec.com/∼ lawrence/inquirus.html) Inquirus ranks pages returned from local search engines based on analyzing the contents of downloaded pages, and it employs a ranking formula that combines similarity and proximity matches (Lawrence & Lee Giles, 1998) In addition to being able to rank results based on desired global similarities, this approach also has some other advantages (Lawrence & Lee Giles, 1998) For example, when attempting to download pages, obsolete URLs can be discovered This helps to remove pages with dead URLs from the final result list In addition, downloading pages on the fly ensures that pages will be ranked based on their current contents In contrast, similarities computed by local search engines may be based on obsolete versions of Web pages The biggest drawback of this approach is its slow speed as fetching pages and analyzing them on the fly can be time consuming Most result merging methods utilize the local similarities or local ranks of returned pages to perform merging The following cases can be identified: Selected Databases for a Given Query Do Not Share Pages, and All Returned Pages Have Local Similarities Attached In this case, each result page will be returned from just one search engine Even though all returned 751 pages have local similarities, these similarities may be normalized using different ranges by different local search engines For example, one search engine may normalize its similarities between and and another between and 1000 In this case, all local similarities should be renormalized based on a common range, say [0, 1], to improve the comparability of these local similarities (Dreilinger & Howe, 1997; Selberg & Etzioni, 1997) Renormalized similarities can be further adjusted based on the usefulness of different databases for the query Recall that when database selection is performed for a given query, the usefulness of each database is estimated and is represented as a score The database scores can be used to adjust renormalized similarities The idea is to give preference to pages retrieved from highly ranked databases In CORI Net (Callan et al., 1995), the adjustment works as follows Let s be the ranking score of local database D and s be the average of the scores of all searched databases for a given query Then the following weight is assigned to D : w = + k * (s − s)/s, where k is the number of databases searched for the given query It is easy to see from this formula that databases with higher scores will have higher weights Let x be the renormalized similarity of page p retrieved from D Then CORI Net computes the adjusted similarity of p by w * x The result merger lists returned pages in descending order of adjusted similarities A similar method is used in ProFusion (Gauch et al., 1996) For a given query, the adjusted similarity of a page p from a database D is the product of the renormalized similarity of p and the ranking score of D Selected Databases for a Given Query Do Not Share Pages, but Some Returned Pages Do Not Have Local Similarities Attached Again, each result page will be returned by one local search engine In general, there are two types of approaches for tackling the result-merging problem in this case The first type uses the local rank information of returned pages directly to perform the merge Note that in this case, local similarities that may be available for some returned pages would be ignored The second type first converts local ranks to local similarities and then applies techniques described for the first case to perform the merge One simple way to use rank information only for result merging is as follows (Meng et al., 2002) First, arrange the searched databases in descending order of usefulness scores Next, a round-robin method based on the database order and the local page rank order is used to produce an overall rank for all returned pages Specifically, in the first round, the top-ranked page from each searched database is taken and these pages are ordered based on the database order such that the page order and the database order are consistent; if not enough pages have been obtained, the second round starts, which takes the second highest-ranked page from each searched database, orders these pages again based on the database order, and places them behind those pages selected earlier This process is repeated until the desired number of pages is obtained In the D-WISE system (Yuwono & Lee, 1997), the following method for converting ranks into similarities is employed For a given query, let ri be the ranking score of 752 WEB SEARCH TECHNOLOGY database Di , rmin be the smallest database ranking score, r be the local rank of a page from Di , and g be the converted similarity of the page The conversion function is g = − (r − 1) * Fi , where Fi = rmin/(m * ri ) and m is the number of documents desired across all searched databases This conversion has the following properties First, all locally top-ranked pages have the same converted similarity, i.e., Second, Fi is the difference between the converted similarities of the jth and the ( j + 1)th ranked pages from database Di , for any j = 1, 2, Note that the distance is larger for databases with smaller ranking scores Consequently, if the rank of a page p in a higher rank database is the same as the rank of a page p in a lower rank database and neither p nor p is top-ranked, then the converted similarity of p will be higher than that of p This property can lead to the selection of more pages from databases with higher scores into the merged result As an example, consider two databases D1 and D2 Suppose r1 = 0.2, r2 = 0.5, and m = Then rmin = 0.2, F1 = 0.25, and F2 = 0.1 Thus, the three top-ranked pages from D1 will have converted similarities 1, 0.75, and 0.5, respectively, and the three topranked pages from D2 will have converted similarities 1, 0.9, and 0.8, respectively As a result, the merged list will contain three pages from D2 and one page from D1 Selected Databases for a Given Query Share Pages In this case, the same page may be returned by multiple local search engines Result merging in this situation is usually carried out in two steps In the first step, techniques discussed in the first two cases can be applied to all pages, regardless of whether they are returned by one or more search engines, to compute their similarities for merging In the second step, for each page p returned by multiple search engines, the similarities of p due to multiple search engines are combined in a certain way to generate a final similarity for p Many combination functions have been proposed and studied (Croft, 2000), and some of these functions have been used in metasearch engines For example, the max function is used in ProFusion (Gauch et al., 1996), and the sum function is used in MetaCrawler (Selberg & Etzioni, 1997) CONCLUSION In the past decade, we have all witnessed the explosion of the Web Up to now, the Web has become the largest digital library used by millions of people Search engines and metasearch engines have become indispensable tools for Web users to find desired information While most Web users probably have used search engines and metasearch engines, few know the technologies behind these wonderful tools This chapter has provided an overview of these technologies, from basic ideas to more advanced algorithms As can be seen from this chapter, Web-based search technology has its roots from text retrieval techniques, but it also has many unique features Some efforts to compare the quality of different search engines have been reported (for example, see (Hawking, Craswell, Bailey, & Griffiths, 2001)) An interesting issue is how to evaluate and compare the effectiveness of different techniques Since most search engines employ multiple techniques, it is difficult to isolate the effect of a particular technique on effectiveness even when the effectiveness of search engines can be obtained Web-based search is still a pretty young discipline, and it still has a lot of room to grow The upcoming transition of the Web from mostly HTML pages to XML pages will probably have a significant impact on Web-based search technology ACKNOWLEDGMENT This work is supported in part by NSF Grants IIS-9902872, IIS-9902792, EIA-9911099, IIS-0208574, IIS-0208434 and ARO-2-5-30267 GLOSSARY Authority page A Web page that is linked from hub pages in a group of pages related to the same topic Collection fusion A technique that determines how to retrieve documents from multiple collections and merge them into a single ranked list Database selection The process of selecting potentially useful data sources (databases, search engines, etc.) for each user query Hub page A Web page with links to important (authority) Web pages all related to the same topic Metasearch engine A Web-based search tool that utilizes other search engines to retrieve information for its user PageRank A measure of Web page importance based on how Web pages are linked to each other on the Web Search engine A Web-based tool that retrieves potentially useful results (Web pages, products, etc.) for each user query Result merging The process of merging documents retrieved from multiple sources into a single ranked list Text retrieval A discipline that studies techniques to retrieve relevant text documents from a document collection for each query Web (World Wide Web) Hyperlinked documents residing on networked computers, allowing users to navigate from one document to any linked document CROSS REFERENCES See Intelligent Agents; Web Search Fundamentals; Web Site Design REFERENCES Bergman, M (2000) The deep Web: Surfacing the hidden value Retrieved April 25, 2002, from http://www completeplanet.com/Tutorials/DeepWeb/index.asp Callan, J (2000) Distributed information retrieval In W Bruce Croft (Ed.), Advances in information retrieval: Recent research from the Center for Intelligent Information Retrieval (pp 127–150) Dordrecht, The Netherlands: Kluwer Academic Callan, J., Connell, M., & Du, A (1999) Automatic discovery of language models for text databases In ACM SIGMOD Conference (pp 479–490) New York: ACM Press REFERENCES Callan, J., Croft, W., & Harding, S (1992) The INQUERY retrieval system In Third DEXA Conference, Valencia, Spain (pp 78–83) Wien, Austria: Springer-Verlag Callan, J., Lu, Z., & Croft, W (1995) Searching distributed collections with inference networks In ACM SIGIR Conference, Seattle (pp 21–28) New York: ACM Press Chakrabarti, S., Dom, B., Raghavan, P., Rajagopalan, S., Gibson, D., Kleinberg, J (1998) Automatic resource compilation by analyzing hyperlink structure and associated text In 7th International World Wide Web Conference, Brisbane, Australia (pp 65–74) Amsterdam, The Netherlands: Elsevier Chakrabarti, S., Dom, B., Kumar, R., Raghavan, P., Rajagopalan, S., et al (1999) Mining the Web’s link structure IEEE Computer, 32, 60–67 Croft, W (2000) Combining approaches to information retrieval In W Bruce Croft (Ed.), Advances in information retrieval: Recent research from the Center for Intelligent Information Retrieval (pp 1–36) Dordrecht: Kluwer Academic Cutler, M., Deng, H., Manicaan, S., & Meng, W (1999) A new study on using HTML structures to improve retrieval In Eleventh IEEE Conference on Tools with Artificial Intelligence, Chicago (pp 406–409) Washington, DC: IEEE Computer Society Dreilinger, D., & Howe, A (1997) Experiences with selecting search engines using metasearch ACM Transactions on Information Systems, 15, 195–222 Fan, Y., & Gauch, S (1999) Adaptive agents for information gathering from multiple, distributed information sources In AAAI Symposium on Intelligent Agents in Cyberspace, Stanford University (pp 40–46) Menlo Park, CA: AAAI Press Gauch, S., Wang, G., & Gomez, M (1996) ProFusion: Intelligent fusion from multiple, distributed search engines Journal of Universal Computer Science, 2, 637– 649 Gravano, L., Chang, C., Garcia-Molina, H., & Paepcke, A (1997) Starts: Stanford proposal for Internet meta-searching In ACM SIGMOD Conference, Tucson, AZ (pp 207–218) New York: ACM Press Hawking, D., Craswell, N., Bailey, P., & Griffiths, K (2001) Measuring search engine quality Journal of Information Retrieval, 4, 33–59 Hearst, M., & Pedersen, J (1996) Reexamining the cluster hypothesis: Scatter/gather on retrieval results In ACM SIGIR Conference (pp 76–84) New York: ACM Press Kahle, B., & Medlar, A (1991) An information system for 753 corporate users: Wide area information servers (Tech Rep TMC199) Thinking Machine Corporation Kirsch, S (1998) The future of Internet search: Infoseek’s experiences searching the Internet ACM SIGIR Forum, 32, 3–7 New York: ACM Press Kleinberg, J (1998) Authoritative sources in a hyperlinked environment In Ninth ACM-SIAM Symposium on Discrete Algorithms (pp 668–677) Washington, DC: ACM–SIAM Koster, M (1994) ALIWEB: Archie-like indexing in the Web Computer Networks and ISDN Systems, 27, 175– 182 Lawrence, S., & Lee Giles, C (1998) Inquirus, the NECi meta search engine In Seventh International World Wide Web Conference (pp 95–105) Amsterdam, The Netherlands: Elsevier Manber, U., & Bigot, P (1997) The search broker In USENIX Symposium on Internet Technologies and Systems, Monterey, CA (pp 231–239) Berkeley, CA: USENIX Meng, W., Yu, C., & Liu, K (2002) Building efficient and effective metasearch engines ACM Computing Surveys, 34, 48–84 Page, L., Brin, S., Motwani, R., & Winograd, T (1998) The PageRank citation ranking: Bring order to the Web (Technical Report) Stanford, CA: Stanford University Pratt, W., Hearst, H., & Fagan, L (1999) A knowledgebased approach to organizing retrieved documents In Sixteenth National Conference on Artificial Intelligence (pp 80–85) Menlo Park, CA: AAAI Press and Cambridge, MA: MIT Press Salton, G., & McGill, M (1983) Introduction to modern information retrieval New York: McCraw-Hill Selberg, E., & Etzioni, O (1997) The MetaCrawler architecture for resource aggregation on the Web IEEE Expert, 12, 8–14 Wu, Z., Meng, W., Yu, C., & Li, Z (2001) Towards a highly scalable and effective metasearch engine In Tenth World Wide Web Conference (pp 386–395) New York: ACM Press Yu, C., Meng, W., Liu, L., Wu, W., & Rishe, N (1999) Efficient and effective metasearch for a large number of text databases In Eighth ACM International Conference on Information and Knowledge Management (pp 217–214) New York: ACM Press Yuwono, B., & Lee, D (1997) Server ranking for distributed text resource systems on the Internet In Fifth International Conference on Database Systems for Advanced Applications (pp 391–400) Singapore: World Scientific Web Services Akhil Sahai, Hewlett-Packard Laboratories Sven Graupner, Hewlett-Packard Laboratories Wooyoung Kim, University of Illinois at Urbana-Champaign Introduction The Genesis of Web Services Tightly Coupled Distributed Software Architectures Loosely Coupled Distributed Software Architectures Client Utility Jini TSpaces Convergence of the Two Independent Trends Web Services Today Web Services Description Web Services Discovery Web Services Orchestration 754 754 754 755 755 755 755 755 755 756 756 757 INTRODUCTION There were two predominant trends in computing over the past decade—(i) a movement from monolithic software to distributed objects and components and (ii) an increasing focus on software for the Internet Web services (or e-services) are a result of these two trends Web services are defined as distributed services that are identified by Uniform Resource Identifiers (URI’s), whose interfaces and binding can be defined, described, and discovered by eXtensible Markup Language (XML) artifacts, and that support direct XML message-based interactions with other software applications over the Internet Web services that perform useful tasks would often exhibit the following properties: Discoverable—The foremost requirement for a Web service to be useful in commercial scenarios is that it be discovered by clients (humans or other Web services) Communicable—Web services adopt a message-driven operational model where they interact with each other and perform specified operations by exchanging XML messages The operational model is thus referred to as the Document Object Model (DOM) Some of preeminent communication patterns that are being used between Web services are synchronous, asynchronous, and transactional communication Conversational—Sending a document or invoking a method, and getting a reply are the basic communication primitives in Web services A sequence of the primitives that are related to each other (thus, conversation) forms a complex interaction between Web services Secure and Manageable—Properties such as security, reliability, availability, and fault tolerance are critical for commercial Web services as well as manageability and quality of service 754 Web Services Platforms Security and Web Services Single Sign-On and Digital Passports Payment Systems for Web Services The Future of Web Services Dynamic Web Services Composition and Orchestration Personalized Web Services End-to-End Web Service Interactions Future Web Services Infrastructures Conclusion Glossary Cross References References 758 760 760 762 763 764 764 764 765 766 766 766 766 As the Web services gain critical mass in the information technology (IT) industry as well as academia, a dominant computing paradigm of that of software as a monolithic object-oriented application is gradually giving way to software as a service accessible via the Internet THE GENESIS OF WEB SERVICES Contrary to general public perception, the development of Web services followed a rather modest evolutionary path The underpinning technologies of Web services borrow heavily from object-based distributed computing and development of the World Wide Web (Berners-Lee, 1996) In the chapter, we review related technologies that help shape the notion of Web services Tightly Coupled Distributed Software Architectures The study of various aspects of distributed computing can be dated back as early as the invention of time-shared multiprocessing Despite the early start, distributed computing remained impractical until the introduction of Object Management Group’s (OMG) Common Object Request Broker Architecture (CORBA) and Microsoft’s Distributed Component Object Model (DCOM), a distributed extension to the Component Object Model (COM) Both CORBA and DCOM create an illusion of a single machine over a network of (heterogeneous) computers and allow objects to invoke remote objects as if they were on the same machine, thereby vastly simplifying object sharing among applications They so by building their abstractions on more or less OS- and platform-independent middleware layers In these software architectures, objects define a number of interfaces and advertise their services by registering the interfaces Objects are assigned identifiers at the time of creation The identifiers are used for WEB SERVICES TODAY discovering their interfaces and their implementations In addition, CORBA supports discovery of objects using descriptions of the services they provide Sun Microsystems’ Java Remote Method Invocation (Java RMI) provides a similar functionality, where a network of platform-neutral Java virtual machines provides the illusion of a single machine Java RMI is a language-dependent solution, though the Java Native Interface (JNI) provides language independence to some extent The software architectures supported by CORBA and DCOM are said tightly coupled because they define their own binary message encoding, and thus objects are interoperable only with objects defined in the same software architecture; for example, CORBA objects cannot invoke methods on DCOM objects Also, it is worth noting that security was a secondary concern in these software architectures—although some form of access control is highly desirable—partly because method-level/object-level access control is too fine-grained and incurs too much overhead, and partly because these software architectures were developed for use within the boundary of a single administrative domain, typically a local area network Loosely Coupled Distributed Software Architectures Proliferation and increased accessibility of diverse intelligent devices in today’s IT market have transformed the World Wide Web to a more dynamic, pervasive environment The fundamental changes in computing landscape from a static client-server model to a dynamic peer-to-peer model encourage reasoning about interaction with these devices in terms of more abstract notion of service rather than a traditional notion of object For example, printing can be viewed as a service that a printer provides; printing a document is to invoke the print service on a printer rather than to invoke a method on a proxy object for a printer Such services tend to be dispersed over a wide area, often crossing administrative boundaries, for better resource utilization This physical distribution calls for more loosely coupled software architectures where scalable advertising and discovery are a must and low-latency, high-bandwidth interprocessor communication is highly desirable As a direct consequence, a number of servicecentric middleware developments have come to light We note three distinctive systems from computer industry’s research laboratories, namely, HP’s client utility (e-Speak), Sun Microsystems’ Jini, and IBM’s TSpaces (here listed in the alphabetic order) These have been implemented in Java for platform independence Client Utility HP’s client utility is a somewhat underpublicized system that became the launching pad for HP’s e-Speak (Karp, 2001) Its architecture represents one of the earlier forms of peer-to-peer system, which is suitable for Web service registration, discovery, and invocation (Kim, Graupner, & Sahai, 2002) The fundamental idea is to abstractly represent every element in computing as a uniform entity called “service (or resource).” Using the abstraction as a building block, it provides facilities for advertising and discovery, 755 dynamic service composition, mediation and management, and capability-based fine-grain security What distinguishes client utility most from the other systems is the fact that it makes advertisement and discovery visible to clients Clients can describe their services using vocabularies and can specifically state what services they want to discover Jini The Jini technology at Sun Microsystems is a set of protocol specifications that allows services to announce their presence and discover other services in their vicinity It advocates a network-centric view of computing However, it relies on the availability of multicast capability, practically limiting its applicability to services/devices connected with a local area network (such as home network) Jini exploits Java’s code mobility and allows a service to export stub code which implements a communication protocol using Java RMI Joining, advertisement, and discovery are done transparently from other services It has been developed mainly for collaboration within a small, trusted workgroup and offers limited security and scalability supports TSpaces IBM’s TSpaces (TSpaces, 1999) is network middleware that aims to enable communication between applications and devices in a network of heterogeneous computers and operating systems It is a network communication buffer with database capabilities, which extends Linda’s Tuple space communication model with asynchrony TSpaces supports hierarchical access control on the Tuple space level Advertisement and discovery are implicit in TSpaces and provided indirectly through shared Tuple spaces Convergence of the Two Independent Trends Web services are defined at the cross point of the evolution paths of service-centric computing and the World Wide Web The idea is to provide service-centric computing by using the Internet as platform; services are delivered over the Internet (or intranet) Since its inception, the World Wide Web has strived to become a distributed, decentralized, all pervasive infrastructure where information is put out for other users to retrieve It is this decentralized, distributed paradigm of information dissemination that upon meeting the concept of service-centric computing has led to the germination of the concept of Web services The Web services paradigm has caught the fancy of the research and development community Many computer scientists and researchers from IT companies as well as universities are working together to define concepts, platforms, and standards that will determine how Web services are created, deployed, registered, discovered, and composed as well as how Web services will interact with each other WEB SERVICES TODAY Web services are appearing on the Internet in the form of e-business sites and portal sites For example, 756 WEB SERVICES priceline.com (http://www.priceline.com) and Expedia com (http://www.expedia.com) act as a broker for airlines, hotels, and car rental companies They offer through their portal sites statically composed Web services that have prenegotiated an understanding with certain airlines and hotels These are mostly a business-to-consumer (B2C) kind of Web services A large number of technologies and platforms have appeared and been standardized so as to enable the paradigm of Web services to support business-to-business (B2B) and B2C scenarios alike in a uniform manner These standards enable creation and deployment, description, and discovery of Web services, as well as communication amongst them We describe some preeminent standards below The Web Services Description Language (WSDL) is a standard to describe service interfaces and publish them together with services’ access points (i.e., bindings) and supported interfaces Once described in WSDL, Web services can be registered and discovered using the Universal Description, Discovery, and Integration (UDDI) After having discovered its partners, Web services use the Simple Object Access Protocol (SOAP), which is in fact an incarnation of the Remote Procedure Call (RPC) in XML, over the HyperText Transfer Protocol (HTTP) to exchange XML messages and invoke the partners’ services Though most services are implemented using platformindependent languages such as Java and C#, development and deployment platforms are also being standardized; J2EE and NET are two well known ones Web services and their users often expect different levels of security depending on their security requirements and assumption The primary means for enforcing security are digital signature and strong encryption using the Public Key Infrastructure (PKI) SAML, XKMS, and XACML are some of recently proposed security standards Also, many secure payment mechanisms have been defined (See Figure 1) Web Services Description In traditional distributed software architectures, developers use an interface definition language (IDL) to define component interfaces A component interface typically describes the operations the component supports by specifying their inputs and expected outputs This enables developers to decouple interfaces from actual implementations As Web services are envisaged as software accessible through the Web by other Web services and users, Net SOAP WSDL J2EE HPPM/ MQSeries UDDII Figure 1: Web services Web Methods Web services need to be described so that their interfaces are decoupled from their implementations WSDL serves as an IDL for Web services WSDL enables description of Web services independently of the message formats and network protocols used For example, in WSDL a service is described as a set of endpoints An endpoint is in turn a set of operations An operation is defined in terms of messages received or sent out by the Web service: Message—An abstract definition of data being communicated consisting of message parts Operation—An abstract definition of an action supported by the service Operations are of the following types: one-way, request–response, solicit–response, and notification Port type—An abstract set of operations supported by one or more endpoints Binding—A concrete protocol and data format specification for a particular port type Port—A single endpoint defined as a combination of a binding and a network address Service—A collection of related endpoints As the implementation of the service changes or evolves over time, the WSDL definitions must be continuously updated and versioning the descriptions done Web Services Discovery When navigating the Web for information, we use key words to find Web sites of interest through search engines Often times, useful links in search results are mixed with a lot of unnecessary ones that need to be sifted through Similarly, Web services need to discover compatible Web services before they undertake business with them The need for efficient service discovery necessitates some sort of Web services clearing house with which Web services register themselves UDDI (http://www.uddi.org) supported by Ariba, IBM, Microsoft, and HP, is an initiative to build such a Web service repository; it is now under the auspice of OASIS (http://www.oasis-open.org) These companies maintain public Web-based registries (operator sites) consistent with each other that make available information about businesses and their technical interfaces and application program interfaces (APIs) A core component of the UDDI technology is registration, an XML document defining a business and the Web services it provides There are three parts to the registration, namely a white page for name, address, contact information, and other identifiers; a yellow page for classification of a business under standard taxonomies; and a green page that contains technical information about the Web services being described UDDI also lists a set of APIs for publication and inquiry The inquiry APIs are for browsing information in a repository (e.g., find business, get businessDetail) The publication APIs are for business entities to put their information on a repository E-marketplaces have been an important development in the business transaction arena on the Internet They are a virtual meeting place for market participants (i.e., Web services) In addition to the basic registration WEB SERVICES TODAY and discovery, e-marketplaces offer their participants a number of value-added services, including the following: Enabling inter-Web service interaction after the discovery (the actual interaction may happen with or without the direct participation of the e-marketplace); Enabling supply and demand mechanisms through traditional catalogue purchasing and request for purchase (RFP), or through more dynamic auctions and exchanges; Enabling supply-chain management through collaborative planning and inventory handling; and Other value-added services, such as rating, secured payment, financial handling, certification services, and notification services Thus, e-marketplaces can be developed as an entity that uses public UDDI registries The e-marketplaces are categorized as vertical and horizontal depending on their target market The vertical e-marketplaces, such as VerticalNet, GlobalNetXChange, and Retailer Market Exchange, target a specific industry sector where participants perform B2B transactions In particular, Chemdex, E-Steel, DirectAg.com, and many more have been successful in their respective markets By contrast, horizontal exchanges, such as eBay, are directed at a broad range of clients and businesses Web Services Orchestration By specifying a set of operations in their WSDL document, Web services make visible to the external world a certain subset of internal business processes and activities Therefore, the internal business processes must be defined and some of their activities linked to the operations before publication of the document This in turn requires modeling a Web service’s back-end business processes as well as interactions between them On the other hand, Web services are developed to serve and utilize other Web services This kind of interaction usually takes a form of a sequence of message exchanges and operation executions, termed conversation Although conversations are described independently of the internal flows of the Web services, they result in executions of a set of backend processes A Web service and its ensuing internal processes together form what is called a global process 757 activity to another These activities may be made visible through one or more operations grouped as endpoints As in WSDL, a set of endpoints defines a service WSFL defines global message flows in a similar way A global flow consists of plug links that link up operations of two service providers Complex services involving more than two service providers are created by recursively defining plug links XLANG developed by Microsoft extends the XML Schema Definition Language (XSDL) to provide a mechanism for process definition and global flow coordination The extension elements describe the behavioral aspects of a service A behavior may span multiple operations Action is an atomic component of a behavior definition An action element can be an operation, a delay element, or a raise element A delay element can be of type delayFor or delayUntil delayFor and delayUntil introduce delays in execution for a process to wait for something to happen (for example, a timeout) and to wait till an absolute date-time has been reached, respectively Raise elements are used to specify exception handling Exceptions are handled by invoking the corresponding handler registered with a raise definition Finally, processes combine actions in different ways: some of them are sequence, switch, while, all, pick, and empty Inter-Web Service Modeling and Interaction Web services must negotiate and agree on a protocol in order to engage in a business transaction on the Web X-EDI, ebXML, BTP, TPA-ML, cXML, and CBL have been proposed as an inter-Web service interaction protocol We focus on ebXML as it is by far the most successful one (See Figure 2.) In ebXML (http://www.ebxml.org/) parties to engage in a transaction have Collaboration Protocol Profiles (CPP’s) that they register at ebXML registries A CPP contains the following: Process Specification Layer—Details the business transactions that form the collaboration It also specifies the order of business transactions Delivery Channels—Describes a party’s message receiving and sending characteristics A specification can contain more than one delivery channel Intra-Web Service Modeling and Interaction The Web Services Flow Language (WSFL) (Leymann, 2001), the Web Services Conversation Language (WSFL) (W3C, 2002), the Web Service Choreography Interface (WSCI) (BEA, 2002) and XLANG (Thatte, 2001) are some of many business process specification languages for Web services WSFL introduces the notion of activities and flows which are useful for describing both local business process flows and global message flows between multiple Web services WSFL models business processes as a set of activities and links An activity is a unit of useful work while a link connects two activities A link can be a control link where a decision of what activity to follow is made, or a data link specifying that a certain datum flows from an o3 o3 o7 o7 P o2 Ptt’.o Pt’ o1 G B F A X Y C Pt’ o Z P o5 Ptt’.o5 Figure 2: Intra and inter-Web service modeling and interaction 758 WEB SERVICES Document Exchange Layer—Deals with processing of the business documents like digital signatures, encryption, and reliable delivery Transport Layer—Identifies the transport protocols to be used with the endpoint addresses, along with other properties of the transport layer The transport protocols could be SMTP, HTTP, and FTP When a party discovers another party’s CPP they negotiate certain agreement and form a Collaboration Protocol Agreement (CPA) The intent of the CPA is not to expose the business process internals of the parties but to make visible only the processes that are involved in interactions between the parties Message exchange between the parties can be facilitated with the ebXML Messaging Service (ebMS) A CPA and the business process specification document it references define a conversation between parties A typical conversation consists of multiple business transactions which in turn may involve a sequence of message exchanges for requests and replies Although a CPA may refer to multiple business process specification documents, any conversation is allowed to involve only a single process specification document Conceptually, the B2B servers of parties involved are responsible for managing CPAs and for keeping track of the conversations They also interface the operations defined in a CPA with the corresponding internal business processes Web Services Platforms Web services platforms are the technologies, means, and methods available to build and operate Web services Platforms have been developed and changed over the course of time A classification into four generations of platform technology should help to structure the space: First Generation: HTML and CGI—Characterized by Web servers, static HTML pages, HTML FORMS for simple dialogs, and the Common Gateway Interface (CGI) to connect Web servers to application programs, mostly Perl or Shell scripts (See Figure 3.) Second Generation: Java—Server-side dynamic generation of HTML pages and user session support; the Java servlet interface became popular for connecting to application programs Third Generation: Application server as Richer development and run-time environments—J2EE as foundation for application servers that later evolved towards the fourth generation ebXML registry Service A CPP CPA CPP Service B Figure 3: ebXML service-to-service interaction front-end Internet FW LB web server app server AS WS AS WS AS WS AS WS back-end Back-End DB DB Figure 4: Basic four-tier architecture for Web services Fourth Generation: Web services—Characterized by the introduction of XML and WSDL interfaces for Web services with SOAP-based messaging A global service infrastructure for service registration and discovery emerged: UDDI Dynamic Web services aggregation— Characterized by flow systems, business negotiations, agent technology, etc Technically, Web services have been built according to a pattern of an n-tier architecture that consists of a frontend tier, firewall (FW), load balancer (LB), a Web-server tier (WS), an application (server) (AS) tier, and a backend tier for persistent data, or the database tier (DB) (See Figure 4.) First Generation: HTML and CGI The emergence of the World Wide Web facilitated the easy access and decent appearance of linked HTML markup pages in a user’s browser In the early days, it was mostly static HTML content Passive information services that provided users with the only capability of navigating though static pages could be built However, HTML supported from the very beginning FORMS that allowed users to enter text or select from multiple-choice menus FORMS were treated specially by Web servers They were passed onto CGI, behind which small applications, mostly Perl or Shell scripts, could read the user’s input, perform respective actions, and return a HTML page that could then be displayed in the user’s browser This primitive mechanism enabled a first generation of services on the Web beyond pure navigation through static contents Second Generation: Java With the growth of the Web and the desire for richer services such as online shopping and booking, the initial means to build Web services quickly became too primitive Java applets also brought graphical interactiveness to the browser side Java appeared as the language of choice for Web services Servlets provided a better interface between the Web server and the application Technology to support dynamic generation of HTML pages at the server side was introduced: JSP (Java Server Pages) by Sun Microsystems, ASP (Active Server Pages) by Microsoft, or PHP pages in the Linux world enabled separation of presentation, the appearance of pages in browsers, from content data Templates and content were then merged on the fly at the server in order to generate the final page returned to the browser Since user identification was critical for business services, user log-in and user sessions were introduced Applications were becoming more complex, and it turned out that there was a significant overlap in common functions needed for many services such as session support, connectivity to persistent databases, and security functions WEB SERVICES TODAY 759 Figure 5: The J2EE platform Third Generation: Application Server The observation that many functions were shared and common among Web services drove the development toward richer development environments based on the Java language and Java libraries A cornerstone of these environments became J2EE (Java Platform, Enterprise Edition), which is a Java platform designed for enterprisescale computing Sun Microsystems (together with industry partners such as IBM) designed J2EE (Figure 5) to simplify application development for Web services by decreasing the need for programming through reusable modular components and by providing standard functions such as session support and database connectivity J2EE primarily manifests in a set of libraries used by application programs performing the various functions Web service developers still had to assemble all the pieces, link them together, connect them to the Web server, and manage the various configurations This led to the emergence of software packages that could be deployed easier on a variety of machines These packages later became application servers They significantly reduced the amount of configuration work during service deployment such that service developers could spend more time on business logic and the actual function of the service Most application server are based on J2EE technology Examples are IBM’s WebSphere suite, BEA’s WebLogic environment, the Sun ONE Application Framework, and Oracle’s 9i application server (See Figure 5.) Fourth Generation: Web Services Prior generations of Web services mostly focused on endusers, people accessing services from Web browsers However, accessing services from services other than browsers turned out to be difficult This circumstance has prevented the occurrence of Web service aggregation for a long time Web service aggregation meant that users would only have to contact one Web service, and this service then would resolve the user’s requests with further requests to other Web services HTML is a language defined for rendering and presenting content in Web browsers It does not allow per se separating content from presentation information With the advent of XML, XML became the language of choice for Web services for providing interfaces that could not only be accessed by users through Web browsers but also by other services XML is now pervasively being used in Web services messaging (mainly using SOAP) and for Web service interface descriptions (WSDL) In regard to platforms, XML enhancements were added to J2EE and application servers The introduction of XML is the major differentiator between Web services platforms of the third and the fourth generation in this classification A major step toward the service-to-service integration was the introduction of the UDDI service (see the above section Web Services Discovery) Three major platforms for further Web services interaction and integration are: Sun Microsystems’ Sun ONE (Open Net Environment), IBM WebSphere, and Microsoft’s NET Sun ONE—Sun’s standards-based software architecture and platform for building and deploying services on demand Sun ONE’s architecture is built around existing business assets: Data, applications, reports, and transactions, referred to as the DART model Major standards are supported: XML, SOAP, J2EE, UDDI, LDAP, and ebXML The architecture is composed of several product lines: the iPlanet Application Framework (JATO), Sun’s J2EE application framework for enterprise Web services development, application server, portal server, integration server, directory server, e-commerce components, the Solaris Operating Environment, and development tools IBM WebSphere—IBM’s platform to build, deploy, and integrate your e-business, including components such as foundation and tools, reach and user experience, business integration, and transaction servers and tools Microsoft NET—Microsoft’s NET platform for providing lead technology for future distributed applications inherently seen as Web services With Microsoft NET, Web services’ application code is built in discrete units, XML Web services, which handle a specified set of tasks Because standard interfaces based on XML simplify communication among software, XML Web services can be linked together into highly specific applications and experiences The vision is that the best XML Web services from any provider around the globe can be used to create a needed solution quickly and easily CURRENT STATE OF THE INTERNET 833 Table World Internet Users Millions of subscribers 1997 1998 1999 2000 2001 2002 2003 2004 2005 55 World 1996 101 150 201 407 518.5 664.7 813.7 978.8 1156a Source: Nua Ltd (up to year 2001) and eTForecasts (forecasts from 2002–2005) a Extrapolated from estimates for 2004 Classless Addressing Each node or a device on the Internet must have a unique IP address Messages are routed through the IP network using the IP address of the destination The most widely deployed version of IP today is version (IPv4) (Postel, 1981a), which uses 32 bits of address space The IP address is typically written as four separate numbers, each coded using bits, separated by periods (an example would be 128.200.222.100) In an isolated network, any IP address that matches the format may be used (except for certain reserved addresses), as long as each node or device has a unique address However, to connect to the Internet, it is necessary to obtain addresses not being used so as to avoid having duplicate addresses with other networks and nodes In the early days, the InterNIC assigned to anyone who asked a range of addresses that belongs to one of three classes Class A is the largest class, which supports up to 16 million hosts on each of 127 networks A Class B network supports up to 65,000 hosts, and a Class C network supports up to 254 hosts These classes were also used to route packets (see the section Dynamic Routing) Hence, this type of address allocation scheme has been termed classful addressing or routing Although simple, this allocation scheme has proven to be very inefficient in terms of actual use of IP addresses For example, a modestly sized network may need to support up to 3,000 nodes in its own network Since a Class C network supports only 254 hosts, a Class B network would have been assigned even though less than 5% of the assigned addresses will be used The possibility of running out of IP addresses became quite real as the growth of the Internet exploded Although IPv6 (Deering & Hinden, 1998) can solve the address crunch by increasing the address space to 128 bits, it does not resolve the nearer term problems described next Laptop Router Since Class A networks are far too large for most organizations, and Class C networks are too small, Class B networks were the most commonly requested and granted type In August 1990 during the Vancouver Internet Engineering Task Force (IETF) meeting, Frank Solensky, Phill Gross, and Sue Hares projected that the Class B space would be depleted by March 1994 This led the InterNIC to force many smaller organizations to accept several Class C networks rather than a single Class B network In the example above, 3,000 nodes would require at least 12 Class C networks The problem with this situation is that each network must be routed individually when classful routing is used, so instead of having a single entry for the network in the routing table, there would now be 12 entries in the table This led to extremely large routing tables in backbone routers, slowing down the whole network Classless Interdomain Routing To address this problem, classless interdomain routing (CIDR) (Hinden, 1993) and classless addressing (Rekhter & Li, 1993) were introduced in 1993 Instead of using one of just three partitions between the network and the host portion of the IP address (determined by the “class” of the network), it became possible to have variable length network identifiers or “prefixes.” CIDR can have any prefix length between 13 and 27 bits instead of just 8, 16, or 24 bits in classful routing Thus, the smallest network can have up to 32 hosts and the largest network can have more than 500,000 hosts A CIDR address includes the standard 32-bit IP address as well as the information on how many bits are used as the network prefix Increasing the efficiency of address allocation has reduced not only the address crunch, but also the size of the routing table CIDR can also be used for route aggregation, which further reduces the size of the routing table Deployment of CIDR has contributed greatly to the continued growth of the Internet Database QoS-enabled wireless backbone network PDA * # QoS-enabled IP network Host Figure 2: Architecture of packet-switched wireline/wireless multimedia network 834 WIRELESS INTERNET Dynamic Routing In a circuit-switched connection as used in a public switched telephone network (PSTN), the path of a call is established at the beginning of a connection and is maintained throughout the duration of the call One of its weaknesses is the vulnerability of the network when central switching stations fail A much worse problem for data traffic has been the fact that such traffic tends to be bursty and the reserved circuit is unused much of the time, thus wasting much of its capacity In order to increase the efficiency and robustness, a packet-switched architecture was envisioned for the Internet Packets from many sources and destinations may share a common transmission circuit but be switched independently, and where the flow of data packets is constantly monitored and adjusted through the network, possibly around any failed nodes This process of directing the switching and flow of packets is called “routing.” Routing in the Internet is performed by a device called a router, which has connections to multiple networks and has the ability to relay data packets between these networks The router decides to which network each packet should be sent based on the information available in the IP header of each packet Currently, most routers look at only the destination address to decide the fate of a packet, even though other information is available (see the section Real-Time Traffic Support) Instead of the entire IP address, only a subset of the address is used by the router In classful routing, only the class portion of the address is used, while in CIDR, only the prefix portion of the address is used Based on the destination address, the router consults its routing table, which lists the port number associated with the precomputed “optimal” path for that destination address The routing table can be configured manually or automatically In either case, the optimal path is determined based on the network topology and different link metrics (e.g., bandwidth, delay, load, reliability, and cost) When the routing table is configured manually, static routing is said to be used It is the preferred mode of operation in networks with few nodes or in stub networks, which is a network with only one or two paths to the rest of the network While simple to configure for small networks, a router using static routing cannot reroute packets automatically if the instantaneous traffic loads change or if a router or a link in a preconfigured route goes down for any reason The destination may remain unreachable until human intervention is made to update the routers, based on new traffic loads or utilization of the failed link or node Dynamic routing addresses these deficiencies by enabling dynamic and automatic update of the routing tables in routers For example, if a router detects failure of a link, it will notify other routers of the condition so that appropriate adjustments to the routing table entries can be made The algorithm for calculating the optimal path and the mechanisms for sharing information among different routers specify numerous routing protocols used in the Internet Routing protocols are generally divided into two groups Interior gateway protocol (IGP) is used within ASs, and exterior gateway protocol (EGP) is used between routers of different ASs, i.e., across the Internet The two most widely used IGPs are routing information protocol (RIP) (Hendrick, 1988) and open shortest path first (OSPF) (Moy, 1998b) Examples of EGP include somewhat confusingly termed exterior gateway protocol (EGP) (Mills, 1984) and border gateway protocol (BGP) (Rekhter & Li, 1994) For details about the protocols, interested readers may consult the RFCs as well as numerous books and articles (Doyle, 1998; Huitema, 2000; Moy, 1998a) Below we present a brief overview of the most popular routing algorithms currently in use in the Internet Routing Information Protocol (RIP) RIP is a distance vector protocol that uses the number of “hops” (i.e., the link between two nodes) to determine the optimal path to the destination Every 30 s, the entire routing table is broadcast, and the routing tables of the listening routers are updated based on the reported hop counts This reliance on second-hand information results in relatively low computation and storage requirements, but it also results in slow convergence RIP is suitable only for relatively small networks (no wider than 15 hops) since the maximum hop count is restricted to 15 (a hop count of 16 is considered infinity) In addition, the high bandwidth overhead of broadcasting entire routing tables makes the algorithm hard to scale Open Shortest Path First (OSPF) OSPF is a link state-based algorithm, in which the link states of each node are flooded to all routers in the network The transmission takes place only when there is a change in network topology The link state may convey information about various link states such as throughput, delay, loss rate, or some cost function Based on the link states, each node in the network computes the complete network topology As all nodes have access to the same link state information, all nodes should arrive at the same network topology Based on this map of the network, the shortest path tree to each destination is computed using a shortest path first algorithm, and the result is used to populate the routing table Such computation places much heavier burden on memory and processing power than algorithms such as RIP, but it also results in faster response to network events such as a link failure, quicker route convergence, and requires less traffic overhead especially for larger ASs OSPF also offers additional advanced features that are described in detail in the literature Such advantages are driving the increasing use of OSPF over RIP Border Gateway Protocol (BGP) BGP is an interautonomous system (inter-AS) routing protocol BGP version has become the main inter-AS routing protocol in the Internet BGP is a path vector protocol, which defines a route as a pairing between a destination and the attributes of the path to that destination A BGP router learns of multiple paths via internal and external BGP speakers and picks the best path, which is then sent to external BGP neighbors A network administrator has control over the policies applied during the best path selection In addition, unlike other routing protocols, CURRENT STATE OF THE INTERNET BGP is connection oriented and uses TCP as the transport protocol BGP supports IP prefixes and path aggregation, which makes it suitable for CIDR Real-Time Traffic Support An important class of applications used over the Internet is real-time Real-time applications generate traffic that must be communicated to the recipient(s) within a very short amount of time The time limit may be milliseconds for interactive applications such as voice and video telephony, but it may be larger for certain streaming or transaction processing applications Despite increasing interest in supporting such applications over the Internet, the Internet is still largely a best-effort network without any delivery guarantees necessary to provide the levels of quality of service (QoS) expected from end-users QoS Support Different applications typically have different traffic characteristics and QoS requirements Applications such as streaming video usually require large bandwidth, and interactive applications such as voice over IP (VoIP) and video over IP (VIP) require tight delay bounds as the data must be played back continuously at the rate they are sampled If the data (packet) does not arrive in time, the playback process at the receiver will be disturbed For example, in voice telephony, human beings can tolerate a latency of up to about 200 ms (Brady, 1971), although in most of today’s voice networks, the latency is limited to around 50 ms If the latency exceeds this value, the degradation in call quality will be noticeable If enough extra bandwidth is available, best-effort service may be able to fulfill the delay, throughput, and other requirements When resources are scarce, however, real-time traffic will suffer from congestion and delay, resulting in degradation in the application quality DiffServ To facilitate end-to-end QoS on IP networks, the IETF has defined two models: Integrated Services (IntServ) (Braden, Clark, & Shenker, 1994) and Differentiated Services (DiffServ) (Blake et al., 1998) IntServ was defined first and follows the signaled-QoS model, where the end-hosts signal their QoS needs to the network using resource reservation protocol (RSVP) (Zhang, Deering, Estrin, Shenker, & Zappala, 1993) RSVP signaling and reservation of the desired QoS is done for each flow in the network A flow or a stream is defined as an individual, unidirectional data stream between two applications, and is uniquely identified by a 5-tuple (source IP address, source port number, destination IP address, destination port number, and the transport protocol) While IntServ provides for a rich end-to-end QoS solution, there are several problems with the approach State information for each reservation needs to be maintained at every router along the path, and each packet must be processed to ensure that the QoS of each flow is being satisfied As there may be hundreds of thousands of simultaneous flows going though a network core (router), it is not clear whether IntServ will scale well in terms of complexity of admission control, memory requirements of maintaining state, and packet processing overhead 835 Since per-flow QoS is difficult to achieve in an endto-end fashion without introducing scalability issues, it naturally leads one to think about classifying flows into aggregates (classes), and providing QoS to aggregates rather than to individual flows For example, all real-time flows could be grouped into a single class, and bandwidth and other resources can be allocated for the class This would reduce the router’s burden on classification of traffic, signaling, and state maintenance requirements This is the approach taken in the DiffServ model In this model, packets are first divided into classes by marking the type of service (ToS) byte in the IP header A 6-bit bit-pattern called the Differentiated Services Code Point (DSCP) in the IPv4 ToS Octet or the IPv6 Traffic Class Octet is used to this end Once packets are classified at the edge of the network, specific forwarding treatments, formally called per-hop behavior (PHB), are applied at each network element, providing the packet with appropriate guarantees (deterministic or statistical) on performance metrics such as delay, delay jitter, and bandwidth This combination of packet marking and utilization of PHBs results in a more scalable QoS solution MPLS Multiprotocol label switching (MPLS) (Rosen, Viswanathan, & Callon, 2001) is another emerging technology that seeks to introduce QoS guarantees on the Internet Existing link state protocols, specifically OSPF and IS-IS, provide the link state information about the underlying IP network Such information is used to determine the best path through the network called label switched paths (LSPs) using criteria such as number of hops and other configurable parameters such as delay and bandwidth An incoming packet to an MPLS network is assigned a “label” by an “edge-label switch router” (Edge-LSR) This label is swapped by intermediate label switch routers (LSRs) as the packet traverses the MPLS network on an LSP, and the final label is removed when leaving the MPLS network The label distribution protocol (LDP) is used to establish label-to-destination network mappings Forwarding of a packet is based solely on the contents of the label, and not on the IP headers as is done in normal IP routing, speeding up the process Such increase in performance, as well as the ability to perform traffic engineering, makes MPLS a strong contender for the converged network Real-time Transport Protocol (RTP) The real-time transport protocol (RTP) (Schulzrinne, Casner, Frederick, & Jacobson, 1996) is an IP-based protocol providing support for the transport of real-time data traffic such as video and audio streams The services provided by RTP include timestamping, sequence numbering, and other mechanisms to take care of the timing issues RTP also provides information about the encoding method used in the underlying data Through these mechanisms, RTP provides end-to-end transport for real-time data over the IP network RTP was primarily designed for multicast of real-time data; unicast is supported as well It also can be used for one-way transport such as videoon-demand service as well as interactive services such as VoIP and VIP 836 WIRELESS INTERNET RTP was designed to work in conjunction with an auxiliary control protocol called real-time control protocol (RTCP) (Schulzrinne et al., 1996) In an RTP session, participants periodically send RTCP packets to convey feedback on quality of data delivery and information of membership RFC 1889 defines five RTCP packet types to carry control information These packets contain information regarding number of packets lost, interarrival jitter, and timestamps Through these control information packets, RTCP provides services such as QoS monitoring, congestion control, intermedia synchronization, and calculation of round-trip delays RTP is typically run on top of user datagram protocol (UDP) (Postel, 1980) to make use of its multiplexing and checksum functions However, use of these protocols introduces bandwidth overhead to each data packet, which is especially important in low-speed wireless links Transmission Control Protocol (TCP) and UDP Transmission control protocol (TCP) (Postel, 1981b) is the most commonly used transport protocol on the Internet TCP provides a connection-oriented and reliable flow between two hosts, while UDP provides a connectionless and unreliable datagram service over the network UDP was chosen as the target transport protocol for RTP because of three reasons First, since RTP was initially designed for multicast, it was realized that connection-oriented TCP does not scale well for a large number of flows and therefore is not suitable Second, for real-time data, 100% reliability is not as important as timely delivery Since TCP provides reliability through retransmissions, it is not suitable for real-time applications, for by the time packet error or loss is detected and the retransmitted packet is received, the playback time of the data contained in the packet would likely have passed Thus, retransmission only increases the network traffic without benefiting the quality of the playback Third, congestion control of TCP does not match well with the needs of real-time applications, as real-time applications cannot tolerate packets being held back even during periods of congestions Internet Applications: World Wide Web, E-mail, Instant Messaging Since the advent of the Internet, a handful of applications have become the drivers of the growth of the Internet Some of these applications include e-mail, file transfer, instant messaging, and the World Wide Web (WWW) File transfer was one of the original applications envisioned for the Internet and continues to be one of the most popular applications Typically, file transfer protocol (FTP) (Postel & Reynolds, 1985) is used to send and retrieve files from a remote computer E-mail was initially used to send text messages between individuals or groups of individuals, but increasingly diverse media such as images and audio are being transmitted using e-mail According to the Nielsen/NetRatings First Quarter 2002 Global Internet Trends report (2002), e-mail was the most dominant online activity in 12 countries over the previous six months Three technologies were invented by 1991 to accommodate the arrival of the WWW Hypertext markup language (HTML) (Raggett, Le Hors, & Jacobs, 1999) is used to produce Web documents, hypertext transfer protocol (HTTP) (Fielding, Gettys, Mogul, Frystyk, & BernersLee, 1997) is used to transport HTML documents from the server to the client, and the client Web browser is used to retrieve, interpret, and display HTML documents Perhaps no application has been more instrumental in piquing the general public’s interest in the possibilities of the Internet In March of 1993, WWW traffic measured mere 0.1% of NSF backbone traffic By February 1995, WWW passed FTP as the largest volume Internet application The main attraction of WWW lies in its flexibility to service documents containing various media including text, graphics, audio, and video, as well as the ease of accessing documents using “hotspots.” Instant messaging (IM) is an application that has been experiencing a tremendous growth of late IM enables individuals to create private “chat” sessions, in which various types of messages may be exchanged It offers all the capabilities of another immensely popular application, email, but with near real-time response The growth of IM has been strong for both home use and business use A November 2001 study from Jupiter Media Metrix showed that the number of unique business users of the top three IM applications (AOL, MSN, and Yahoo!) increased from 10 million in September 2000 to 13.4 million in September 2001 During the same time period, the total usage time increased from 2.3 billion minutes per month to 4.9 billion minutes per month The growth in home use has been equally impressive with the number of users reaching 53.8 million and the total usage time reaching 13.6 billion minutes a month IM service is very similar to short messaging service (SMS) available in the wireless networks, and the increasing use of IM over the Internet will only increase the demand for SMS (Instant messaging, n.d.) CURRENT STATE OF CELLULAR SYSTEMS (FOCUS ON 2G) Cellular Layouts Most wide area wireless networks today are cellular The service area is divided into smaller service areas called cells In contrast, the first mobile telephony systems in operation up to late 1970s did not use the cellular layout Instead, they relied on a high-power transmitter to service a large area Because only a fixed number of frequency channels were assigned to each service area, increasing demand for mobile telephony meant increasing competition for the available channels, resulting in excessively high call blocking rates The cellular system was designed to address this issue, to increase capacity through the employment of frequency reuse The first systems that used cells were based on analog technology and are referred to as first generation (1G) systems In some places, such as the United States, the analog network still exists to serve customers that have not yet transitioned to the next generation of cellular systems that use digital transmission technology The digital cellular service is referred to as the second generation (2G) system By going over to CURRENT STATE OF CELLULAR SYSTEMS (FOCUS ON 2G) digital transmission, the cellular systems gained tremendous flexibility because of the inherent intelligence that could now be built into the management and control of the system In addition, digital signals can be pre- and postprocessed to enhance the received signal quality Third generation systems, besides being all digital, promise to offer higher bit rates for data services and to switch from circuit- to packet-based transmission To address the need for higher bit rates for data services, some intermediate technologies that fit into the 2G architecture were defined We discuss those in the section Higher Bit Rates for Data—GPRS and HDR A more detailed description of 3G is given in the section 3G Cellular Systems and in the final section of the chapter we very briefly discuss the directions for 4G The advantage of creating cells is that after dividing the service area into cells, the same frequency sets can be used and reused systematically Because the same set of frequency bands can be reused many times, a larger number of users could be supported (Jakes, 1993; Yacoub, 1993) Each cell has a base station, which contains antenna and radio equipment, as well as a high-speed, highcapacity connection to the network Since the area of the cell is typically much smaller than the carrier’s coverage area, lower transmit power can be used to communicate with the mobile station (MS) However, even with lower transmit power, signals still propagate into neighboring cells, causing interference Thus, cells with the same frequency sets are spaced many cells apart, and immediately neighboring cells use different sets of frequency channels to reduce interference (one exception is CDMA systems, such as IS-95, in which the same frequencies can be used in all cells due to its ability to work at low carrierto-interference ratios) (Viterbi, 1995) A group of nearest cells that use disjoint sets of frequency bands is called a cluster The service area is composed of these clusters, which reuse all frequency bands allocated to the service provider Most current cellular networks utilize an additional level of subdivision to further increase capacity beyond that achieved using the cellular concept Instead of making cells even smaller (which introduces the problem of frequent handoffs, which will be discussed later), each cell is divided into or more sectors (Yacoub, 1993) Previously, an omnidirectional antenna was used at the base station to transmit signal equally in all directions, which also distributes interference equally in all directions The omnidirectional antenna is replaced by several directional antennas, each of which can direct the radio wave to a certain direction For example, an omnidirectional antenna may be replaced by three directional antennas, each with a beam width of 120◦ Thus, the cell is effectively sectorized into three distinct areas, each with its own set of frequencies Since the signal is transmitted only in the sector that contains the MS, sectorization reduces the interference between cells in adjacent clusters Reduced interference means that clusters can be located closer together, enabling more frequent reuse of the same frequencies and subsequently increasing traffic capacity A major obstacle to the cellular concept working effectively involved a mobile user traveling from the coverage area of one cell to that of another during a call This 837 problem did not exist, at least not to the same extent, in noncellular systems where the service areas were much larger However, as the sizes of cells shrank (in order to maximize frequency reuse), it became very likely for a user to travel outside the coverage of a cell, and simply dropping those calls was unacceptable from the point of view of the mobile user To address this concern, handoff techniques were developed as a part of different wireless standards By using handoffs, it became possible to automatically transfer a call from a radio channel in one cell to that in another without disrupting the ongoing connection This ability made the adaptation of the cellular concept practical and realizable Mobility and Roaming One of the primary benefits of wireless telephony is the ability to move around without the concern of losing connectivity, at least most of the time in most areas However, typical mobile calling plans specify a home calling area, such as a particular metropolitan area, a state, or even the entire United States When a mobile user travels outside this area, he is said to be “roaming.” Even if the home calling area covers a large area, it is possible and quite likely that in certain geographical locations, the user’s carrier does not have coverage while another carrier may Thus, roaming capability is necessary in order to increase the level of any time, anywhere access, to which mobile users have become accustomed and expect The latter type of roaming is made possible through the business agreements between carriers and service providers to grant each other’s customers access to their networks, in addition to technologies and standards summarized below Two of the most popular technologies enabling the roaming service are GSM Mobile Application Part (MAP) (MAP Specification, 1997) and IS-41 (Telecommunications Industries Association, 1991) Both have become standards adopted in their respective areas of application MAP is used in GSM networks (Mouly & Pautet, 1992), while IS-41 is used in IS-136 (Telecommunications Industries Association, 1996) and IS-95 networks (Telecommunications Industries Association, 1993) IS-41 was developed a few years after the development of GSM MAP, and adopted terminology, network architecture, and some protocol details from MAP Three major components are used to enable the roaming service: the home location register (HLR), the visitor location register (VLR), and the mobile switching center (MSC) The HLR is the database that contains information about subscribers in the network, including the current locations of the subscribers For roaming subscribers, the location is in the form of the signaling address of the VLR associated with the subscriber The VLR temporarily stores a subset of information available in the HLR for those mobile users whose HLRs are located elsewhere The MSC is a telephone exchange that is able to set up and route mobile calls For each user in its service area, the MSC utilizes either the HLR or the VLR to setup and route calls In both IS-41 and GSM MAP, common channel signaling system number (SS7) (Black, 1997) is used to exchange call setup and routing information, including HLR and VLR access, over a digital signaling network 838 WIRELESS INTERNET Voice Telephony as the Primary Service The current generation of the wide area wireless networks, i.e., the cellular networks, has been developed primarily to serve voice telephony From the physical layer (e.g., channel coding and transmission) to the application layer (e.g., voice compression), all aspects of the system were designed and optimized for the purpose of maximizing the capacity and the quality of phone calls Despite the increasing use of services such as wireless Web and SMS (to be discussed later), the bulk of present-day wireless traffic remains voice telephony For example, according to Strand Consult’s report from 2002, more than 85% of mobile service revenues in Europe were from voice telephony in year 2002 (How to make money, 2002) This is especially noteworthy since Western Europe is one of the leading markets in use of non-voice wireless services Another leading market is South Korea SK Telecom, Korea’s leading mobile carrier with 11 million subscribers which claims to have the most advanced mobile network based on CDMA 2000 1X (Telecommunications Industries Association, 2001), recently reported that it expects to gather only 5.4% of its expected total sales from mobile Internet services Even though this figure represents a large increase over previous years, it remains small in comparison to the revenue from voice telephony The lesson here is that while the growth of nonvoice services over cellular networks will continue, it is important to continue to serve what is currently the most valuable market, i.e., voice telephony Popularity of Instant Messaging (SMS) Since its launch in 1995 as a part of the original GSM specifications, short messaging service has become a tremendously popular service offered on wireless telephone networks Although its nascent growth was fueled by the younger generation, by the year 2000 the popularity of SMS has spread beyond the original group Worldwide, the use of SMS has skyrocketed The GSM Association announced in February 2001 that 15 billion messages were sent over the world’s GSM wireless networks during December 2000, compared to only billion messages a year before In terms of revenue, research outfit IDC expects SMS revenue to reach US$6.5 billion worldwide by 2002 SMS provides the ability to send and receive text messages to and from mobile handsets with message lengths ranging from 120 to 256 characters The communication is near real-time as in Internet-based IM In addition to GSM, SMS or SMS-like services are available in other popular wireless standards, such as in IS-136 and IS-95 networks Virtually any type of information based on text can be sent using SMS, including e-mails, news headlines, and some games The range of applications of SMS has yet to be thoroughly explored and continues to expand The tremendous success of SMS has resulted in development of more advanced versions of SMS The enhanced messaging service (EMS) is an open 3rd Generation Partnership Project (3GPP) standard (3GPP TS 23.040, 2002) that allows a mobile phone to send and receive not only plain text, but text enhanced with different fonts and sizes, images, sounds, and animation Since EMS also utilizes the signaling channels for transport and the same SMS Centers, no network modifications are needed to support SMS, which allows a relatively painless upgrade path In comparison, multimedia messaging service (MMS) offers more drastic changes both in terms of functionalities and requirements (3GPP TS 23.140, 2002) For example, MMS data are delivered over the traffic channel rather than the signaling channel This would require one of the new mobile network infrastructures, such as General Packet Radio Service (GPRS) (Cai & Goodman, 1997) or 3G, as well as new network elements such as multimedia messages relays and servers to fully utilize the capabilities of MMS In return, rich media messages composed of text, images, audio, and video will be made possible Some of the envisioned applications include maps, cartoons, games, and interactive videos Just as the current success of SMS was totally unexpected, it is imprudent to speculate on the future of these newfangled technologies With the development of suitable applications, however, these services have a good chance of success Web Access Based on the growth from the traditional methods of Web access based on modems and other wireline devices, wireless Web access was expected to be the next killer application on the wireless cellular networks Who wouldn’t want to access all those Web sites with the conveniences of not being tethered to any sockets or switches on walls? However, the lofty expectations of exponential growth have yet to be realized Problems such as limited speed, incompatible data standards, and poor user interfaces have hindered the growth Even so, there were already more than 28 million subscribers in Japan, or close to 1/5 of the population, who subscribe to NTT DoCoMo’s i-mode mobile Internet service at the end of April 2002 (technically, i-mode is a specification, not a standard) Fortunately, there has been much ongoing progress in addressing some of the difficulties that have plagued wireless Web access Higher bandwidth technologies such as GPRS and high data rate (HDR) CDMA (Bender et al., 2000) are beginning to be deployed and will increase data transmission speeds to more than 171 Kbps Advances in user interfaces such as voice recognition and better displays, in conjunction with increasing use of industry standards such as WAP and i-mode, will continue to improve accessibility of the Web via the wireless network Higher Bit Rates for Data—GPRS and HDR Two of the more promising wireless technologies slated to become the main drivers of wireless data access are GPRS and HDR Whether they are true 3G service or mere stopgap 2.5G is up for debate, but both are able to provide significantly enhanced data rates, which will allow more diverse applications to be served over the wireless network GPRS GPRS builds upon the tremendous success of GSM technology, which, according to the GSM Association, claims more than 825 million users in over 193 countries as of March 2003 The major advantage of GPRS over the existing GSM data services such as circuit switched data (CSD) WIRELESS INTERNET—IS IT HAPPENING? and SMS include much enhanced speed, “always-on” connection, and spectrum efficiency Although the underlying GSM networks uses a circuitswitched architecture, GPRS uses packet-switching technology to achieve “always-on” connectivity and higher spectrum efficiency All eight timeslots in a frame can be used in GPRS to achieve the peak rate of 171.2 Kbps Unlike in CSD, the timeslots are not set aside to a user for the duration of the connection Instead, radio resources are used only when data are actually being transmitted Thus, the same channel can be shared among many users concurrently This efficient sharing of the limited bandwidth allows the network operator to maximize use of the limited radio resources, and lower the cost for the mobile users Since the existing GSM MSCs are based on circuitswitching technology, they cannot handle the operation of packet-switched GPRS connections Thus, two new network components called serving GPRS support node (SGSN) and gateway GPRS support node (GGSN) have been added to the GSM architecture The SGSN behaves much like the MSC, but for GPRS traffic It is responsible for delivering packets to the mobile users in its service area, and also handles queries to HLR for roaming operation The GGSN is the interface to the external networks, such as the Internet It maintains address and routing information necessary to tunnel data packets to appropriate SGSNs and onto the MSs This makes it possible to use existing IP applications over the GPRS network This and other advantages of GPRS already mentioned may help initiate the long-awaited proliferation of wireless Web access HDR HDR is based on CDMA technology and achieves a peak data rate of 2.4 Mbps Instead of sharing the channel with voice data as is done in GPRS, in HDR the entire frequency channel is allocated to data traffic By decoupling the data service from voice service, the overhead required to meet strict latency requirements of voice calls no longer degrades the system’s ability to handle packet data efficiently Large improvement in spectrum efficiency is achieved by measuring the signal-to-noise-plus-interference ratio between the BS and the MS, and adapting the modulation scheme and the forward error correction to achieve the optimal efficiency for the given channel condition Since only a single user is served at any given time (in units of 1.67-ms packets), there is no degradation in capacity due to inference among the MSs In addition, by taking advantage of varying and more relaxed latency requirements of data traffic, further increase in throughput can be achieved (Bender et al., 2000) The network architecture of HDR has been designed with the Internet in mind Selection of the point-topoint protocol (PPP) (Simpson, 1994) and the PPP multilink protocol (MP) (Sklower, Lloyd, McGregor, Carr, & Coradetti, 1996)were based on the need to support IP traffic with different QoS requirements while utilizing low overhead In addition, the radio link protocol (RLP) has been designed to achieve the level of data fidelity, i.e., bit error rate, which PPP and IP experience in wireline 839 networks This is important since many upper layer protocols, such as transmission control protocol (TCP), had been designed and optimized for the conditions observed in wireline networks Such network architecture and enhanced data rates are expected to improve and enhance the usability of wireless data services WIRELESS INTERNET—IS IT HAPPENING? Mobile IP (MIP) Mobility is not a feature that was incorporated into IP when it was conceived several decades ago Mobile IP consists of the necessary extensions needed to support mobility in the Internet (Perkins, 1997) IPv4, currently the most prevalent IP version on the Internet, has no provisions for mobility Each host computer on the Internet has a unique address in the hierarchical IP addressing space Each address consists of a network prefix and a host number The network prefix of the address determines the location (i.e., campus network) of the host computer Routers not contain the address of each individual host in their routing tables; instead, the network prefix of each address is used to forward packets to the next hop en route to the destination network To reach a host, you must know its IP address; if a host moves, all traffic addressed to it will be sent to its home location following the rules of network prefix routing (see Dynamic Routing for more details) If a host is assigned an address at its current new location, the sender must be made aware of it so that the appropriate destination address can be used in the data packets As the host moves, its IP address will change to reflect its new location This means that the sender must be informed of every change in location/address to maintain the connection and data flow Packets in transit will be lost unless some provision is made to have them follow the host to the next location For TCP connections the problem is further exacerbated as the IP address of the host is used in the TCP connection for session identification purposes Thus, if the host moves and changes its IP address, the connection identifier will no longer be valid, thereby causing the session to be terminated It is for these reasons that MIP was conceived For true mobility, the whereabouts and mobility of a host should not affect its ability to be reached by any sender, and in addition, the sender should not have to be responsible for tracking a remote host as it moves about the Internet When designing MIP, it was obvious that it had to be Compatible with the existing installed base of IP and the layers below it; Transparent to the layers above IP; Scalable and efficient, and capable of supporting large numbers of hosts and not impede the functionality of the Internet in anyway; and Secure, and the forwarding of connection control information must be authenticated to prevent traffic from being diverted to other destinations The design of MIPv4 accomplished the above four requirements By all means, it is not the most optimal solution 840 WIRELESS INTERNET Fortunately, IPv6 was designed with flexibility and mobility in mind and, as such, can support mobility in a more optimal/natural manner Unfortunately, IPv6 is not being adopted as quickly as expected, and is currently only being used in some small isolated locations and uses IPv4 tunnels for connectivity Below we will highlight some of the major features of MIPv4 MIP Terminology In RFC 2002 describing IP mobility support (Perkins, 1996), the following entities were identified in conjunction with MIP: Mobile node (MN)—A device that supports MIP and can change its location without affecting its communication abilities so long as layer connectivity is available Home agent (HA)—A device in the home network (i.e., the subnet to which the MN’s IP address belongs) of the MN that keeps track of the location of an MN It tunnels packets destined for the MN to its new address The HA is typically a router on the home network Foreign agent (FA)—A device in the current foreign network (i.e., the network that the MN is currently visiting) of the MN that can forward packets sent to it for the MN to the MN if it terminates the tunnel set up by the HA The FA is typically the default router on the foreign network Care of address (COA)—IP address that defines the current location of the MN It is the address to which the HA forwards all packets for the MN, thus terminating the tunnel There are two possible scenarios for COA: The COA is located at the FA; i.e., it is the IP address of the FA The FA then terminates the tunnel and forwards the packets to the MN This approach allows many MN to share one IP address The MN temporarily acquires a new IP address For this scenario, the MN terminates the tunnel Although a convenient approach, it does require that several IP addresses be made available for mobile devices, which may not always be the case This is referred to as a colocated COA Correspondent node (CN)—A device that communicates with the MN It is unaware of the location of the MN and just simply uses the MN’s original IP address for packet forwarding From the above we see that a tunnel starts at a HA and terminates at either the FA or the MN The HA keeps track of where the MN is The FA is not always needed for MIP functionality although it may be necessary for security purposes Operation of MIP The operation of MIP consists of three steps: agent discovery, registration, and encapsulation/routing/tunneling Below, we discuss each step: Agent discovery—Consists of broadcast messages used by the MNs to detect whether they have moved These messages are sent out periodically by the default router (FA) on a subnet If the MN has not heard an advertisement, it will solicit for one Registration—All MNs are required to register with the HA and the FA (if used) As registrations expire, MNs must re-register periodically Any move to a new location requires a new registration Encapsulation/routing/tunneling—All packets that arrive for the MN on the home subnet are claimed by the HA The HA proceeds to encapsulate them to reflect the new COA of the MN and then routes all the traffic to the MN on the foreign network The CN sends all messages to the MN’s IP address, and the HA relays them, via the tunnel, to the COA The MN sends its messages directly to the CN This type of communication results in what is referred to as triangular routing: CN to HA, HA to MN, MN to CN To improve the performance of MIP, route optimization has been proposed, which allows the CN to learn the COA of the MN and correspond with it directly This does mean that the CN must be informed of any change in location of the MN and requires additional authentication procedures Security in MIP Security is one of the main concerns in any mobile environment It is necessary that all devices involved with data reception and forwarding be authenticated to ensure their identities MNs must register periodically with the HA, which involves an authentication process If an MN moves, it must re-register at the new location If route optimization is used, the CN must authenticate itself before being capable of communicating directly with the MN TCP for Wireless Networks The TCP protocol was not designed to operate over channels that are lossy in nature It uses timers at the sender to determine the state of congestion in the network If an ACK does not come back before the timer expiration, it assumes that the link is congested and shuts itself down by decreasing the congestion window size to one It then proceeds to retransmit the unacknowledged packet TCP does not take into account the possibility of a lost packet due to channel conditions as a lost packet is directly interpreted as congestion Because of the proliferation of wireless networks, it is imperative that the transport layer understands the difference in performance of the wireless link and not make erroneous assumptions as to the state of the network Several papers that propose modifications to TCP to improve its performance over the wireless channels have appeared We describe some of these approaches below: Indirect TCP (I-TCP) (Bakre & Badrinath, 1995)—It segments the TCP connection into two portions: one for the wireline transmission, and one for the wireless segment It uses the traditional TCP over the wireline segment, and a modified version for the wireless segment Assuming the use of MIP, the FA is the most likely candidate for acting as the proxy and acknowledging all packets as well as terminating the connection Over the wireless segment, the proxy communicates with the WIRELESS INTERNET—IS IT HAPPENING? MN using local ACKs not relayed to the CN This mechanism shortens the retransmit time (most retransmissions occur only over the short wireless portion) However, should the FA fail, packets will be lost and neither side of the connection will be aware that this happened This scheme violates the end-to-end semantics of TCP Snooping TCP (Balakrishnan, Seshan, & Katz, 1995; Brewer et al., 1998)—This variation on TCP does not violate the end-to-end semantics of TCP as I-TCP does It resides on an intermediate node (such as a FA in the case of MIP) that is attached to the wireless link It buffers all data that are meant to be transmitted over the wireless channel so that if a loss occurs, it can retransmit the buffered data immediately, thereby avoiding the long end-to-end delay that otherwise would occur and cause the session to time-out Although it does not violate the end-to-end semantics, it is not as efficient as I-TCP as it does not completely hide the wireless link from TCP; the retransmissions incur a delay that could impact the performance of the end-to-end TCP session Mobile TCP (M-TCP) (Brown, & Singh, 1997)—It is similar in operation to I-TCP However, no data are buffered or retransmitted by the proxy The proxy only monitors the link, and if it determines that a loss has occurred, it will freeze the connection so that TCP does not go into slow start On the wireline side it will close down the window, forcing the sender to go into persistent mode, and on the wireless side it uses a fast recovery TCP that does not use slow start The end-to-end semantics are not violated as the proxy does not retransmit any data; it only interferes in the transmission by forcing certain TCP behavior on the detection of packet loss Other techniques include selective TCP (SACK) (Mathis, Mahdavi, Floyd, & Romanow (1996), which requires only the lost packets to be transmitted, and fast recovery/fast transmit (Caceres & Iftode, 1995), which does not allow TCP to go into slow start by sending three duplicate acknowledgments All of the solutions above provide some features to prevent TCP from entering the slow start mode due to packet loss on the wireless channel IEEE 802.11b/g/a 841 and an access point (AP), which acts as a bridge between the wireless and the wireline networks An access point usually consists of a radio, a wireline network interface (e.g., 802.3), and bridging software conforming to the 802.1d bridging standard The access point acts as the BS for the wireless network, aggregating access for multiple wireless stations onto the wireline network Wireless stations can be 802.11 PC Card, PCI, or ISA NICs, or embedded solutions in non-PC clients (such as an 802.11-based telephone handset) The 802.11 standard defines two modes: infrastructure mode and ad hoc mode In infrastructure mode, the wireless network consists of at least one access point connected to the wireline network infrastructure and a set of wireless stations This configuration is called a basic service set (BSS) An extended service set (ESS) is a set of two or more BSSs forming a single subnetwork Since most corporate WLANs require access to the wireline LAN for services (e.g., file servers, printers, and Internet links) they will operate in infrastructure mode Ad hoc mode (also called peer-to-peer mode or an independent basic service set, or IBSS) is simply a set of 802.11 wireless stations that communicate directly with one another without using an access point or any connection to a wireline network This mode is useful for quickly and easily setting up a wireless network where a wireless infrastructure does not exist or is not required or where access to the wireline network is barred 802.11 Physical Layer The three physical layers originally defined in 802.11 included two spread-spectrum radio techniques and a diffuse infrared specification The radio-based standards operate within the 2.4-GHz industrial, scientific and medical (ISM) band These frequency bands are recognized by international regulatory agencies, such as the FCC (USA), the ETSI (Europe), and the MKK (Japan) for unlicensed radio operations Spread-spectrum techniques, in addition to satisfying regulatory requirements, increase reliability, boost throughput, and allow many unrelated products to share the spectrum without explicit cooperation and with minimal interference The original 802.11 wireless standard defines data rates of and Mbps via radio waves using frequency-hopping spread spectrum (FHSS) or direct sequence spread spectrum (DSSS) Wireless LANs (WLANs) constitute an alternative technology to wireline LANs They use radio signals to transmit data Wireless technologies offer exceptional flexibility and mobility, thus making WLANs very attractive for many environments, including office, home, and public places IEEE 802.11 (Institute of Electrical and Electronics Engineers, 1999)-type technologies are the common standard used in WLANs Like all IEEE 802 standards, the 802.11 standards focuses on the lower two layers of the International Standards Organization (ISO) reference model, the physical layer and the data link layer 802.11b Enhancements to the PHY Layer The key contribution of the 802.11b addition to the wireless LAN standard was to standardize the physical layer support of two new access speeds, i.e., 5.5 and 11 Mbps To accomplish this, DSSS was selected as the sole physical layer technique for the standard since frequency hopping cannot support the higher speeds without violating current FCC regulations The implication is that 802.11b systems will interoperate with 1- and 2-Mbps 802.11 DSSS systems, but will not work with 1- and 2-Mbps 802.11 FHSS systems Operation of 802.11 802.11 Operating Modes 802.11 defines two pieces of 802.11 Data Link Layer The data link layer within 802.11 consists of two sublayers: logical link control (LLC) and media access control (MAC) 802.11 uses the same 802.2 LLC and 48-bit addressing as other 802 equipment, a wireless station, which is usually a PC equipped with a wireless network interface card (NIC), 842 WIRELESS INTERNET LANs, allowing for very simple bridging from wireless to IEEE wireline networks, but the MAC is unique to WLANs For 802.3 ethernet LANs, the carrier sense multiple access with collision detection (CSMA/CD) protocol regulates how ethernet stations establish access to the wire and how they detect and handle collisions that occur when two or more devices try to simultaneously communicate over the physical medium In an 802.11 WLAN, collision detection is not possible due to the near/far problem To detect a collision, a station must be able to transmit and listen at the same time, but in radio systems the transmission (near signal) drowns out the ability of the station to detect a collision (far signal) Since collision detection is not possible, the stations only use collision avoidance; they sense the channel before transmitting If the channel is busy, stations back off and try again at a later time If the channel is idle, a station will transmit its frame Since several stations may be sensing the channel at the same time and all detect it to be idle, they will start to transmit concurrently, thereby causing collisions Because stations are unable to detect collisions, CSMA/CA systems need to use explicit packet acknowledgments (ACK) In other words, an ACK packet is sent by the receiving station to confirm that the data packet arrived intact Another MAC-layer problem specific to wireless is the hidden-terminal issue, in which two stations can both hear activity from the access point, but not from each other, usually due to distance or a physical obstruction To solve this problem, 802.11 specifies an optional request to send/clear to send (RTS/CTS) protocol at the MAC layer When this feature is in use, a sending station transmits an RTS and waits for the access point to reply with a CTS Since all stations in the network can hear the access point, the CTS causes them to delay any intended transmissions, allowing the sending station to transmit and receive a packet acknowledgment without any chance of collision Since RTS/CTS adds additional overhead to the network by temporarily reserving the medium, it is typically used only on the largest-sized packets, for which retransmission would be expensive from a bandwidth standpoint Security 802.11 provides for MAC layer (OSI Layer 2) access control and encryption mechanisms, which are jointly known as wired equivalent privacy (WEP), with the objective of providing WLANs with security equivalent to their wireline counterparts For the access control, the ESSID (also known as a WLAN service area ID) is programmed into each access point A wireless client must know the ESSID to associate with an access point No communication can occur unless there is an association between a client and the access point In addition, there is provision for a table of MAC addresses called an access control list to be included in the access point, restricting access to only those clients whose MAC addresses are on the list For data encryption, the standard provides for optional encryption using a 40-bit shared-key RC4 PRNG algorithm from RSA Data Security All data sent and received while the end station and access point are associated can be encrypted using this key In addition, when encryption is in use, the access point will issue an encrypted challenge packet to any client attempting to associate with it The client must use its key to encrypt the correct response in order to authenticate itself and gain network access Unfortunately, beginning with an internal study in 2000 (Walker, 2000) to a highly publicized study in 2001 (Borisov, Goldberg, & Wagner 2001), WEP has been shown to fall well short of accomplishing its security goals Some of the problems of WEP that have been identified by researchers include the following: WEP uses RC4, a synchronous stream cipher, but it is difficult to ensure synchronization during a complete session over the unreliable wireless link, leading to the use of a separate key for each packet—a clear violation of one of the most important requirements of RC4 A very limited key-space is used, which is problematic since a separate key is needed for each packet 802.11 does not provide any mechanism for sharing keys over an insecure channel There is no mechanism for a mobile to authenticate the network Checksum (CRC-32) used for integrity check is linear; thus, it is relatively easy to make undetected changes in the message Such weaknesses combine to result in a network that is vulnerable to several types of attacks and intrusions There are several ongoing efforts to secure the 802.11 network, one of which is the robust security network (RSN) In RSN, a recently proposed 802.1x standard (Institute of Electrical and Electronics Engineers, 2001) forms the basis for access control, authentication, and key management In addition, a number of protocols such as extensible authentication protocol–transport layer security (EAP-TLS) (Aboba & Simon 1999; Diersk & Allen, 1999; Zorn, 1999) are being considered to provide strong authentication between the MS and the AP Timing and Power Management Synchronization of all clocks within a BSS is maintained by periodic transmission of beacons containing timestamp information In the infrastructure mode, the AP serves as the timing master and generates all timing beacons Synchronization is maintained to within ms plus propagation delay Timing beacons also play an important role in power management There are two power saving modes defined: awake and doze In the awake mode, stations are fully powered and can receive packets at any time In the doze mode, it is unable to transmit or receive data and consumes very little power A station must inform the AP that it is entering the doze mode The AP does not send packets to stations in the doze mode, but instead buffers them for transmission at a designated time Comparison of 802.11b and 802.11a Two advanced WLAN standards, 802.11b and 802.11a, were developed by the IEEE’s 802.11 working group At WIRELESS INTERNET—IS IT HAPPENING? Table Features of 802.11b Advantages 2.4-GHz band is almost universally available Table Comparison of 802.11b and 802.11a Technologies Disadvantages Prone to interference from other devices that operate in the 2.4-Ghz band the MAC layer, they both use the CSMA/CA protocol At the physical layer, 802.11b uses the DSSS radio transmission method and operates in the 2.4-GHz ISM band (see Table 3), while 802.11a uses orthogonal frequency division multiplexing (OFDM)—a more recent modulation scheme that is claimed to offer better performance at higher data rates (Bingham, 1990)—and operates in the 5-GHz UNII (unlicensed national information infrastructure) band (see Table 4) 802.11b—Offers data rates of up to 11 Mbps (see Table 5) 802.11a—Offers data rates of up to 54 Mbps due to higher carrier frequency and a more sophisticated encoding technology (see Table 5) 802.11g While the 802.11g standard is yet to be finalized (draft standard was ratified in November 2001), it seeks to offer current users of 802.11b data rates up to 54 Mbps in the same 2.4-GHz band Because 802.11b and 802.11g use the same frequency band, an 802.11b radio interface will work with an 802.11g access point albeit at an 802.11b rate However, the physical range of 802.11g will be less than 802.11b, so higher concentration of access points will be necessary to obtain the full 54-Mbps rate throughout the service area Two mandatory and two optional modes are part of the draft standard Use of OFDM (similar to one in use for 802.11a) is mandatory for data rates greater than 20 Mbps and support for complementary code keying (CCK) is necessary to ensure backward compatibility with existing 802.11b radios The RTS/CTS mechanism described previously is used to ensure that both OFDM and CCK can coexist in the same 2.4-GHz band An additional benefit of 802.11g is that since OFDM is already required for 802.11a, it is possible to build dual-band Table Features of 802.11a Advantages Better performance in office environment (multipath reflection recovery) Higher data rates and less congestion in the UNII band Some chip makers promise proprietary modes that will deliver up to 72 Mbps 843 Disadvantages Limited number of channels available outside the United States Standard 802.11a 802.11b Speed, up to Range Radio technology MAC protocol Frequency Power management 54 Mbps 300 feet OFDM CSMA/CA GHz 11 Mbps 300 feet DSSS CSMA/CA 2.4 GHz Embedded power saving protocol (2.4 and GHz) radios without extra hardware complexity The popularity of IEEE 802.11a/b/g makes it a very serious challenger for 3G It is being deployed in many public places to provide high-speed access to mobile users The ease of deployment and maintenance of 802.11 is at the heart of its success It is cheap (compared to 3G), robust, and simple to use, three cornerstones of any networking technology Bluetooth and PANs The concept of personal area networks (PANs) has been introduced to enable wireless communication between devices in direct vicinity of a user Examples of such devices are laptop computers, personal digital assistants (PDAs), cellular phones, printers, and photo cameras PAN technology is characterized by low cost, low power consumption, and ad hoc network organization Wireless PANs are being standardized in the IEEE 802.15 working group The Bluetooth technology is being promoted as an industry standard for PANs, and is forming the basis for IEEE 802.15 standardization Bluetooth is a low-power, low-cost wireless technology for distances up to 10 m, at data rates up to Mbps (and higher in newer versions of the standard) (Haartsen, 1998) It operates in the same 2.4-GHz ISM band as the IEEE 802.11b standard, using fast (1600 hops/s) FHSS Bluetooth nodes are organized in so-called piconets, consisting of a master node and up to slave nodes (that number is now being increased to 25 in the newer version of the standard) A slave can simultaneously be a slave or a master in another piconet, with a different frequency hopping sequence This allows the construction of larger networks, called scatternets, where communication between nodes is carried out using multiple hops across piconets Within piconets, the Bluetooth MAC uses polling to regulate access to the radio interface A slave is only allowed to transmit a slot of data when polled by the master Transmission of data from master to slave is considered as an implicit poll Using this polling scheme, Bluetooth can provide both synchronous connection-oriented (SCO) links, e.g., for support of voice telephony applications, and asynchronous connectionless (ACL) links, e.g., for IP packet transfer Bluetooth products are rapidly being introduced into the market Many new cellular telephone models are 844 WIRELESS INTERNET MSC CS-domain PSTN/ ISDN HLR M S RAN GSN PS-domain IP Network Core Network Figure 3: A 2.5G cellular network equipped with the technology as are some new digital cameras Furthermore, products such as headsets, PCMCIA cards for laptops and PDAs are appearing on the market 3G CELLULAR SYSTEMS Second generation cellular networks are possessed with very limited data capabilities Apart from special vertical services such as the SMS, these capabilities are restricted to circuit-switched data up to a data rate of 14.4 Kbps 2G cellular networks have been extended later with enhanced data capabilities The GSM standard, for instance, has been extended with a packet-switched mode, GPRS, where timeslots can be assigned toMSs on an ondemand basis Also, an MS can combine several timeslots per frame, in order to achieve higher data rates (theoretically up to 171.2 Kbps) This GPRS is much better suited for the support of end-to-end IP-based services For the migration to GPRS, only few changes are required in the GSM base stations The core network consisting of circuit-switched MSCs, on the other hand, needs to be enhanced by a packet-switched network In the core network the switches (MSCs) are augmented by specialized routers, the so-called GSNs From this so-called 2.5 generation on, the core network of a cellular system consists of a circuit-switched part with MSCs, and a packetswitched part with GSNs (see Figure 3) The major change going from 2.5G to 3G cellular networks is a complete new RAN For 2G and 2.5G cellular systems, a number of incompatible standards and systems are used throughout the world For 3G, the International Telecommunications Union (ITU) has set up IMT 2000 (International Mobile Telecommunications 2000) as a framework for worldwide wireless access by linking the diverse system of terrestrial and/or satellite based networks (ITU-T Recommendation Q.1701, 1999) The vision for IMT 2000 is to support advanced applications by providing higher data rates, from 384 Kbps global coverage to Mbps indoor or low-range outdoor coverage Further, the systems should be highly flexible, providing support for both applications that traditionally use circuit-switched networks and applications that traditionally use packet-switched networks A wide range of date rates is to be supported, with a high granularity IMT 2000 comprises a number of cellular systems (De Vriendt, Lain´ , Lerouge, & Xu, 2002) First, the e CDMA2000 system is an evolution from the American CDMA system, IS-95 Further, a number of IMT 2000 systems are defined by (re)using the core network from GSM Three different radio technologies have been defined for use with this core network EDGE (enhanced data rates for GSM evolution) is an evolution from the GSM (and GPRS) technology, using new modulation techniques, to provide data rates up to 384 Kbps The two other radio technologies have been developed in the context of UMTS (universal mobile telecommunication system), and are based on CDMA principles The main difference between the two is the duplexing technique used Duplexing the two directions of communication can be done in either frequency or time, resulting in an FDD and a TDD UMTS variant, respectively The first one, which is currently most widely implemented, uses a technique called wideband CDMA (WCDMA), using 5-MHz carriers to provide data rates up to Mbps WCDMA and CDMA2000 both rely on CDMA technology, but differ from each other in various design and implementation aspects, such as clock and CONVERGENCE OF IP AND CELLULAR SYSTEMS—TOWARD THE MOBILE INTERNET 845 IP Network HLR M S BS RNC SGS GGS Backbone Network IP Figure 4: Architecture of a UMTS network chip rates, synchronization approach, pilot channels, and frame duration, and are incompatible with each other Both WCDMA and CDMA2000 provide the user with channels, either shared with other users or dedicated, with different maximum data rates, depending on the spreading factor used It provides a wide range of channel bit rates, and is very well suited for bursty packetbased traffic These radio standards offer quite some room for further improvement New modulation techniques, more advanced scheduling, dynamic link adaptation, and multiple-input-multiple-output antenna techniques are some of the improvements currently under study (Honkasalo, Pehkonen, Niemi, & Leino, 2002) The high-speed downlink packet access channel (HSDPA) in WCDMA, with data rates up to 10 Mbps, has been standardized as a first step in this direction 2.5G systems are widely available now throughout the world, although the use of these systems is not (yet) overwhelming 3G systems are currently starting service in limited areas, especially in Japan (WCDMA), and are being expanded in functionality, performance, and coverage Availability of terminals is still very limited CONVERGENCE OF IP AND CELLULAR SYSTEMS—TOWARD THE MOBILE INTERNET It is expected that the future will show a convergence between the Internet and cellular systems They will converge both in the services provided and in the technology used The converged system will combine the wide range of horizontal services offered by the Internet with mobility and highly integrated services and devices offered by the cellular networks As for the technology, it is expected that future systems will combine radio technology from cellular systems with switching and routing technology based on the IP protocol suite Looking at a possible net- work architecture of a UMTS network (Figure 4), we see two distinct IP networks One is the external IP network, the other an IP-based transport network The external IP network is a network providing services and connectivity from remote end-users to the user of the MS For the MS, the GGSN is a gateway to the external IP network, and provides the MS with an IP address valid on that network IP headers of packets on their way from the MS to the external IP network (or vice versa) are only processed by the GGSN, not by all the intermediate nodes From the point of view of the external IP network, the UMTS network is a single subnetwork The IP-based transport network is typically owned by the cellular operator, and provides connectivity between the various nodes in the UMTS network These are, for the packet-switched core network, the SGSN and the GGSN, and for the RAN, the BSs and Radio Network Controllers (RNCs) The BS provides the main radio (physical layer) functions, whereas the RNC provides the higher layer functions, including radio resource management, and soft handover For the packet-switched core network, it has been specified that the SGSNs and GGSNs should be interconnected by an IP-based transport network User-level IP packets are encapsulated and tunneled using the so-called GPRS tunneling protocol (GTP) on top of the transport level IP GTP provides for the routing of the user level IP packets to the appropriate SGSN, using the database of the HLR, for mobility management functions As such, GTP provides an alternative to MIP In the first release of UMTS, ATM is used as transport technology for the RAN However, it is foreseen that IP will also be introduced in this part of the cellular network in the near future The use of IP as transport technology in a RAN, besides enabling the operators to provide new packet-based services, is especially beneficial since it provides the means for statistical aggregation of traffic, which leads to increased transmission efficiency and 846 WIRELESS INTERNET consequently reduced leasing costs A radio frame is a short data segment coded/decoded and transmitted/received by the base station These radio frames must be delivered from RNC to BS and vice versa in a timely fashion with limited delay Otherwise, the BS or RNC will discard them Due to the time constraints on the delivery of radio frames the majority of the traffic in an IP-based RAN can be considered to be real-time traffic Seen from the architecture of a RAN and the nature of the transported data, the IP-based RAN has different characteristics when compared to traditional IPnetworks Typically, the wireline transmission in a radiobreak access network contains a relatively high volume of leased lines The fact that thousands of radio BSs are spread over a wide geographical area and are in general situated at large distances from the backbone typically results in high cost for the transmission links Further, the majority of the traffic transported on the wireline transmission links used by the RAN is radio frames This means that the traffic is very sensitive to delays and delay variation (jitter) Deploying resource management schemes in this environment is therefore essential The introduction of IP-based transport in the RAN indicates that an IP QoS-capable domain will have to be managed in the radio access network Currently, DiffServ (Blake et al., 1998) as a scalable IP QoS architecture is the favorite one to be used in an IP-based RAN The scalability is achieved by offering services on an aggregate basis rather than per flow and by forcing as much of the per-flow state as possible to the edges of the network, that is, to the edge nodes In order to allow for dynamic resource management in DiffServ, an extension, called RMD (resource management in DiffServ), has been proposed (Heijenk, Karagiannis, Rexhepi, & Westberg, 2001) RMD extends the DiffServ architecture with new reservation concepts and features, such that the IP-based RAN resource management requirements are met These trends will lead to a cellular network architecture, where all nodes are interconnected using an QoSenabled IP-based transport network On top of this IP protocol, protocols for cellular specific functions, related to mobility and radio, will be running Further, an endto-end IP protocol will run on top of this to enable endusers to use IP-based services, and to connect to other end-users Currently, IP-based services over cellular networks are mainly best-effort type of services, both interactive, such as Web browsing, and background, such as e-mail downloading For the near future, IP-based services might be extended with more streaming type of services for audio and video A situation where all services, including conversational services, in a cellular network are IP-based is somewhat further away Prerequisites for such a situation are very efficient header compression techniques, and a migration to a cellular network architecture where also signaling is IP-based, e.g., based on SIP Efforts in these directions are being made in both research and standardization Cellular networks provide wide-area coverage for mobile users at moderate data rates Besides cellular networks, other wireless systems are gaining popularity, in particular WLAN and short-range technologies WLAN systems provide wireless ethernet extension to notebooks WLANs based on the IEEE 802.11b standard (Institute of Electrical and Electronics Engineers, 1999) and operating in the 2.4-GHz ISM band are widely available in the market They offer data rates up to 11 Mbps and have a range of 50 to 300 m New products operating in the 2.4-GHz ISM and 5-GHz band offering data rates up to 54 Mbps are starting to appear on the market These systems have been primarily designed for nomadic applications and, consequently, their support for mobility is very poor Short-range technologies, e.g., Bluetooth, have a more limited coverage, support lower data rates, and consume less power than WLANs (Haartsen, 1998) Operating in the ISM band, these technologies originally designed to replace the cabling connecting peripherals and other devices are also suitable for inexpensive communication between portable devices The above-mentioned technologies have been optimized for different applications Even as they evolve, it is not expected that they will be replaced by a common multipurpose technology in the future On the contrary, it is expected that the next generation wireless systems will integrate different and complementary technologies Wireless devices able to operate with different radio interfaces will access the communication facilities using the “always best connected” paradigm This means that a wireless device that has the choice will use that radio interface that it deems most appropriate for its purposes, e.g., highest performance or lowest cost In the future, a person might use multiple personal devices, such as laptop, phone, and organizer, that are mutually interconnected, forming a PAN (see Figure 5) using, for example, Bluetooth technology These devices will all have one or more wireless interfaces At a certain moment, the person might use his laptop and be connected to his company network via a wireless LAN interface When moving out of the office, he may want to stay connected using a Bluetooth link between his laptop and phone, where his phone will act as an intermediate hop, via a UMTS link to the fixed network In this scenario, the PAN acts as a moving network with multiple interfaces, which moves along different wireless APs to the fixed network, and which may merge with another moving network, e.g., the network in a vehicle The challenge in these scenarios is to achieve seamless integration, meaning that from the point of view of the application, switching from one network technology to the other is imperceptible and the level of security is maintained This requires measures to be taken at different levels of the protocol stacks Suitable techniques for supporting mobility and smooth transitions between different technologies and systems as well as in between private and public networks are under study (Wireless World Research Forum, 2001) Even as wide deployment of 3G has been experiencing delays, 4G wireless technology is in an active research stage 4G is intended to provide much of what 3G had originally envisioned, i.e., a broadband cellular service providing high-speed capacity at low cost, along with IP-based applications and services Data rates of up to 20 Mbps are targeted, even as the MS moves at up to 200 km/h, and GLOSSARY 847 Fixed IP Network UMTS WLAN Bluetooth Figure 5: Depiction of a personal area network (PAN) the entire network will use packet-switching techniques However, because there is not a single 3G standard upon which to build 4G, there are significant challenges ahead To support such high data rates, significant advances in baseband processing are necessary One of the most widely considered transmission schemes is multicarrier modulation (MCM), which uses many parallel subchannels to transmit information MCM’s advantages include better protection against intersymbol interference (ISI) and implementation efficiencies possibly using fast Fourier transform (FFT) techniques At least two different types of MCM are being considered for 4G, including multicarrier CDMA (MC-CDMA) and OFDM (recall that OFDM is already used in several wireless LAN standards such as 802.11a and 802.11g) Regardless of the multiplexing scheme, techniques such as smart antenna and multiuser detection will become integral parts of the standard, as they enable the technology necessary to support the requirements of 4G Additional information about 4G issues may be found in Lu (2002), Lu and Berezdivin (2002), and Mahonen and Polyzos (2001) ACKNOWLEDGMENTS We thank the following colleagues for their assistance with the authoring of this chapter: Jasmine Zan, Vlora Rexhepi, Georgios Karagiannis, and Sonia Heemstra de Groot We also are very appreciative of all the insightful comments from the reviewers of this chapter and the guidance provided to us by Dr H Bidgoli GLOSSARY ATM A network technology designed to transport cells, where unlike IP, cell sizes are fixed and circuitswitching circuits are used Autonomous system (AS) A network or a group of networks operated by a common administrator CDMA A digital cellular technology that uses spreadspectrum techniques to multiplex numerous concurrent connections Cellular network A wireless network that divides geographical regions into cells, to which subsets of available frequency bands are allocated Duplexing Transmission of data in two directions, from node A to node B and vice versa Encapsulation The process of embedding an additional network protocol on top of the existing one(s) possibly for the purposes of tunneling Hop A connection between two adjacent network devices such as a router Hypertext markup language (HTML) A formatting language used in creating documents for the World Wide Web Hypertext transport protocol (HTTP) A transport protocol used in the World Wide Web to deliver documents to browsers Internet A global network of computers and smaller networks Internet protocol (IP) A specification for data format and addressing scheme used on the Internet ... DS4 1. 544 2.048 6. 31 2 8.448 32 .064 34 .36 8 44. 736 97 .728 13 9. 264 274 .17 6 DS, North America; E, Europe; J, Japan Number of DS0s 24 32 96 12 8 512 672 2,048 4, 032 Designation OC -1 OC -3 OC -9 OC -12 ... OC -12 OC -18 OC-24 OC -36 OC-48 OC -96 OC - 19 2 Line rate SDH equivalent 51. 840 Mbps 15 5.250 Mbps 466.560 Mbps 622.080 Mbps 93 3 .12 0 Mbps 1. 24 416 Gbps 1. 86624 Gbps 2.48 832 Gbps 4 .97 664 Gbps 9. 9 532 8 Gbps... Encryption of Network Transmissions Routing and Remote Access Service (RRAS) 792 792 792 7 93 797 797 797 798 798 799 799 799 WHAT IS W2K? Microsoft’s Windows 2000 (W2K) is an operating system product