services in WiMAX

Tài liệu tham khảo ngành viễn thông services in WiMAX

2.2.2 Overview of Admission Control 24

2.2.3 Admission Control Policy 24

2.3 Services and service flows 26

2.3.1 Connections and service flow 26

2.3.2 Connection Identifiers (CIDs) 27

2.3.3 Service Flows 29

2.4 QoS architecture model 35

Chapter 3 Scheduling and Admission for Real-Ttime Traffic 39

3.1 Scheduling and admission for real-time traffic 39

3.1.1 System models [18] 40

3.1.2 Loss rate for preemptive EDF 44

3.1.3 Non-preemptive EDF 47

3.2 Some current scheduling algorithms for real-time polling services 49

3.2.1 EDF Broadband Wireless Access Scheduling Algorithm 51

3.2.2 Admission Control 54

Chapter 4 Simulation Results 57

4.1 Theoretical Performance of single queue EDF scheduling algorithms 57

4.2 Simulation of NP-EDF scheduling algorithm for rtPS services in WiMAX 59

4.3 Conclusion 61

References 62

Nowadays, the demand for Internet broadband access is growing rapidly,which results in lots of new standards of accessing Internet broadband Together withthe increasing development of traditional wired broadband networks, wirelessnetwork access is expanding more and more Wireless broadband access standardIEEE 802.16 came into existence as a result of this fact IEEE 802.16 standards areestablished for Wireless Metropolitan Area Network- WMAN, however, the main partof 802.16 packet scheduling still remains unspecified This thesis, thus, reviewsanalytical methods to evaluate the efficiency of real time system models with the use ofsingle- server queue that have been used by many researchers The service principlein the queue is EDF (Earliest Deadline First) Real-time jobs with exponentiallydistributed deadlines arrive according to a Poisson process, all jobs have deadlinesuntil the end of service The thesis also introduces a non-preemptive EDF schedulingalgorithm and admission control for real-time polling services in WiMAX

It has been an honor that I have had the chance to study in the field that I havebeen taught and interested in It was obvious that I had to try my best to finish mywork, but without the help of many people, this could not have been completed.

Firstly, I would like to thank Prof Dr Nguyen Dinh Thong from the Universityof Technology, Sydney, for his helpful and enthusiastic instructions throughout theperiod of time that I worked on this thesis

I am also very grateful to Mr Nguyen Quoc Tuan, M.E of the Department ofTelecommunications, College of Technology, Vietnam National University Hanoi, forhis sincere help and providing the best setting within the time I studied and researchedin the Faculty of Electronics and Telecommunications.

Next, I am very thankful to Dr Nguyen Thi Minh Tam, Ms Nguyen Hai Ha,B.A and Ms Tran Thanh Thu, B.A from the College of Foreign Languages, VietnamNational University Hanoi, who have helped me correct the English of this thesis.

I would also like to thank all my teachers and peers that helped me during thetime that I have spent here at the university

Last but not least, I also want to give my sincere thanks to my family who haveconstantly supported me during my four years studying far away from home.

Hanoi, date month year 2008 Nguyen Minh Khue

List of figures

Figure 1.1: Simple scheduling architecture……… 15

Figure 2.1: TDD frame structure……… 23

Figure 2.2: Correspondence between the CID and SFID……….28

Figure 2.3: Illustration of service flows and connections……….28

Figure 2.4: Classification and CID mapping……… 32

Figure 2.5: Header suppression at the sending entity ……… 33

Figure 2.6: Header suppression mechanism at the receiving entity……… 33

Figure 2.7: Illustration of PHS operation……… 35

Figure 2.8: DSC MAC management message used for the signalling of a PHS rule……… 36

Figure 2.9 Quality of Service model in IEEE 802.16 ……… 37

Figure 2.10 Service flow transition diagram……….39

Figure 2.11 Medium Access Control architecture of the Base and Subscriber Station 40

Figure 3.1 State-transition-rate diagram for Markov chain M………43

Figure 3.2 The modified view to the non-preemptive EDF queue …………50

Figure 3.3 Hierarchical structure of bandwidth allocation [24]………55

Figure 3.4 rtPS database structure in scheduling database module [24] ………… 56

Figure 3.5 Token Bucket mechanism ……… 57

Figure 4.1 P-EDF analytic with µθ=4……… 59θ=4……… 59

Figure 4.2 P-EDF analytic with µθ=4……… 59θ=8……… 60

Figure 4.3 Loss probability analytic….……….61

Figure 4.4 Bandwidth allocation for rtPS connection ……… …62

Figure 4.5 Arrival and service curve for rtPS connection ……… 63

List of tables

Table 1: CID ranges as defined in Reference IEEE 802.16-2004……… 29Table 2: Possible values of the PHS support field……… 33

Base Station (BS): Generalized equipment set providing connectivity, management,

and control of the subscriber station (SS).

Broadband: having instantaneous bandwidths greater than around 1MHz and

supporting data rates greater than about 1.5Mbps.

Broadband Wireless Access (BWA): Wireless access in which the connection(s)

capabilities are broadband.

Burst Profile: set of parameters that describe the uplink or downlink transmission

properties associated with an interval usage code Each profile contains parameterssuch as modulation type, forward error correction (FEC) type, preamble length, guardtimes, etc.

Connection: A unidirectional mapping between base station (BS) and subscriber

station (SS) medium access control (MAC) peers for the purpose of transporting trafficflow of a service Connection are identified by a connection identifier (CID) Alltraffic is carried on a connection even for service flows that implement connectionlessprotocols, such as Internet Protocol (IP).

Connection Identifier (CID): A 16-bit value that identifiers a connection to

equivalent peers in the MAC of the base station (BS) and subscriber station (SS) Itmaps to a service flow identifier (SFID), which defines the Quality of Service (QoS)parameters of the service flow associated with that connection.

Downlink: The direction from the base station (BS) to the subscriber station (SS).Downlink Map (DL-MAP): A MAC messages that defines burst start times for both

time division multiple and time division multiple access (TDMA) by a subscriberstation (SS) on the downlink.

Frame: A structured data sequence of fixed duration used by some PHY

specifications A frame may contain both an uplink subframe and a downlinksubframe.

Frequency Division Duplex (FDD): A duplex scheme in which uplink and downlink

transmissions use different frequencies but are typically simultaneous.

Protocol Data Unit (PDU): The data unit exchanged between peer entities of the

same protocol layer In the downward direction, it is the data unit generated for the

next lower layer On the upward direction, it is the data unit received from theprevious lower layer.

Service Access Point (SAP): The point in a protocol stack where the services of a

lower layer are available to its next higher layer.

Service Data Unit (SDU): The data unit exchanged between two adjacent protocol

layer On the downward direction, it is the data unit received from the previous higherlayer On the upward direction, it is the data unit sent to the next higher layer.

Service Flows (SF): A unidirectional flow of medium access control (MAC) service

data units (SDUs) on a connection that is provided by a particular Quality of Service(QoS).

Service Flow Identifier (SFID): A 32-bit quantity that uniquely identifiers a service

flow to both the subscriber station (SS) and base station (BS).

Subscriber Station (SS): Generalized equipment set providing connectivity between

subscriber equipment and a base station.

Time Division Duplex (TDD): A duplex scheme where uplink and downlink

transmissions occur at different times about may share the same frequency.

Uplink: The direction from a subscriber station to the base station

Uplink MAP (UL-MAP): A MAC messages that defines the uplink usage in terms of

the offset the burst relative to the allocation start time (AST).

Acronyms

IEEE Institute of Electrical and Electronic Engineers

RTG Receive/Transmit Transition Gap

Chapter 1.

Introduction to WiMAX Broadband Wireless Access and QoS Scheduling

There has recently been a considerable growth in demand for high-speedwireless Internet access, which has caused the emergence of new wireless technologiesfor short distances (i.e IEEE 802.11) and also wireless technologies for long andmedium distances (i.e IEEE 802.16).

Wireless technologies for long and medium distances, in particular IEEE802.16, offer another option apart from the current wired access networks such ascable modem and digital subscriber line (DSL) links The IEEE 802.16 has become ahighly applicable option, as it can be deployed rapidly even in areas where wiredinfrastructures could hardly reach Moreover, it covers broad geographical area in amore economical and time efficient manner than traditional wired systems.

In comparison with 802.11 standard, 802.16 can serve a much greater numberof simultaneous users and approximately 50 times greater (radial) coverage The IEEE802.16a standard has a range of up to 30miles with data transfer speeds of up 70 Mbps.At the same time, customer demand for quality of service (QoS) hassignificantly increased as a result of the growth in broadband wireless access (BWA).BWA p providers are critically concerned about the provision of QoS The IEEE802.16 standards appear to offer a solution to this problem, by establishing a numberof unique and guaranteed QoS parameters in terms of delay, jitter and throughput Thisenables service providers to offer flexible and enforceable QoS guarantees, the benefitthat has never been available with other fixed broadband wireless standards [1].

The IEEE 802.16 has great potential in the broadband market Recentinvestment in Australia [2] indicates that IEEE 802.16 may become prevalent in thecoming years IEEE 802.16 standard are now supported by both the IEEE and theEuropean Telecommunication Standard Institute (ETSI) HiperMan standard Thechanges adopted by either of the two bodies are being reflected in the baselinetechnical requirements It is also supported by parallel interoperability efforts throughthe industry forum known as Worldwide Interoperability for Microwave Access(WiMAX).

1.1 Quality of Service Fundamentals

Integrated telecommunication networks carry traffic of several different classes,including one or more real-time traffic classes, each with its own set of trafficcharacteristics and performance requirements Two different approaches have beendeveloped to deal with this phenomenon The first approach is circuit-switched, inwhich sufficient resources are allocated to each connection to handle its peak rate Thisguarantees that the connection will obtain the quality of service (QoS) it requires, butat the cost of under-utilizing the network resources The second popular approach isthe packet-switched approach, in which traffic from all sources is broken into packetsand statistical multiplexing techniques are used to combine all the network trafficthrough single switching nodes This allows higher network utilization, but requiresmore sophisticated controls to ensure that the appropriate QoS is provided [3].

Packet-switched networks were originally designed to provide best effortservices, but later the demand for differentiated services caused the packet-switchednetworks to be integrated with QoS architecture QoS architecture introduces tools totreat packets in different ways, for example, real-time packets will be given priorityover non-real-time packets allowing them to traverse the network faster and arrive atthe destination within their required delay bounds

QoS in packet-switched networks can be characterized in terms of a specific setof parameters including delay, delay jitter, bandwidth and loss or error rate Delay isthe time that it takes for the packet to traverse from source to destination, it consists oftransmission delay, propagation delay, and queuing delay in intermediate routers.Delay jitter is the fluctuation or variation in end-to-end delay from one packet to thenext within the same packet flow Bandwidth is a measure of the amount of data that anetwork allows one flow to transmit over a period of time Drop rate of one flowmeasures the number of packets which is dropped due to buffer overflow, transmissionerror or expiry caused by over-staying than their maximum delay bound

The ability to manage congestion and maintain QoS in a packet-switched networkrequires the collaboration of many components in the QoS architecture The threemain components are as follows:

 Admission Control determines whether a new request for resources can be

granted or not based on the knowledge of total network capacity and the alreadyaccepted flows It has a critical role of limiting the number of flows admittedinto the network so that each individual flow obtains its required QoS.

 Packet Scheduling is a critical component in any QoS architecture, as the

packets traverse different switches (or routers) along their way through thenetwork It determines the order in which packets belonging to different flowstransmit on the output link, thus ensuring that packets from different applicationmeet their QoS constraints The basic scheduler architecture is shown in figure1.1, in which several inputs are buffered and there is a single scheduler (server).An optimal scheduling mechanism will provide the necessary QoS guaranteesrequired by different classes of traffic while efficiently utilizing the networkresources.

 Buffer Management has the responsibility of discarding one or more in

coming packets before the output buffer overflows, in order to improve theperformance of the network One of the most common packet drop strategies isRandom Early Detection (RED) [4].

Figure 1.1 Simple scheduling architecture.

1.2 QoS in IEEE 802.16

The IEEE 802.16 is designed with the aim of providing a guaranteed QoS Thereare a number of features included in the current standard However, the details of someof these features are still unspecified In the section, we describe initially the currentQoS architecture of the IEEE 802.16 and subsequently the unspecified features.

In IEEE 802.16, the QoS architecture is part of a MAC layer For wirelessnetworks it is natural to integrate the QoS architecture with the MAC protocol TheMAC protocol coordinates communication over the shared wireless medium The802.16 MAC provides QoS differentiation for different types of applications thatmight operate over 802.16 networks IEEE 802.16 defines five classes of service to

support the QoS in 802.16-2004 fixed WiMAX and one extra class for 802.18e-2005Mobile WiMAX [5].

 Unsolicited Grant Service (UGS) is designed to support real-time data

streams consisting of fixed-size data packets issued at periodic intervals,such as T1/E1 and Voice over IP without silence suppression UGS isprohibited from using any contention requests, and there is no explicitbandwidth request issued by subscriber station (SS) The base station (BS)provides fixed size access slots at periodic intervals to the UGS flows.However, the reserved bandwidth may be wasted when a correspondingUGS flow is inactive.

 Real-Time Polling Service (rtPS) is designed to support real-time data

streams consisting of variable-sized data packets that are issued at periodicintervals, such as moving pictures expert group (MPEG) video Themandatory QoS service flow parameters for this scheduling service areminimum reserved traffic rate, which is defined as the minimum amount ofdata transported on the connection over period of time and maximumlatency, which is the upper bound on the waiting time of a packet in thenetwork The rtPS flows are polled by BS through a unicast request pollingfrequently enough to meet the delay requirement of the service flows.

 Extended Real-Time Polling Service (ertPS) The ertPS is designed for

realtime traffic with variable data rate (such as VOIP service with silencesuppression) over the WiMAX network.

 Non Real-time Polling Service (nrtPS) is designed to support delay

tolerant data streams consisting of variable-sized data packets for which aminimum data rate is required The nrtPS flows like an rtPS flow are polledthrough a unicast request polling but at time-sale of one second or less ThenrtPS flows can also receive a few request polling opportunities duringnetwork congestion, and are also allowed to use the contention request  Best Effort Service (BE) is designed to support data streams for which no

minimum service level is required and therefore may be handled on abandwidth available basic The BE flows are allowed to use contentionrequest opportunity The applications in this class receive any bandwidthremaining after it has been allocated to all other classes.

1.2.1 Admission Control

In IEEE 802.16 before a subscriber station (SS) can initiate a new connection, itmust first make a request to the base station (BS) with the service contracts required.The BS may reject the request based on its ability to uphold the requirements Theadmission control mechanism at the BS is not specified in the standard It is possiblethat the BS admits a connection based on statistical QoS, where, on average, allconnections would not have their QoS satisfied, or that the BS could have a strictermodel where QoS is absolutely guaranteed, even in the worst case scenario Theadmission control problem will be discussed in more detail in chapter 2.

1.2.2 Scheduling

IEEE 802.16 specifies only the outbound traffic scheduling goals, and not themethodology Essentially, traffic is to be serviced such that service contracts have noor minimal disruption.

Bandwidth can be requested in the initialization of a connection, such as anUGS traffic class connection, or, due to the uplink burst profile being so dynamic inother classes of traffic, the SSs would indicate current uplink bandwidth requirementsfor each.

When a SS is granted bandwidth it may choose how this is allocated among itsconnections This means that connections may “borrow” bandwidth from otherconnections within the same SS The only exception is UGS traffic, which is granted afixed amount of bandwidth per frame and may not use more than this Of course itmay use less, in which case the unused bandwidth passes to other connections to beused

1.3 Research Motivation and objectives

This research is motivated by the fact that user expectations of wired andwireless networks have increased with regard to a large variety of services andapplications with different QoS requirements The implementation of QoS guaranteesin such networks is a prerequisite for best supporting services and applications oversuch networks It is believed that this topic requires extensive research, bothqualitative and quantitative, to find out the most suitable model of QoS architecture.

Another motivation of this research is the fact that an ideal QoS architectureneeds to seamlessly support the same QoS constraints across heterogeneous types ornetworks including those designed for the wired and wireless environment The

emphasis of the research would be on broadband wireless access networks in whichstringent requirements are imposed on the QoS architecture due to the numerouslimitations of the wireless channel.

A further motivating factor is that IEEE 802.16 does not specify themethodology to use for scheduling and admission control Previous research in thisarea has only recently gained momentum, with many researchers proposing algorithmsfor efficient bandwidth allocation; however, limitations are still apparent in their work.

The objective of this thesis is to give an evaluation of existing schedulingalgorithms for real-time service flows for WiMAX uplink A new priority admissioncontrol strategy is also introduced, which admits the connections into the systemaccording to their QoS requirements The new strategy can be used to minimize theconnection blocking probability of higher priority connections which is acceptablefrom the viewpoints of both users and service providers.

1.4 Thesis organization

This thesis is divided into 4 chapters: Chapter 1 introduces Broadband WirelessAccess (BWA) Chapter 2 describes the MAC layer functionalities of the IEEE802.16 It also presents an overview of literature in the area of traffic scheduling andadmission control and QoS architecture model Chapter 3 is the description of theproposed scheduling algorithms and admission control framework for broadbandwireless access (BWA) Finally, Chapter 4 explains the simulation models and thesimulation methodology used in this study

Chapter 2 IEEE 802.16 Standards2.1 Protocol layer in 802.16

2.1.1 Physical layer

The WiMAX physical layer is based on Orthogonal Frequency DivisionMultiplexing (OFDM) OFDM is the transmission scheme of choice to enable high-speed data, video, and multimedia communications and is used by a variety ofcommercial broadband systems, including DSL, Wi-Fi, Digital Video Broadcast-Handheld (DVB-H), and MediaFLO, besides WiMAX OFDM is an elegant andefficient scheme for high data rate transmission in a non-line-of-sight or multipathradio environment In this section, we cover the basics of OFDM and provide anoverview of the WiMAX physical layer

The IEEE 802.16-2004 standard specifies 5 variants of the PHY layerdistinguished by whether the PHY layer is Single Carrier (SC) or uses OFDMtechnology The variants with a brief description follow:

Wireless Metropolitan Area Network – Orthogonal Frequency DivisionMultiplexing (WirelessMAN-OFDM): The WirelessMAN-OFDM PHY is based on

OFDM technology designed mainly for fixed SSs, where the SSs are deployed inresidential areas and businesses The OFDM PHY supports sub-channelization in theuplink with 16 sub-channels It also supports TDD and FDD frame structures withboth FDD and H-FDD options The modulation schemes supported are BPSK, QPSK,16-QAM and 64-QAM

Wireless Metropolitan Area Network – Orthogonal Frequency DivisionMultiple Access (WirelessMAN-OFDMA): This variant is based on OFDMA

technology and offers sub-channelization in both uplink and downlink The OFDMAPHY supports both TDD and FDD frame structures, with both FDD and H-FDDoptions The variant is different from WirelessMAN-OFDM in that it supports sub-channelization in both the uplink and downlink directions resulting in broadcastmessages being transmitted at the same time as data

Wireless High Speed Unlicensed Metro Area Network (WirelessHUMAN):

This specification of the PHY layer is similar to the OFDM based layer except it isfocused on Unlicensed National Information Infrastructure (UNII) devices and otherunlicensed bands

Wireless Metropolitan Area Network – Single Carrier (WirelessMAN-SC):

This variant specifies the use of the technology in the frequency range 10-66GHz ThePHY Layer design supports Point to Multi-Point (PMP) architecture whereby the BSacts as the coordinator for all the SSs in its cell In this design, the BS transmits a TimeDivision Multiplexing (TDM) signal in which the SSs are allocated time slots serially.This variant provides support for both TDD and FDD frame structures Both TDD andFDD support adaptive burst profiles whereby the modulation and coding options canbe dynamically assigned on a burst by burst basis

Wireless Metropolitan Area Network - Single Carrier Access(WirelessMAN-SCa): This variant of the PHY layer uses single carrier modulation in

the 2-11GHz frequency range and it is intended for Non Line-Of-Sight (NLOS)operations It supports both FDD and TDD frame structures with TDMA in the uplinkand TDM or TDMA in the downlink The PHY specification includes Forward ErrorCorrection (FEC) coding for both uplink and downlink and framing structures thatallow improved channel estimation performance over NLOS operations.

In the IEEE 802.16e-2005, PHY layer has been improved to adapt to the

mobility, that's due to the exploitation of SOFDMA S-OFDMA improves the

performance of OFDMA256 for NLOS applications by:

 Improving NLOS coverage by using advanced antenna diversity schemes, andhybrid ARQ (HARQ),

 Improving coverage by achieving higher capacity through the use of AdaptiveAntenna Systems (AAS) and MIMO technology,

 Increasing system gain by using denser subchannelization, thereby improvingindoor penetration,

 Allowing tradeoff between coverage and capacity by using DLsubchannelization,

 Introducing high-performance coding techniques such as Turbo Coding andLow-density Parity Check (LDPC), enhancing security and NLOS performance.

OFDM allows smart antenna operations to be perform on complex flat sub-carriers

and therefore complex equalizers are not required to compensate for selective fading In fact, MIMO-OFDM/OFDMA is envisioned to be the corner-stone

frequency-for 4G broadband communication systems The full range of smart antennatechnologies supported includes:

 Beamforming; multiple antennas transmit weighted signals to improve coveragedirectivity.

 Space-time code (STC): orthogonal coding for multiple transmit antenna forspatial diversity gain and reducing fade margin, e.g Alamouti code for 2-antenna case.

 Spatial Multiplexing (SM): multiple data streams are transmitted over multipleantennas If the receiver also has multiple antennas, i.e the MIMO case, it can

separate the different streams to add linearly to give higher throughput.

Unfortunately, MIMO, being dependent on the channel as well, is affected bypoor fading On the other hand, STC, being a transmit diversity, provides largecoverage regardless of channel condition, but does not improve the peak datarate In UL, MS only has one antenna, and two MSs can collaborativelytransmit in the same slot as if two data streams are spatial multiplexed onto thesame one antenna an UL collaborative SM.

2.1.2 MAC Layer

The Mac layer in WiMAX basically provides intelligence to the PHY layer Itcontains 3 sublayers: service-specific convergence sublayer, MAC common partsublayer and the privacy sublayer

Service specific convergence sublayer: the IEEE 802.16-2004 standard

specifies 2 types of service convergence sublayer for mapping service to and from theMAC layer; the ATM sublayer for mapping ATM service and the packet sublayer formapping packet services such as IPv4, IPv6 and Ethernet The major task of thesublayer is to map Service Data Units (SDUs) to MAC connections, and to enable QoSand bandwidth allocation based on the parameters received from the upper layers Theconvergence sublayer also has the ability to perform more complicated tasks such aspayload header compression.

Mac common part sublayer: the MAC protocol, according to the IEEE

802.16-2004 standard, is mainly designed for Point to Multi-Point (PMP) operation.The MAC layer is connection-oriented including connection-less services mapped to aconnection which allows a way of requesting bandwidth and mapping the QoS

parameters of the connection Connections contain a 16 bits Connection Identifiers(CIDs) which act as the primary addresses used for all operations Each SS has a 48-bitMAC address which is mainly used as an equipment identifier Upon entering thenetwork, an SS is assigned 3 management connections in each of the uplink anddownlink directions These connections are:

Basic connection: this connection is responsible for transferring critical MAC

and Radio Link Control (RLC) messages.

Primary Management Connection: this connection is responsible for

transferring longer and more delay-tolerant control messages such as those used forauthentication and connection setup

Secondary Management Connection: this connection is used for transfer of

standard based management messages such as Dynamic Host Configuration Protocol(DHCP), Trivial File Transfer Protocol (TFTP) and Simple Network ManagementProtocol (SNMP).

The functionalities supported by the Common Part Sublayer are:

Channel Acquisition: the MAC protocol includes an initialization procedure

that allows as SS, upon installation, to scan its frequency list to find an operatingchannel Once the SS is synchronized, it will periodically search for DownlinkChannel Descriptor (DCD) and Uplink Channel Descriptor (UCD) messages thatinform the SS about the modulation scheme used on the channel.

Ranging and Negotiation: after determining the parameters to be used forranging, the SS will search for ranging opportunities by scanning the UL-MAPmessages in every frame The ranging response from the SS allows the BS to find outthe distance between itself and the SS.

SS authentication and Registration: each SS contains a manufacturer issued

digital certificate that establishes a link between the 48-bit MAC address of the SS andits public RSA key After verifying the identity of the SS, if the SS is admitted into thenetwork, the BS will respond to the request with an Authorization Reply containing anAuthorization Key (AK), encryptedwith the SS’s public key.

Bandwidth Grant and Request: the MAC layer distinguishes between 2 classes

of SS, one accepts bandwidth for a connection and the other accepts bandwidth for theSS For the Grant per Connection (GPC) class, bandwidth is granted explicitly to theconnection of the SS With the Grant per SS (GPSS) class, bandwidth is granted to the

SS and then the SS decides how to distribute the bandwidth among its connections.The SS can use the bandwidth for the connection that requested it or for anotherconnection The IEEE 802.16-2004 standard allows SSs to request bandwidth viacontention or piggyback mechanism In contention mechanism, the SSs will attachtheir bandwidth request onto data packets A bandwidth request can also be eitherincremental or aggregate When the BS receives an incremental bandwidth request, itwill add the quantity of the bandwidth requested to its current perception of thebandwidth needs of the SS When the BS receives an aggregate bandwidth request, itwill replace its perception of the SS with the quantity of bandwidth requested.Piggyback bandwidth requests are always incremental.

Frame structure and MAP messages: IEEE 802.16 MAC support both TDD

and FDD frame structures The MAC starts building the downlink subframe with MAP (Downlink Map ) and UL-MAP (Uplink MAP) messages The DL-MAPindicates the PHY transitions on the downlink while the UL-MAP indicates bandwidthallocation and burst profiles in the uplink In a Time Division Duplexing (TDD) framestructure, the frame is divided into uplink and downlink subframes along the time axis(see Figure 2.1) The frame starts with the downlink sub-frame followed by a short gapcalled transmit/receive transition gap (TTG) The downlink sub-frame contains apreamble followed by a header and one or more downlink bursts Each uplink burstcontains a preamble that allows the BS to synchronize with each SS The uplink sub-frame is then followed by a short gap called the receive/transmit transition gap (RTG)before the BS can start transmitting again.

DL-Figure 2.1 TDD frame structure

Packing and Fragmentation of MAC SDUs: this functionality is executed in

tandem with the bandwidth allocation process to maximize efficiency, flexibility andeffectiveness Fragmentation is the process by which a MAC SDU is divided into oneor more SDU fragments Packing is the process in which multiple MAC SDUs are

packed into a single MAC PDU payload Either functionality can be initiated by theBS in the downlink or by the SS in the uplink

Scheduling Services: the common part sublayer maps each connection to a

scheduling service, where a scheduling service is associated with pre-determined QoSparameters The IEEE 802.16-2004 standard specifies five scheduling services:Unsolicited Grant Service (UGS), real-time Polling Service (rtPS), extended real-timePolling Service (ertPS), non-real time Polling Service (nrtPS) and Best Effort (BE).The bandwidth allocation mechanism for the UGS service as specified in the IEEE802.16-2004 standard requires the BS to send fixed size grants to the SSs periodically.

Privacy sublayer: provides authentication, secure key exchange and

encryption IEEE 802.16’s privacy protocol is based on the Privacy Key Management(PKM) protocol of the Data Over Cable Service Interface Specification (DOCSIS)Baseline Privacy Interface (BPI+) but has been enhanced to fit seamlessly into theIEEE 802.16 MAC protocol and to better accommodate stronger cryptographicmethods, such as the recently approved Advanced Encryption Standard[6].

2.2 Admission Control

Our approach towards supporting QoS requirements of different classes inIEEE 802.16 BWA is in two directions One is in the form of an admission controlmechanism and the other in the form of packet scheduling Scheduling medicates thelow level contention for service between packet of different classes, while admissioncontrol determines the acceptance or blocking of a new connection These two levelsof control are related for example, if too much traffic is allowed to enter the networkby an admission control policy, then no scheduler would be able to provide therequested QoS of all classes A functioning admission control is thus a prerequisite forany guarantee of packet level QoS Both are important in ensuring that a requestedQoS can be provided for a certain number of connections, a number that is dictated bythe amount of availablebandwidth in the system.

This chapter describes admission control policy which describes the techniquesused to determine the availability of uplink resources for providing the QoSrequirements specified by the requesting connection

2.2.1 Objective

There is no question about the necessity of using admission control in any kindof networks that aims to provide QoS support for its connections As the IEEE 802.16

is aiming to do so, it is essential to have some sorts of admission control beingembedded in its QoS architecture, however, there is no admission control procedure isdefined in the current standard In this work we have suggested an efficient admissioncontrol module, targeting network utilization while giving a priority to the BE and rtPSconnection request, since they carry a data of real-time applications.

2.2.2 Overview of Admission Control

I express the need for the admission control in order to control the usage andallocation of bandwidth resources for various traffic classes requiring certain QoSguarantee Admission control is a key component in determining whether a newrequest for a connection can be granted or not according to the current traffic load ofthe system This assumes great significance when the BS needs to maintain a certainpromised level of service for all the connections being admitted (served) If theadmission control admits too few connections, it results in wastage of systemresources On the other hand If too many connections are allowed to contend forresources, then the performance of the already admitted connections degrades rapidlyin the presence of new connections Therefore, judicious decision making mechanismfor allocating bandwidth to different classes of service is needed.

In IEEE 802.16, before an SS can initiate a new connection or changing ordeleting an already admitted connection, it must first make a request to the BS Asmentioned earlier in chapter 1, four types of MAC layer services exist in IEEE 802.16.These service flows can be created, changed, or deleted through the issue of dynamicservice addition (DSA), dynamic service creation (DSC) and dynamic service deletion(DSD) messages Each of these actions can be initiated by the SS or the BS and arecarried out through a two or three-way-handshake.

2.2.3 Admission Control Policy

The task of admission controller is to accept or reject the arriving requests for aconnection in order to maximize the channel utilization, by accepting as manyconnections as possible, while keeping the QoS level of all connections at the levelspecified in their traffic profile In other words it ensures that already admittedconnections QoS will not be affected by the decision made Although it may seems tobe very intuitive and simple procedure, it has great influence on QoS of the admittedconnections.

This issue have been studied and researched extensively in the context of wiredand wireless networking Although the focus of this research was not admissioncontrol, whenever it comes to IEEE 802.16, putting a well rounded introduction seemsto be indispensable, thus we have concisely introduced some of the fundamental worksthat need to be done in this area as there is no defined procedure in IEEE 802.16.

Admission control algorithms can be categorized into three classes of completesharing, complete partitioning and hybrid policies which is a combination methods ofthe other two.

 Complete sharing (CS): allows all users equal access to the bandwidth

available at all times This strategy results in maximum utilization of theavailable bandwidth, specially in high traffic networks, which is what networkproviders aiming at However, at the same time, it does not differentiatebetween connections of different priority that is a perverse outcome whenconnection of one class needs significantly less bandwidth than others At thissituations it might be desirable to reject calls of this type to increase theprobability of future acceptance of a larger call In other word it is not fairstrategy to the wider bandwidth users as all request would be dealt with thesame priority.

 Complete partitioning (CP): divides up the available bandwidth into

non-overlapping pools of bandwidth in accordance with the type of user’sconnection Therefore, number of existing users in each class would beprohibited to a maximum number M which admission decision will be madeupon This policy allows for more control of the relative blocking probability atthe expense of overall usage of the network.

 Hybrid policies: basically provide a compromise between the different policies

by subdividing the available bandwidth into sections Part of the bandwidth incompletely shared and the other part is completely partitioned Depending onthe policy adopted, the partitioned division would be dedicated to some or allclasses This allows more live up to the QoS requirements of the different usertype while maintaining higher network utilization.

In the above mentioned admission control policies, CP and CS, have nocomplexities to be elaborated and decision making would be a matter of checking asingle condition, though, the hybrid strategy is a more open ended problem As an

example, in the following we suggest an algorithm that could be considered as ahybrid method for admission control

2.3 Services and service flows2.3.1 Connections and service flow

The CS provides any transformation or mapping of external network data receivedthrough the CS Service Access Point (SAP) into MAC SDUs received by the MACCommon Part Sublayer (CPS) through the MAC SAP (see Figure 1) This includesclassifying external network Service Data Units (SDUs) and associating them with theproper MAC Service Flow Identifier (SFID) and Connection Identifier (CID).Classification and mapping are then based on two 802.16 MAC layer fundamentalconcepts

 Connection: A connection is a MAC Level connection between a BS and an

SS (or MS) or inversely It is a unidirectional mapping between a BS and an SSMAC peers for the purpose of transporting a service flow's traffic A connectionis only for one type of service (e.g voice and email cannot be on the sameMAC connection) A connection is identified by a CID (Connection IDentifier),an information coded on 16 bits.

 Service flow: A Service Flow (SF) is a MAC transport service that provides

unidirectional transport of packets on the uplink or on the downlink A serviceflow is identified by a 32-bit SFID (Service Flow IDentifier) The service flowdefines the QoS parameters for the packets (PDUs) that are exchanged on theconnection.

Figure 2.2 shows the relation between the SFID and CID The relation between thetwo is the following: only admitted and active service flows (see the definitions below)are mapped to a CID, i.e a 16-bit CID In other terms:

Figure 2.2: Correspondence between the CID and SFID

 A SFID matches to zero (provisioned service flows) or to one CID (admitted oractive service flow)

 A CID maps to a service flow identifier (SFID), which defines the QoSparameters of the service flow associated with that connection.

The definitions of connection and service flow in the 802.16 standard allowdifferent classes of QoS to be found easily for a given element (SS or BS), withdifferent levels of activation (see Figure 2.3) More details will now be given aboutconnections (and CIDs) and service flows.

Figure 2.3: Illustration of service flows and connections

2.3.2 Connection Identifiers (CIDs)

A Connection IDentifier (CID) identifies a connection where every MAC SDUof a given communication service is mapped into The CID is a 16-bit value thatidentifies a unidirectional connection between equivalent peers in the MAC layers of aBS and an SS All 802.16 traffic is carried on a connection Then, the CID can beconsidered as a connection identifier even for nominally connectionless traffic like IP,since it serves as a pointer to destinations and context information The use of a 16-bitCID permits a total of 64K connections within each downlink and uplink channel.There are several CIDs defined in the standard (see Table 1) Some CIDs have aspecific meaning Some of the procedures introduced in this table, such as ranging,basic, primary and secondary management.

Table 1: CID ranges as defined in Reference IEEE 802.16-2004.

Initial ranging 0 × 0000 Used by SS and BS during the initial ranging processBasic CID 0 × 0001 –

m

Each SS has a basic CID and has a short delay The same CID value is assigned to both the downlink and uplink connections

Primary management m + 1 − 2m The primary management connection is used to exchange

longer, more delaytolerant MAC management messagesTransport CIDs and

AAS initial ranging CID

0 × FEFF A BS supporting AAS (Advanced Antenna System) uses this CID when allocating an AAS ranging period (using AAS_ Ranging_Allocation_IE)

Multicast polling CIDs 0 × FFOO−0 × FFF9

An SS may be included in one or moremulticast polling groups for the purposes of obtaining bandwidth via polling.These connections have no associated service flow

Normal mode multicast CID

0 × FFFA Used in DL-MAP to denote bursts for transmission of downlink broadcast information to normal mode SSSleep mode multicast

0 × FFFB Used in DL-MAP to denote bursts for transmission of downlink broadcast information to sleep mode SS May also be used in MOB_TRF-INO messages

Idle mode multicast CID

0 × FFFC Used in DL-MAP to denote bursts for transmission of downlink broadcast information to idle mode SS May also be used in MOB_PAG-ADV messages

Fragmentable broadcast CID

0 × FFFD Used by the BS for transmission of management broadcast information with fragmentation The fragment subheader should use an II-bit long FSN on this connection

Padding CID 0 × FFFE Used for transmission of padding information by the SS and BS

Broadcast CID 0 × FFFF Used for broadcast information that is transmitted on a

downlink to all SSs

2.3.3 Service Flows

A Service Flow (SF) is a MAC transport service that provides unidirectionaltransport of packets on the uplink or on the downlink It is identified by a 32-bit SFID(Service Flow IDentifier) A service flow is characterized by a set of QoS parameters.The QoS parameters include details of how the SS may request uplink bandwidthallocations and the expected behaviour of the BS uplink scheduler.

2.3.3.1 Service Flow Attributes

A service flow is partially characterised by the following attributes:

 Service Flow ID An SFID is assigned to each existing service flow The SFIDserves as the identifier for the service flow in the network.

 CID Mapping a CID to an SFID exists only when the connection has anadmitted or active service flow (see below).

 ProvisionedQoSParamSet This defines a QoS parameter set that is provisionedvia means that the standard assumes to be outside of its scope The standardstates that this could be part of the network management system For example,the service (or QoS) class name is an attribute of the ProvisionedQoSParamSet.There are five QoS classes, the fifth having been added by the 802.16eamendment.

 AdmittedQoSParamSet This defines a set of QoS parameters for which the BS,and possibly the SS, are reserved resources The principal resource to bereserved is bandwidth, but this also includes any other memory or time-basedresource required to subsequently activate the flow.

 ActiveQoSParamSet This defines a set of QoS parameters defining the serviceactually being provided to the service flow Only an active service flow mayforward packets The activation state of the service flow is determined by theActiveQoSParamSet If the ActiveQoSParamSet is null, then the service flow isinactive.

 Authorisation module This is a logical function within the BS that approves ordenies every change to QoS parameters and classifiers associated with a service

flow As such, it defines an ‘envelope’ that limits the possible values of theAdmittedQoSParamSet and ActiveQoSParamSet.

2.3.3.2 Types of Service Flow

The standard has defined three types of service flow:

 Provisioned service flows This type of service flow is known via provisioningby, for example, the network management system Its AdmittedQoSParamSetand ActiveQoSParamSet are both null.

 Admitted service flow The standard supports a two-phase activation model thatis often used in telephony applications In the two-phase activation model, theresources for a call are first ‘admitted’ and then, once the end-to-endnegotiation is completed, the resources are ‘activated’.

 Active service flow This type of service flow has resources committed by theBS for its ActiveQoSParamSet Its ActiveQoSParamSet is non-null.

2.3.3.3 Classification and Mapping

Classification is the process by which a MAC SDU is mapped on to a particularconnection for transmission between MAC peers The mapping process associates aMAC SDU with a connection, which also creates an association with the service flowcharacteristics of that connection Classification and mapping mechanisms exist in theuplink and downlink In the case of a downlink transmission, the classifier will bepresent in the BS and in the case of an uplink transmission it is present in the SS Aclassifier is a set of matching criteria applied to each packet entering theWiMAX/802.16 network The set of matching criteria consists of some protocol-specific packet matching criteria (a destination IP address, for example), a classifierpriority and a reference to a CID If a packet matches the specified packet matchingcriteria, it is then delivered to the SAP for delivery on the connection defined by theCID The service flow characteristics of the connection provide the QoS for thatpacket The classification mechanism is shown in figure 2.4.

Payload Header Suppression (PHS)

The packets delivered to OSI model layer 2 may have very large headers, sometimesas long as 120 bytes This is the case for some RTP/UDP/IPv6 packets (RTP, Real-Time Protocol, UDP, User Datagram Protocol) This is very often repetitive(redundant) information and so should not be transmitted on a scarce resource such as

a radio channel, which should be used for useful information This process is known asheader compression and decompression in 3G-cellular systems In the 802.16 standard,the PHS process suppresses repetitive (redundant) parts of the payload header in theMAC SDU of the higher layer

Figure 2.4: Classification and CID mapping

The receiving entity restores the suppressed parts Implementation of the PHScapability is optional Figure 2.5 shows the PHS mechanism at the sending entity.Suppression of parts of the header leads to a compressed header The receiver has torestore the header before properly using the received packet (Figure 2.6)

Figure 2.5: Header suppression at the sending entity

Figure 2.6: Header suppression mechanism at the receiving entity

To indicate whether the PHS is present or not, the PHS support field is used Thisparameter indicates the level of PHS support The PHS support field is a field in someMAC management messages, Registration Request and Registration Response Table2 shows the possible values of the PHS support field The default value is 0 (no PHS).

Table 2: Possible values of the PHS support field