660 PHOTONIC PACKET SWITCHING 12.4 12.5 and find an expression for the control inputs Cl Ck. Assume that if ci = 1, switch i is set in the bar state, and if ci = 0, switch i is set in the cross state. Consider the fiber loop mirror shown in Figure 12.8, and show that the nonlinear element should introduce a phase shift of zr between the clockwise and counterclock- wise signals in order for all the energy entering the directional coupler from arm A to be transferred to arm B. We have seen that many photonic packet-switching proposals use a lower-rate header compared to the payload. Suppose the maximum header bit rate is 1 Gb/s and headers are 10 bytes long. The payload data rate is 100 Gb/s. (a) We would like the duration of the payload to be 90% of the overall packet duration (including header and payload). What size does the payload need to be? (b) If we wanted the maximum payload size to be 1000 bytes and maintain the same efficiency, at what rate would the header have to be transmitted? (c) Suppose we need a minimum of 1 #s to process the header. This time is accounted for as an additional guard band in the overall packet, in addition to the header and payload. Again, if we want to maintain the payload at 90% of the overall packet, and the header at 10 bytes at 1 Gb/s, what size does the payload need to be? References [Ams83] S. Amstutz. Burst switching an introduction. IEEE Communications Magazine, 21:36-42, Nov. 1983. [AS92] A.S. Acampora and S. I. A. Shah. Multihop lightwave networks: A comparison of store-and-forward and hot-potato routing. IEEE Transactions on Communications, 40(6):1082-1090, June 1992. [Bar64] P. Baran. On distributed communications networks. IEEE Transactions on Communications, pages 1-9, March 1964. [Bar96] R.A. Barry et al. All-optical network consortium~ultrafast TDM networks. IEEE JSAC/JLT Special Issue on Optical Networks, 14(5):999-1013, June 1996. [BCM+92] D.J. Blumenthal, K. Y. Chen, J. Ma, R. J. Feuerstein, and J. R. Sauer. Demonstration of a deflection routing 2 • 2 photonic switch for computer interconnects. IEEE Photonics Technology Letters, 4(2):169-173, Feb. 1992. [BDG95] C. Baransel, W. Dobosiewicz, and P. Gburzynski. Routing in multihop packet switching networks: Gb/s challange. IEEE Network, pages 38-61, May/June 1995. References 661 [BDN90] K.J. Blow, N. J. Doran, and B. P. Nelson. Demonstration of the nonlinear fibre loop mirror as an ultrafast all-optical demultiplexer. Electronics Letters, 26(14):962-964, July 1990. [BFP93] A. Bononi, E Forghieri, and P. R. Prucnal. Synchronisation in ultrafast packet switching transparent optical networks. Electronics Letters, 29( 10):872-873, May 1993. [BIPT98] D.J. Blumenthal, T. Ikegami, P. R. Prucnal, and L. Thylen, editors. IEEEIOSA Journal of Lightwave Technology: Special Issue on Photonic Packet Switching Technologies, Techniques and Systems, volume 16, Dec. 1998. [Bor95] E Borgonovo. Deflection routing. In M. Steenstrup, editor, Routing in Communication Networks. Prentice Hall, Englewood Cliffs, NJ, 1995. [BP96] A. Bononi and P. R. Prucnal. Analytical evaluation of improved access techniques in deflection routing networks. IEEEIA CM Transactions on Networking, 4(5):726-730, Oct. 1996. [BPS94] D.J. Blumenthal, P. R. Prucnal, and J. R. Sauer. Photonic packet switches: Architectures and experimental implementations. Proceedings of IEEE, 82:1650-1667, Nov. 1994. [Bur94] M. Burzio et al. Optical cell synchronization in an ATM optical switch. In Proceedings of European Conference on Optical Communication, pages 581-584, 1994. [CHI+92] M.W. Chbat, B. Hong, M. N. Islam, C. E. Soccolich, and P. R. Prucnal. Ultrafast soliton-trapping AND gate. IEEEIOSA Journal on Lightwave Technology, 10(12):2011-2016, Dec. 1992. [CHK+96] R.L. Cruz, G. R. Hill, A. L. Kellner, R. Ramaswami, and G. H. Sasaki, editors. IEEE JSA CIJLT Special Issue on Optical Networks, volume 14, June 1996. [Ch196] I. Chlamtac et al. CORD: Contention resolution by delay lines. IEEE JSACIJLT Special Issue on Optical Networks, 14(5):1014-1029, June 1996. [CLM97] D. Cotter, J. K. Lucek, and D. D. Marcenac. Ultra-high bit-rate networking: From the transcontinental backbone to the desktop. IEEE Communications Magazine, 35(4):90-95, April 1997. [Cot95] D. Cotter et al. Self-routing of 100 Gbit/s packets using 6 bit "keyword" address recognition. Electronics Letters, 31 (25):2201-2202, Dec. 1995. [Dan97] S.L. Danielsen et al. WDM packet switch architectures and analysis of the influence of tuneable wavelength converters on the performance. IEEE/OSA Journal on Lightwave Technology, 15(2):219-227, Feb. 1997. 662 PHOTONIC PACKET SWITCHING [Dan98] S.L. Danielsen et al. Analysis of a WDM packet switch with improved performance under bursty traffic conditions due to tuneable wavelength converters. IEEE/OSA Journal on Lightwave Technology, 16(5):729-735, May 1998. [DW88] N.J. Doran and D. Wood. Nonlinear-optical loop mirror. Optics Letters, 13(1):56-58, Jan. 1988. [Eis92] M. Eiselt. Optical loop mirror with semiconductor laser amplifier. Electronics Letters, 28(16):1505-1506, July 1992. [ENW96] A. Erramilli, O. Narayan, and W. Willinger. Experimental queueing analysis with long-range dependent packet traffic. IEEE/A CM Transactions on Networking, 4(2):209-223, Apr. 1996. [FBP95] E Forghieri, A. Bononi, and P. R. Prucnal. Analysis and comparison of hot-potato and single-buffer deflection routing in very high bit rate optical mesh networks. IEEE Transactions on Communications, 43(1):88-98, Jan. 1995. [FGO+96] M. Fujiwara, M. S. Goodman, M. J. O'Mahony, O. K. Tonguez, and A. E. Willner, editors. IEEE/OSA JLT/JSA C Special Issue on Multiwavelength Optical Technology and Networks, volume 14, June 1996. [FHHH90] M.E. Fermann, E Haberl, M. Hofer, and H. Hochreiter. Nonlinear amplifying loop mirror. Optics Letters, 15(13):752-754, July 1990. [FV88] G.J. Foschini and G. Vannucci. Using spread spectrum in a high-capacity fiber-optic local network. IEEE/OSA Journal on Lightwave Technology, 6(3):370-379, March 1988. [Gam98] E Gambini et al. Transparent optical packet switching: Network architecture and demonstrators inthe KEOPS project. IEEE JSAC: Special Issue on High-Capacity Optical Transport Networks, 16(7):1245-1259, Sept. 1998. [GG93] A.G. Greenberg and J. Goodman. Sharp approximate models of deflection routing in mesh networks. IEEE Transactions on Communications, 41(1):210-223, Jan. 1993. [GH92] A.G. Greenberg and B. Hajek. Deflection routing in hypercube networks. IEEE Transactions on Communications, 40(6):1070-1081, June 1992. [Gre93] P.E. Green. Fiber-Optic Networks. Prentice Hall, Englewood Cliffs, NJ, 1993. [GSP94] I. Glesk, J. P. Sokoloff, and E R. Prucnal. All-optical address recognition and self-routing in a 250 Gb/s packet-switched network. Electronics Letters, 30(16):1322-1323, Aug. 1994. [Gui98] C. Guillemot et al. Transparent optical packet switching: The European ACTS KEOPS project approach. IEEE/OSA Journal on Lightwave Technology, 16(12):2117-2134, Dec. 1998. References 663 [Gui00] K. Guild et al. Cascading and routing 14 optical packet switches. In Proceedings of European Conference on Optical Communication, 2000. [Gun97a] P. Gunning et al. 40 Gbit/s optical TDMA LAN over 300m of installed blown fibre. In Proceedings of European Conference on Optical Communication, volume 4, pages 61-64, Sept. 1997. [Gun97b] P. Gunning et al. Optical-TDMA LAN incorporating packaged integrated Mach-Zehnder interferometer channel selector. Electronics Letters, 33(16):1404-1406, July 1997. [HA00] D.K. Hunter and I. Andonovic. Approaches to optical Internet packet switching. IEEE Communications Magazine, 38(9):116-122, Sept. 2000. [Hal97] K.L. Hall. All-optical buffers for high-speed slotted TDM networks. In IEEE/LEOS Summer Topical Meeting on Advanced Semiconductor Lasers and Applications, page 15, 1997. [HC93] B. Hajek and R. L. Cruz. On the average delay for routing subject to independent deflections. IEEE Transactions on Information Theory, 39(1):84-91, Jan. 1993. [HCA98] D.K. Hunter, M. C. Chia, and I. Andonovic. Buffering in optical packet switches. IEEE/OSA Journal on Lightwave Technology, 16(12):2081-2094, Dec. 1998. [Hi185] W.D. Hillis. The Connection Machine. MIT Press, Cambridge, MA, 1985. [Hi187] W.D. Hillis. The connection machine. Scientific American, 256(6), June 1987. [HK88] M.G. Hluchyj and M. J. Karol. Queuing in high-performance packet switching. IEEE JSAC, 6(9):1587-1597, Dec. 1988. [HMY98] K. Habara, T. Matsunaga, and K I. Yukimatsu. Large-scale WDM star-based photonic ATM switches. IEEEIOSA Journal on Lightwave Technology, 16(12):2191-2201, Dec. 1998. [HR98] K.L. Hall and K. T. Rauschenbach. All-optical buffering of 40 Gb/s data packets. IEEE Photonics Technology Letters, 10(3):442-444, Mar. 1998. [Hun99] D.K. Hunter et al. WASPNET a wavelength switched packet network. IEEE Communications Magazine, 37(3):120-129, Mar. 1999. [KGSP94] M.G. Kane, I. Glesk, J. P. Sokoloff, and P. R. Prucnal. Asymmetric loop mirror: Analysis of an all-optical switch. Applied Optics, 33(29):6833-6842, Oct. 1994. [KH90] A. Krishna and B. Hajek. Performance of shuffle-like switching networks with deflection. In Proceedings of IEEE Infocorn, pages 473-480, 1990. [LGM+97] J.K. Lucek, P. Gunning, D. G. Moodie, K. Smith, and D. Pitcher. Synchrolan: A 40 Gbit/s optical-TDMA LAN. Electronics Letters, 33(10):887-888, April 1997. 664 PHOTONIC PACKET SWITCHING [LNGP96] E. Leonardi, E Neri, M. Gerla, and P. Palnati. Congestion control in asynchronous high-speed wormhole routing networks. IEEE Communications Magazine, pages 58-69, Nov. 1996. [Mas96] E Masetti et al. High speed, high capacity ATM optical switches for future telecommunication transport networks. IEEE JSA C/JLT Special Issue on Optical Networks, 14(5):979-998, June 1996. [Max89] N. E Maxemchuck. Comparison of deflection and store-and-forward techniques in the Manhattan Street and shuffle-exchange networks. In Proceedings of IEEE Infocom, pages 800-809, 1989. [MEM98] D.D. Marcenac, A. D. Ellis, and D. G. Moodie. 80 Gbit/s OTDM using electroabsorption modulators. Electronics Letters, 34(1):101-103, Jan. 1998. [Mid93] J.E. Midwinter, editor. Photonics in Switching, Volume II: Systems. Academic Press, San Diego, CA, 1993. [Mik99] B. Mikkelsen et al. Unrepeatered transmission over 150 km of nonzero-dispersion fibre at 100 Gbit/s with semiconductor based pulse source, demultiplexer and clock recovery. Electronics Letters, 35(21):1866-1868, Oct. 1999. [MS88] J.E. Midwinter and E W. Smith, editors. IEEE JSAC" Special Issue on Photonic Switching, volume 6, Aug. 1988. [PF95] V. Paxon and S. Floyd. Wide area traffic: The failure of Poisson modelling. IEEE/ACM Transactions on Networking, 3(3):226-244, June 1995. [PHR97] N.S. Patel, K. L. Hall, and K. A. Rauschenbach. Optical rate conversion for high-speed TDM networks. IEEE Photonics Technology Letters, 9(9):1277, Sept. 1997. [Pru93] E R. Prucnal. Optically processed self-routing, synchronization, and contention resolution for 1-d and 2-d photonic switching architectures. IEEE Journal of Quantum Electronics, 29(2):600-612, Feb. 1993. [PSF86] P.R. Prucnal, M. A. Santoro, and T. R. Fan. Spread spectrum fiber-optic local area network using optical processing. IEEE/OSA Journal on Lightwave Technology, LT-4(5):547-554, May 1986. [QY99] C. Qiao and M. Yoo. Optical burst switching (OBS): A new paradigm for an optical Internet. Journal of High Speed Networks, 8(1):69-84, 1999. [RMGB97] M. Renaud, E Masetti, C. Guillemot, and B. Bostica. Network and system concepts for optical packet switching. IEEE Communications Magazine, 35(4):96-102, Apr. 1997. [Sa189] J.A. Salehi. Code division multiple-access techniques in optical fiber networks Part I: Fundamental principles. IEEE Transactions on Communications, 37(8):824-833, Aug. 1989. References 665 [SB89] J.A. Salehi and C. A. Brackett. Code division multiple-access techniques in optical fiber networksmPart II: Systems performance analysis. IEEE Transactions on Communications, 37(8):834-842, Aug. 1989. [SBP96] S W. Seo, K. Bergman, and P. R. Prucnal. Transparent optical networks with time-division multiplexing. IEEE JSA C/JLT Special Issue on Optical Networks, 14(5):1039-1051, June 1996. [SID93] J.R. Sauer, M. N. Islam, and S. P. Dijaili. A soliton ring network. IEEE/OSA Journal on Lightwave Technology, 11(12):2182-2190, Dec. 1993. [Smi81] B. Smith. Architecture and applications of the HEP multiprocessor system. In Real Time Signal Processing IV, Proceedings of SPIE, pages 241-248, 1981. [SPGK93] J.P. Sokoloff, P. R. Prucnal, I. Glesk, and M. Kane. A terahertz optical asymmetric demultiplexer (TOAD). IEEE Photonics Technology Letters, 5(7):787-790, July 1993. [To198] P. Toliver et al. Routing of 100 Gb/s words in a packet-switched optical networking demonstration (POND) node. IEEE/OSA Journal on Lightwave Technology, 16(12):2169-2180, Dec. 1998. [Tur99] J.S. Turner. Terabit burst switching. Journal of High Speed Networks, 8(1):3-16, 1999. [Yam98] Y. Yamada et al. Optical output buffered ATM switch prototype based on FRONTIERNET architecture. IEEE JSA C" Special Issue on High-Capacity Optical Transport Networks, 16(7):2117-2134, Sept. 1998. [YQD01] M. Yoo, C. Qiao, and S. Dixit. Optical burst switching for service differentiation in the next-generation optical Internet. IEEE Communications Magazine, 39(2):98-104, Feb. 2001. [ZT98] W.D. Zhong and R. S. Tucker. Wavelength routing-based photonic packet buffers and their applications in photonic switching systems. IEEE/OSA Journal on Lightwave Technology, 16(10):1737-1745, Oct. 1998. This Page Intentionally Left Blank Deployment Considerations I N THIS CHAPTER, we will study some of the issues facing network operators as they build new networks or upgrade their networks to higher and higher capacities. We will start by understanding how the network is changing from a services perspective, and then understand the changes happening to the network infrastructure. Chapter 1 provided an overview of some of these changes, but we will examine them in detail in this chapter. We will try to understand the various architectural choices available to carriers planning their next-generation networks, in terms of the roles played by SONET/SDH, IP, and ATM. We will discuss the role played by the optical layer and the economic considerations underlying the deployment of WDM and TDM optical layer technologies in the network. We will see that long-haul networks and metro networks have different requirements that influence the choice of technology deployed. In general, it is difficult to decide between the different technologies, and network operators often employ sophisticated network design tools to help them understand the cost trade-offs between different approaches. The examples and problems in this chapter will help the reader gain a better understanding of these trade-offs. 13.1 The Evolving Telecommunications Network The legacy transport network in place in networks run by established carriers is based on SONET and SDH. Over the past decade, we have seen the WDM optical layer play an increasing role in these networks. 667 668 DEPLOYMENT CONSIDERATIONS Several factors are causing service providers to reexamine the way they build their transport network. The first driver is obviously the enormous growth in network traffic. Not only is the traffic doubling every year, but the traffic mix is unpredictable and changing. Another driver is the increasing dominance of data traffic, particularly Internet traffic, relative to voice traffic. Data traffic now exceeds voice traffic on the public network. This trend is likely to continue for at least the next several years. A third driver is the advent of increased competition, which is causing service providers to rethink how they deploy services. In contrast to a world where a new service request for bandwidth could take weeks to months to be fulfilled and require long-term contractual agreements, service providers are increasingly entering a world where services need to be deployed rapidly without long-term contracts at highly competitive rates. Moreover, there is now a new generation of carriers who operate under signifi- cantly different business models than the established carriers. These different business models require different architectures. A carrier providing services to interconnect Internet service providers has very different requirements than a traditional carrier servicing voice and private circuit-switched lines. We also now have a new set of carriers' carriers. These are carriers providing bulk bandwidths (say, at 622 Mb/s and above) primarily to other carriers. These carriers' carriers have different require- ments from carriers delivering low-speed services (such as 1.5 Mb/s lines) to their customers. Before we delve into the evolution of the network, it is worth looking at what carriers look for when they deploy equipment in their network. At the end of the day, what they deploy must either enable them to reduce the cost of their network, or enable them to generate revenue from new services enabled by the equipment deployed. From a cost perspective, carriers look at capital cost and operations cost. Capital cost is the up-front cost of deploying the equipment, and operations cost represents the recurring cost of maintaining and operating the network. Capital cost includes the cost of the equipment, as well as the cost of real estate, providing for appropriate power and cooling and the fiber facilities. In the case of transmission equipment, the goal is to minimize the cost per bit transmitted per mile in the network. It is important to look at the initial entry cost, as well as the cost to add incremental capacity to already-deployed equipment. Operations cost includes real estate rental/lease costs; recurring costs of power and cooling; labor costs to provision, maintain, and service the equipment; and costs associated with replacing failed equipment and missing service-level agreements on network availability. While most carriers will say that operations costs dominate over capital costs in their networks, capital costs are usually much easier to quantify, and hence many carriers use capital costs as the primary basis for making purchasing decisions. 13.1 The Evolving Telecommunications Network 669 Looking at the revenue side of the equation, carriers are always on the lookout for generating new revenue streams by deploying new services. These might include services tailored toward enabling new applications, for instance, providing storage networks between data centers, or modified versions of traditional services. For instance, deploying equipment that enables a carrier to set up and take down private line circuit-switched services in minutes where needed would enable a carrier to offer short-term tariffs on these services, as opposed to requiring its customers to buy the service for extended durations. Another benefit of this capability is that it reduces the time to deploy a service and extracts more revenue as a result. Yet another benefit is that it allows a carrier to better utilize its existing network resources, without having stranded bandwidth due to an inability to anticipate the traffic pattern in the network. The factors described above are forcing carriers to deploy networks that can scale in capacity, networks that are flexible in that they are able to deliver a wide variety of services where needed when needed. The optical layer provides carriers with the ability to deliver these high-speed circuit-switched services, and also serves as the transport mechanism for carrying multiplexed low-speed packet and circuit-switched services. 13.1.1 The SONET/SDH Core Network Figure 13.1(a) shows the core network of a typical established carrier. The network consists of interconnected SONET rings. Given today's capacity demands, many of the rings actually consist of multiple rings connecting the same set of nodes. These are called stacked rings. These rings operate over different fibers, or more commonly, wavelengths within the same fibers using WDM. Figure 13.1(b) shows a blowup of a large node in this network. The node has multiple WDM terminals (OLTs). Each ring passing through the node requires a SONET ADM. These ADMs are connected to the OLTs and operate at line rates of OC-48 (2.5 Gb/s) or OC-192 (10 Gb/s). The ADMs drop lower-speed traffic streams, ranging from 45 Mb/s DS3 streams to higher-speed 622 Mb/s OC-12 streams. The lower-speed traffic is handled by digital crossconnect systems (DCSs). Data traffic is brought into the network through these lower-speed signals and multiplexed to higher speeds by the SONET ADMs and the DCSs. This data enters the network typically in the form of private lines, such as DS1, DS3 or El, E3 lines, or directly at other SONET/SDH rates. These rates are well defined and mapped into the SONET/SDH multiplexing structure. Other data traffic, such as IP traffic from routers or ATM traffic from ATM switches, can be brought into the network via DS1/DS3 lines or higher-speed optical signals such as OC-3, OC-12, and carried over the SONET/SDH infrastructure. . the header and payload. Again, if we want to maintain the payload at 90% of the overall packet, and the header at 10 bytes at 1 Gb/s, what size does the payload need to be? References [Ams83]. packet switching. IEEE JSAC, 6(9):158 7-1 597, Dec. 1988. [HMY98] K. Habara, T. Matsunaga, and K I. Yukimatsu. Large-scale WDM star-based photonic ATM switches. IEEEIOSA Journal on Lightwave. S. Amstutz. Burst switching an introduction. IEEE Communications Magazine, 21:3 6-4 2, Nov. 1983. [AS92] A. S. Acampora and S. I. A. Shah. Multihop lightwave networks: A comparison of store-and-forward