IEEE Communications Magazine • March 1999 120 WASPNET: A Wavelength Switched Packet Network 0163-6804/99/$10.00 © 1999 IEEE avelength-division multiplexing (WDM) is cur- rently being deployed in telecommunications networks in order to satisfy the increased demand for capacity brought about by both narrowband services and new broadband services such as high-speed Internet. While it is thought that WDM will ultimately evolve to intercon- nected rings or perhaps a mesh network, the objective of the Wavelength Switched Packet Network (WASPNET) project is to gain a more long-term understanding of how optical networks will develop. WASPNET is a WDM trans- port network that uses optical packet switching, resulting in greater flexibility, functionality, and granularity than possi- ble with the current generation of WDM networks. These optical packets may be used to carry asynchronous transfer mode (ATM) or IP, for example, and the network is also designed to support synchronous digital hierarchy/syn- chronous optical network (SDH/SONET) traffic, thus per- mitting a smooth upgrade path. Optical packet switches [1–3] have attracted considerable research interest internationally due to their potential for overcoming projected difficulties with very large electronic switching cores, such as connection, pinout, and electromag- netic interference (EMI) problems. A key problem when designing packet switches of any kind is contention resolu- tion, since multiple packets may arrive asynchronously at the same time to go to the same output. Buffering is often employed to solve this problem, but since optical random access memory (RAM) does not exist, delay lines (usually made of optical fiber) must be used to store optical packets and implement buffering. Various solutions to optical pack- et switching have been proposed, dictated by the buffering strategy [1]. Implement Medium to Large Buffers — The switches implemented by this technique may be cascaded to implement very large buffers, suitable for bursty traffic. Use No Buffers in the Switch Nodes, but Employ Deflection Routing — When multiple packets arrive destined for a given output, all but one are “deflected” to other out- puts, to find their way to the destina- tion by another route through the network. This not only provides fast and flexible routing, but also allows nodes to have no buffering. However, each packet transmitted from a node may be routed across a different path to the same destination. Some packets may wander within the network and waste bandwidth. Consequently, each packet will experience different propagation delays, and the traffic may not arrive at the destination node in sequence. Compromise by Using a Small Amount of Buffering with Deflection Routing — There are various such 2 x 2 buffered switches consisting of a chain of 2 x 2 switch devices and delay lines. Here a new approach, WASPNET, is proposed as a solu- tion to the ever-increasing demand for telecommunications transport capacity. A key feature is its use of statistical multi- plexing over many wavelengths to reduce the amount of pack- et contention, and hence the amount of optical buffering required — which is difficult to implement — for a given quality of service (QoS). It is a packet-based transport net- work, which is designed both to support conventional optical paths for the transport of SDH and also to switch optical packets. Unlike other optical packet network proposals, WASPNET is a reconfigurable multiwavelength transport net- work. Hence, it provides a smooth upgrade path from SDH over WDM while still supporting legacy SDH equipment, and possesses greater flexibility and granularity than existing WDM network proposals. In WASPNET, not only are node design and routing considered, but also network control and operation, device fabrication, and demonstrator construction. Two possible network control methodologies were identified: the scat- tered wavelength path (SCWP) and shared wavelength path (SHWP) schemes. These are compared in the follow- ing section, and the problems of control, packet ordering, resilience, and sequencing are then addressed. We describe candidate packet formats, and detail the switch architec- ture that has been proposed. The article goes on to describe the devices that will be used in the demonstrator, followed by a description of the demonstrator, which is under construction. Finally, the last section contains the conclusions. David K. Hunter, Mohamed H. M. Nizam, Meow C. Chia, and Ivan Andonovic, University of Strathclyde Ken M. Guild, Anna Tzanakaki, and Mike J. O’Mahony, Essex University John D. Bainbridge, Marc F. C. Stephens, Richard V. Penty, and Ian H. White, University of Bristol W WASPNET is an EPSRC-funded collaboration between three British universities: the University of Strathclyde, Essex Universi- ty, and Bristol University, supported by a number of industrial institutions. The project — which is investigating a novel packet-based optical WDM transport network — involves determining the management, systems, and devices ramifications of a new network con- trol scheme, SCWP, which is flexible and simplifies optical hardware requirements. The principal objective of the project is to understand the advantages and potential of optical packet switching compared to the conventional electronic approach. Several schemes for packet header implementation are described, using subcarrier multiplexing, separate wave- lengths, and serial transmission. A novel node design is introduced, based on wavelength router devices, which reduce loss, hence reducing booster amplifier gain and concomitant ASE noise. The fabrication of these devices, and also wavelength converters, are described. A photonic packet switching testbed is detailed which will allow the ideas developed with- in WASPNET to be tested in practice, permitting the practical problems of their implemen- tation to be determined. ABSTRACT T OPICS IN L IGHTWAVE