No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
Throughput analysis of WDM-based dual-bus local area networks
Abstract
This paper investigates the possibility of using wavelength division multiplexing to enhance the system capacity of local area networks with the bus topology. We first consider the choice between wavelength fixed and wavelength tunable transceivers based on practical considerations, and we conclude that the use of fixed wavelength transceivers allows one to employ WDM while still using standard interface cards. This requires that wavelength translation be implemented in the network. The paper then considers the problem of wavelength translation and introduces three approaches: the min-max approach that minimizes the maximum load on channels; the Translation at Destination approach in which wavelength translation is done only at the ingress point to the receiving channel, and the Translation at Source approach in which the wavelength translation is done at the egress point from the source. Practical algorithms for implementing these optimal allocations are considered. We then present throughput analysis of dual-bus networks employing the proposed translation approaches
ResearchGate has not been able to resolve any citations for this publication.
This article investigates the possibility of using wavelength division multiplexing to enhance the system capacity of local area networks with the bus topology, and which employ some type of distributed or centralized reservation. We first consider the choice between wavelength fixed and wavelength tunable transceivers based on practical considerations, and we conclude that the use of fixed wavelength transceivers allows one to employ WDM while still using standard interface cards. This requires that wavelength translation be implemented in the network. The article then considers the problem of wavelength translation and introduces three approaches: the min–max approach that minimizes the maximum load on channels; the Translation at Destination approach where the wavelength translation is carried out only at the ingress point to the receiving channel, and the Translation at Source approach in which the wavelength translation is carried out at the egress point from the source. Practical algorithms for implementing these optimal allocations are considered. We consider heterogeneous systems within the scope of two problems: the load distribution on different channels, and the system capacity. A linear programming approach is introduced in order to assess the capacity enhancement in the general heterogeneous case. In the homogeneous case, closed form expressions can be obtained. Finally, we present simulation results based on these algorithms for a DQDB network in order to assess the degree of performance enhancement.
Cyclic-reservation multiple access is presented for networks based
on slotted unidirectional bus structure, both folded and dual-bus
configurations. The cyclic-reservation access provides both throughput
efficiency and flexibility in capacity allocation, which can be used to
achieve any set of fairness conditions. The reservation-cancellation
backpressure minimizes the worst-case access delay. CRMA provides the
capability of reserving consecutive slots for transmitting a packet,
thus facilitating reassembly, and is well suited to
supercomputer-networking applications such as visualization,
high-resolution graphics, and image-retrieval systems. The recovery from
transmission errors is simple owing to the centralized-global-queue
approach. These properties make CRMA suitable for both LANs and MANs in
the gigabit-per-second, as well as the 100-Mb/s speed range
The Distributed Queue Dual Bus (DQDB) protocol, which has been specified by the IEEE 802.6 as the Metropolitan Area Network (MAN) standard, has the performance limiting feature in which busy slots must continue to propagate downstream, even if they have already been read. We address the issue of slot reuse in DQDB. We introduce and evaluate the performance of a new erasure node algorithm. The algorithm effectively balances the slot reuse and request cancellation functions by buffering the request stream and only passing those requests that have not been satisfied by erased slots. We show that our algorithm possesses a number of desirable features which makes it superior to existing algorithms.
Distributed-queue multiple-access (DQMA) and cyclic-reservation
multiple-access (CRMA) access schemes for gigabit-per-second local and
metropolitan area networks are described. These schemes are based on a
slotted unidirectional bus structure, with both folded and dual-bus
configurations. Like the distributed-queue dual-bus (DQDB) scheme, both
the DQMA and the CRMA access schemes achieve full throughput independent
of network speed and distance. They have two significant advantages over
DQDB. They provide throughput fairness even at high speeds and large
distances, where DQDB exhibits dramatic unfairness, and, by allowing
reservation of multiple consecutive slots, they make segment labeling
unnecessary and facilitate packet reassembly significantly
For pt.I see ibid., vol.6, no.3, p.12-27, 1992. A survey of wavelength-division-multiplexing (WDM)-based local lightwave networks is presented. The general characteristics of multihop systems are discussed, and various multihop approaches are reviewed. The construction of optimal structures based on minimizing the maximum link flow and optimizations based on minimization of the mean network packet delay are also reviewed. Regular topologies that have been studied as candidates for multihop lightwave networks, including the perfect shuffle, the de Bruijn graph, the toroid, and the hypercube, are discussed. Near-optimal node placement algorithms and shared-channel multihop systems are also discussed.< >
An overview of emerging all-optical networks is given. The characteristics and alternative architectures for single-hop systems are discussed. The characteristics of lightwave technology that facilitate the design of wavelength-division-multiplexing (WDM) networks are reviewed, and it is explained how WDM local networks can be built based on the single-hop and multihop approaches. Various categories of single-hop systems are discussed: experimental systems, systems based on no pretransmission coordination, and systems based on pretransmission coordination, which also require a separate control channel. A simple classification for single-hop systems is provided.< >
Following a brief introduction to the applications for wavelength
conversion and the different available conversion techniques, the paper
gives an in depth analysis of cross gain and cross phase wavelength
conversion in semiconductor optical amplifiers. The influence of
saturation filtering on the bandwidth of the converters is explained and
conditions for conversion at 20 Gb/s or more are identified. The cross
gain modulation scheme shows extinction ratio degradation for conversion
to longer wavelengths. This can be overcome using cross phase modulation
in semiconductor optical amplifiers that are integrated into
interferometric structures. The first results for monolithic integrated
interferometric wavelength converters are reviewed, and the quality of
the converted signals is demonstrated by transmission of 10 Gb/s
converted signals over 60 km of nondispersion shifted single mode fiber
WDM networks make a very effective utilization of the fiber
bandwidth and offer flexible interconnections based on wavelength
routing. In high capacity, dynamic WDM networks, blocking due to
wavelength contention can he reduced by wavelength conversion.
Wavelength conversion addresses a number of key issues in WDM networks
including transparency, interoperability, and network capacity. Strictly
transparent networks offer seamless interconnections with full
reconfigurability and interoperability. Wavelength conversion may be the
first obstacle in realizing a transparent WDM network. Among numerous
wavelength conversion techniques reported to date, only a few techniques
offer strict transparency. Optoelectronic conversion (O/E-E/O)
techniques achieve limited transparency, yet their mature technologies
allow deployment in the near future. The majority of all-optical
wavelength conversion techniques also offer limited transparency but
they have a potential advantage over the optoelectronic counterpart in
realizing lower packaging costs and crosstalk when multiple wavelength
array configurations are considered. Wavelength conversion by
difference-frequency generation offers a full range of transparency
while adding no excess noise to the signal. Recent experiments showed
promising results including a spectral inversion and a 90 nm conversion
bandwidth. This paper reviews various wavelength conversion techniques,
discusses the advantages and shortcomings of each technique, and
addresses their implications for transparent networks
The very broad bandwidth of low-loss optical transmission in a
single-mode fiber and the recent improvements in single-frequency
tunable lasers have stimulated significant advances in dense wavelength
division multiplexed optical networks. This technology, including
wavelength-sensitive optical switching and routing elements and passive
optical elements, has made it possible to consider the use of wavelength
as another dimension, in addition to time and space, in network and
switch design. The independence of optical signals at different
wavelengths makes this a natural choice for multiple-access networks,
for applications which benefit from shared transmission media, and for
networks in which very large throughputs are required. Recent progress
in multiwavelength networks are reviewed, some of the limitations which
affect the performance of such networks are discussed, and examples of
several network and switch proposals based on these ideas are presented.
Discussed also are critical technologies that are essential to progress
in this field