An increasing number of high performance internetworking protocol routers, LAN and asynchronous transfer mode (ATM) switches use a switched backplane based on a crossbar switch. Most often, these systems use input queues to hold packets waiting to traverse the switching fabric. It is well known that if simple first in first out (FIFO) input queues are used to hold packets then, even under benign conditions, head-of-line (HOL) blocking limits the achievable bandwidth to approximately 58.6% of the maximum. HOL blocking can be overcome by the use of virtual output queueing, which is described in this paper. A scheduling algorithm is used to configure the crossbar switch, deciding the order in which packets will be served. Previous results have shown that with a suitable scheduling algorithm, 100% throughput can be achieved. In this paper, we present a scheduling algorithm called iSLIP. An iterative, round-robin algorithm, iSLIP can achieve 100% throughput for uniform traffic, yet is simple to implement in hardware. Iterative and noniterative versions of the algorithms are presented, along with modified versions for prioritized traffic. Simulation results are presented to indicate the performance of iSLIP under benign and bursty traffic conditions. Prototype and commercial implementations of iSLIP exist in systems with aggregate bandwidths ranging from 50 to 500 Gb/s. When the traffic is nonuniform, iSLIP quickly adapts to a fair scheduling policy that is guaranteed never to starve an input queue. Finally, we describe the implementation complexity of iSLIP. Based on a two-dimensional (2-D) array of priority encoders, single-chip schedulers have been built supporting up to 32 ports, and making approximately 100 million scheduling decisions per second.

/pdf/the-islip-scheduling-algorithm-for-input-queued-switches-38q8iabi27.pdf

The iSLIP scheduling algorithm for input-queued switches

It is well known that head-of-line (HOL) blocking limits the throughput of an input-queued switch with FIFO queues. Under certain conditions, the throughput can be shown to be limited to approximately 58%. It is also known that if non-FIFO queueing policies are used, the throughput can be increased. However it has not been previously shown that if a suitable queueing policy and scheduling algorithm are used then it is possible to achieve 100% throughput for all independent arrival processes. In this paper we prove this to be the case using a simple linear programming argument and quadratic Lyapunov function. In particular we assume that each input maintains a separate FIFO queue for each output and that the switch is scheduled using a maximum weight bipartite matching algorithm.

/pdf/achieving-100-throughput-in-an-input-queued-switch-30gnma36nn.pdf

Achieving 100% throughput in an input-queued switch

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. >

WDM-based local lightwave networks. I. Single-hop systems

A network management system includes a domain administrating server (DAS) that stores a virtual catalog representing an overview of all files distributively stored across a network domain currently or in the past. The current and historical file information is used for assisting in auditing or locating files located anywhere in the domain. The current file information is used for assisting in transferring files across the domain. The domain administrating server (DAS) also includes a rule-base driven artificial administrator for monitoring and reacting to domain-wide alert reports and for detecting problematic trends in domain-wide performance based on information collected from the network domain.

Network management system having virtual catalog overview of files distributively stored across network domain

The algorithms described in this thesis are designed to schedule cells in a very high-speed, parallel, input-queued crossbar switch. We present several novel scheduling algorithms that we have devised, each aims to match the set of inputs of an input queued switch to the set of outputs more efficiently, fairly and quickly than existing techniques.
In Chapter 2 we present the simplest and fastest of these algorithms: SLIP--a parallel algorithm that uses rotating priority ("round-robin") arbitration. SLIP is simple: it is readily implemented in hardware and can operate at high speed. SLIP has high performance: for uniform i.i.d. Bernoulli arrivals, SLIP is stable for any admissible load, because the arbiters tend to desynchronize. We present analytical results to model this behavior. However, SLIP is not always stable and is not always monotonic: adding more traffic can actually make the algorithm operate more efficiently. We present an approximate analytical model of this behavior. SLIP prevents starvation: all contending inputs are eventually served. We present simulation results, indicating SLIP's performance. We argue that SLIP can be readily implemented for a 32 x 32 switch on a single chip.
In Chapter 3 we present i-SLIP, an iterative algorithm that improves upon SLIP by converging on a maximal size match. The performance of i-SLIP improves with up to log$\sb2N$ iterations. We show that although it has a longer running time than SLIP, an i-SLIP scheduler is little more complex to implement.
In Chapter 4 we describe maximum or maximal weight matching algorithms based on the occupancy of queues, or waiting times of cells. These algorithms are stable over a wider range of traffic loads. We describe two algorithms, longest queue first (LQF) and oldest cell first (OCF) and consider their performance. We prove that LQ although too complex to implement in hardware, is stable under all admissible i.i.d. offered loads. We consider two implementable, iterative algorithms i-LQF and i-OCF which converge on a maximal weight matching. Finally, we present two interesting implementations of the Gale-Shapley algorithm, designed to solve the stable marriage problem.

/pdf/scheduling-algorithms-for-input-queued-cell-switches-5becja7d2v.pdf

Scheduling algorithms for input-queued cell switches

A number of cases of slotted Aloha protocols for high-speed optical fiber local area networks (LANs) are presented and analyzed. The results are compared to the cases of Slotted Aloha/Aloha schemes presented by I.M.I. Habab et al., (1987) and N. Mehravori (1990). The proposed schemes were found to offer better throughput and delay characteristics. These cases are also compared with Aloha/N-server and CSMA/N-server protocols presented by Habab and Mehravori. Some of the considered protocols offer better throughput when the data packet lengths are small. Two cases of reservation Aloha protocols are presented and analyzed. >

Slotted Aloha and reservation Aloha protocols for very high-speed optical fiber local area networks using passive star topology

A simultaneous real-time data delivery (SRTDD) control scheme that is based on a real application scenario developed in the Multimedia Communications Research Laboratory at the University of Ottawa is addressed. SRTDD refers to simultaneously delivering multimedia data in different data streams belonging to the same time interval. It is one of the basic synchronization problems involved in multimedia distributed systems. The details of the SRTDD problem are investigated. Based on the new control scheme for multimedia data delivery constraints on the size of the synchronization control unit are obtained. The segment delivery protocol is developed. A practical segmentation method is introduced for real-time voice and video. A performance analysis of the delivery protocol with this segmentation method is also conducted. >

Synchronization in real time multimedia data delivery

Collaborative virtual environment (CVE) concepts have been used in many systems. The architectures available today provide support for a number of users but they fail if too many users are together in a small "space" in the virtual world. This paper introduces VELVET, an adaptive hybrid architecture which allows a greater number of users to interact through a CVE. VELVET introduces a novel adaptive area of interest management, which supports heterogeneity amongst the various participants.

VELVET: an adaptive hybrid architecture for very large virtual environments

An investigation is made of the performance of a nonblocking packet switch having input buffers and a limited amount of buffers within the switch fabric, where contention for the output ports occurs. This technique improves the performance significantly, while maintaining the switch fabric speed equal to that of the port speed. For uniform traffic, a 16*16 switch with head of line priority scheduling has an achievable throughput equal to 87.5%. The simulation results suggest that for unbalanced and bursty traffic, the performance of the switch does not degrade appreciably. >

16*16 limited intermediate buffer switch module for ATM networks

Multi-channel/port communication structures are often found in multiaccess networks and switching systems, where data packets from input channels/ports are directed to their destinations of output channels/ports, following certain traffic assignment rules. The authors propose a simple algorithm, the K-HOL algorithm, to expedite the processing of multi-channel/port traffic assignment. Recursive equations are given to find the parameter K and the maximum achievable throughput for the algorithm. They compare this algorithm with two typical algorithms found in the literature. Simulation results show that the proposed algorithm has close performance to the two algorithms whereas the computational complexity is reduced greatly. >

http://ieeexplore.ieee.org/iel2/2977/8446/00369015.pdf

N.D. Georganas

Papers

Slotted Aloha and reservation Aloha protocols for very high-speed optical fiber local area networks using passive star topology

Synchronization in real time multimedia data delivery

VELVET: an adaptive hybrid architecture for very large virtual environments

16*16 limited intermediate buffer switch module for ATM networks

A fast algorithm for multi-channel/port traffic assignment