Lower Average Delay Research Articles

Analysis of storage and retrieval policies for ATM buffers

To satisfy the different performance (QoS) requirements of the different types of ATM traffic, several control strategies have been described in the literature. The control strategies that operate on cell level are combinations of storage and retrieval priority policies. In the comparisons, two classes of traffic are distinguished, one class (LL) requiring a low loss probability and one class (LD) requiring a low average delay (jitter). By using the delay-loss plane -- plotting the loss of the LL cells against the delay of the LD cells -- the performance of the various storage and (state-dependent) retrieval policies has been compared for both stationary and time-varying traffic. In a relative ranking of the different policies, the LDOLL threshold policies give the best results, in addition to creating the possibility of setting the threshold according to the desired delay-loss trade-off.

The fair distributed queue (FDQ) protocol for high-speed metropolitan-area networks

A new protocol, fair distributed queue (FDQ), suitable for very high-speed metropolitan-area networks (MANs), is presented. FDQ is a slotted system implemented on a unidirectional fiber bus. It has similarities to distributed queue dual bus (DQDB), the IEEE 802.6 Standard for MANs, including the bus topology and same nodal hardware. Like DQDB, FDQ achieves full throughput efficiency independent of the bus length, the transmission speed, and the number of nodes. Unlike DQDB, FDQ allocates equal bandwidth under heavy load to all active users in a time period less than or equal to the round-trip propagation delay without wasting bandwidth. Its delay characteristics are studied via simulation and compared to DQDB. FDQ has lower average delays under Poisson load than DQDB with or without the bandwidth balancing (BWB) mechanism. Two distinct implementations of priority levels are given and their characteristics are discussed. It is shown that FDQ's delay and throughput characteristics are little affected with increasing distances or transmission rates. Thus, FDQ possesses excellent scalability properties which allow its total length to extend over 100 km and transmission rate well above 1 Gb/s.

Temporal Patterning in Choice among Delayed Outcomes

Undergraduates chose repeatedly between two equal monetary outcomes. One outcome was more immediate, but the other resulted in less delay on future trials and a lower average delay per outcome overall. Minimizing average delay per outcome maximized overall reinforcement rate (money earned). The difference in immediacy between the two alternatives on a given trial was easily perceptible but the (opposed) difference in overall reinforcement rate was gradual and difficult to perceive. Subjects frequently chose the more immediate outcome and thus failed to maximize overall reinforcement rate. However, when choice-outcome (CO) units were grouped in triplets (as opposed to being presented singly at fixed intervals) subjects chose the more immediate outcome less often and thereby increased overall reinforcement rate. Findings are discussed in terms of impulsiveness and risk-aversion in laboratory studies and in everyday life.

Bandwidth allocation algorithms for packet video in ATM networks

Abstract We consider a two-node ATM network. Several video sources are multiplexed at the source node. Five alternative algorithms for bandwidth allocation to the video sources are considered: the Static Slot Assignment Algorithm, the Buffer-Population-Based Dynamic Slot Assignment Algorithm, the Adaptive Slot Assignment Algorithm, the First-Come First-Serve Slot Assignment Algorithm, and the Rate-Based Dynamic Slot Assignment Algorithm. The average cell delay and the coefficient of variation of the delay for the five algorithms are determined by computer simulation and compared. The Adaptive Slot Assignment Algorithm and the Buffer-Population-Based Dynamic Slot Assignment Algorithm have a built-in ability to detect the presence of abnormally high traffic at the point of multiplexing sent by a video source, and initiate corrective actions. Moreover, these two algorithms result in low average cell delay. We conclude that the Adaptive Slot Assignment Algorithm and the Buffer-Population-Based Dynamic Slot Assignment Algorithms are suitable for packet video in ATM networks.

Bandwidth Allocation for Broadband Multichannel Systems

Future broadband networks will need to support transmission and switching at rates ranging from hundreds of Gbits/s to a Tbits/s or more. A feasible approach in the realization of such networks is the use of sets of parallel links or channels which are termed trunk groups or channel groups. In this paper, we examine an optimal bandwidth allocation scheme for multichanne1 systems to switch trunk groups. This new multichannel switching system features a simple architecture, high throughput, high channel utilization, and other properties such as preserving packet sequence, being robust to unbalanced input traffic loads, providing fair access to network transmission resources, and fast packet processing via a pipeline structure. Numerical results show that the multichannel system can achieve a throughput of 91% per channel for a trunk group of 32 channels while also providing very low average cell delay performance in the presence of correlated inputs.

Packet reservation protocols for multichannel satellite networks

Owing to system economics, many satellite data networks have star configurations, with low cost user terminals communicating with high performance hub earth stations over inbound multiaccess channels and outbound broadcast channels. The inbound channel data rate is often limited by the power budget, so that multiple satellite channels are required to accommodate the overall traffic in a large network. Most packet multiaccess protocols previously considered, when applied to multichannel satellite networks, enable distinct groups of users to share individual channels. In contrast, this paper presents a new class of multichannel reservation protocols (MRPs) which improves channel utilisation by enabling complete sharing of all channels among all users. For good throughput efficiency and reduced delay, the MRPs employ slotted random access with immediate retransmissions for first-packet and reservation request transmissions, and reservation of consecutive time slots for transmissions of remaining packets in multipacket messages. Delay throughput characteristics are derived under steady state conditions. Numerical results show that the MRPs achieve higher throughput capacity than random access protocols and lower average delay than other reservation protocols.

Real-time packet switching: a performance analysis

The authors model the internal structure of a packet-switching node in a real-time system and characterize the tradeoff between throughput, delay, and packet loss as a function of the buffer size, switching speed, etc. They assume a simple shared-single-path switch fabric, though the analysis can be generalized to a wider class of switch fabrics. They show that with a small number of buffers the node will provide a guaranteed delay bound for high-priority traffic, a low average delay for low-priority traffic, no loss of packets at the input and low probability of packet loss at output.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

A Study of Data Multiplexing Techniques and Delay Performance

This paper is concerned with the study of data multiplexing techniques which provide local distribution for user populations whose source characteristics may be categorized as Inquiry/Response. The techniques studied are Polling, Random Access, and a Loop System. In each method a group of users is multiplexed onto the same line which is connected to a Central Processor. The Central Processor forms the interface between the local system and long haul facilities, but does no computation on arriving messages other than to route them appropriately. The advantage of such systems is flexibility of operation and economy through sharing of equipment at the Central Processor. We focus our attention on the average round-trip message delay and use this as a measure for comparisons of the three techniques. Source models, in terms of calls per busy hour and number of bits per message for users as well as for computer responses, are developed which are appropriate in an Inquiry/Response context. Other factors taken into consideration are transmission rates, synchronization delay time, and allocation of available capacity to the transmission of overhead information, such as polling messages and acknowledgment traffic. The results of the study are presented in graphic form, where the average delay due to traffic in the system is plotted as a function of the number of stations. Principal conclusions can be summarized as follows: Polling — It is found that this system is sensitive to the synchronization delay which takes place each time a user station transmits to the central facility. For reasonable choices of system parameters as many as 100 stations can share the Central Processor without exceeding an average of 1-second round trip message delay. Random Access — This system is not sensitive to synchronization delay. Over most of the range of load parameters, the Random Access system shows lower average delay than the other two systems. Loop System — The study revealed that this system has average delay

Routing in computer networks

AbstractThe problem of routing flow in a network of computers is extremely complex. This is especially formidable when routing is to be incorporated in iterative analysis and design. Among the properties of desirable flow patterns is low average delay from message inception to arrival. In this paper, we discuss procedures for minimizing average delay subject to a set of flow constraints. Heuristic routing procedures are presented and compared to optimum routing procedures. Computational experience is given.

Letter to the Editor—A Proof of the Optimality of the Shortest Remaining Processing Time Discipline

Letter to the Editor—A Proof of the Optimality of the Shortest Remaining Processing Time Discipline