Network support for variable bit-rate video needs to consider (i) properties of workload induced (e.g., significant auto-correlations into far lags and different marginal distributions among connections) and (ii) application specific bounds on delay-jitter and statistical cell-loss probabilities. The objective of this paper is to present a quality-of-service solution for such traffic at each multiplexing point in the network. Heterogeneity in both offered workload and quality-of-service needs are addressed. The framing strategy introduced by Golestani is generalized by inducing the framing structure on virtual circuits (VCs) rather than on communication links. This results in eliminating the phase mismatch between incoming and outgoing frames at the switches. Hence, a significant reduction in delay and buffer requirements at each switch traversed by the VC is achieved. An architecture design that can be implemented in a moderate cost is also presented. The framing strategy is combined ...

A dynamic bandwidth allocation strategy to support variable bit rate #VBR# video tra#c is proposed. This strategy predicts the bandwidth requirements for future frames using either adaptive or non-adaptive least mean square #LMS# error linear predictors. The adaptive technique does not require any prior knowledge of the statistics, nor assumes stationarity. Several reservation schemes and pro-active congestion approach are also presented. Analysis using six one-half hour video traces indicate that prediction errors for the bandwidth required for the next frame are almost white noise. By reserving bandwidth equal to the predicted value, only the prediction errors need to be bu#ered. Because the errors are almost white noise, small bu#ers size, high utilization, and small delay are achieved. Simulation results show that for the same expected cell loss, bu#ers size is reduced by more than a factor of 100 and network utilization is increased more that 250 # as compared to a #xed...

this report three specific traffic types have been chosen as representative: Ethernet local area network (LAN) traffic, MPEG (Motion Pictures Experts Group) compressed video, and Internet WWW traffic. Video traffic is important because it is one of the most bandwidth-intensive applications that ATM networks are being used for. Of the different types of video, variable bit rate (i.e. compressed) video has been identified as being important because of its large and widely varying bandwidth requirements.

Future integrated-services networks are expected to support applications with a wide range of service requirements. The most demanding applications require a bounded-delay service that provides deterministic (i.e., worst-case) bounds on network latencies for all packets. To provide such delay guarantees, a network must allocate network resources such as bandwidth and buffer space to individual connections. However, since resource availability is limited, the network must carefully manage its resources in order to ensure a high achievable network utilization.

also sponsored through generous grants from the EMC Corporation and the Intel

Video is becoming the most important data in asynchronous transfer mode (ATM) networks. In ATM networks, image quality remains almost the same by encoding a video signal at variable bit rates (VBRs). Moving picture experts group (MPEG) video consists of three different frames: intra (I), predictive (P), and bidirectional (B). The important feature of VBR MPEG video traffic is the long-range dependence (LRD) characteristic. To examine the LRD characteristic of real MPEG video sequences, the Hurst parameter is employed. This paper presents a wavelet method, a line length method, and a Fourier filtering method for Hurst parameter estimation and compares their performance of LRD analysis with various video data. The relationship between the Hurst parameter and parameters in fractal modeling is also investigated. C ○ 2001 Academic Press 1.

Abstract—Size-based policies have been shown in the literature to effectively balance load and improve performance in cluster environments. Size-based policies assign jobs to servers based on the job size and their performance improvements are an outcome of separating “short ” from “long ” jobs, by avoiding having short jobs waiting behind long jobs for service. In this paper, we present evidence that performance improvements due to this separation quickly vanish if the arrival process to the cluster is autocorrelated. Based on our observations, we devise a new size-based policy called D_EQAL that still strives to separate jobs to servers according to job size but this separation is now biased by an effort to reduce performance loss due to autocorrelation in the arrival flows of jobs that are directed to each server. As a result of this bias, all servers may not be equally utilized (i.e., load in the system may be “unbalanced”), but performance benefits become significant. D_EQAL can be used on-line as it does not assume any a priori knowledge of the incoming workload. Extensive simulations show the effectiveness of D_EQAL under autocorrelated and uncorrelated arrival streams and illustrate that the policy successfully self-adjusts the degree of load unbalancing based on monitored performance measures. Index Terms—Load balancing, autocorrelated arrivals, highly variable service times, self adaptive policies. Ç 1

Recent studies indicate that positive auto-correlations in input traffic dramatically increase queue length statistics at a multiplexor. Further, traffic over several data networks has been found to possess a slowly decaying auto-correlation structure. Accurate studies of these and similar environments need to be based on realistic traffic. This paper introduces a traffic model for network performance analysis that appears to preserve temporal dependencies and the marginal distribution of the packet arrival process for TCP traffic over the NSFNET T3 backbone network.

this report, all stationary processes are denoted by X(t) and their mean, EfX(t)g, are taken as 0. The organization of this report is as follows: In Section 2, we briefly review (i) ARIMA processes, (ii) analytical derivations and properties of fARIMA processes, and (iii) self-similar processes. In Section 3, we describe the parameter estimation methods for fARIMA process. In Section 4, we describe the synthesis of fARIMA process and present the results of performance evaluation of estimation methods described in Section 3. In Section 5, we present an application of fARIMA processes in network traffic modeling. Finally, conclusions are drawn in Section 6. 2 fARIMA Processes

Traffic models are at the heart of any performance evaluation of telecommunications networks. An accurate estimation of network performance is critical for the success of broadband networks. Such networks need to guarantee an acceptable quality of service (QoS) level to the users. Therefore, traffic models need to be accurate and able to capture the statistical characteristics of the actual traffic. In this paper we survey and examine traffic models that are currently used in the literature. Traditional short range and non-traditional long range dependence traffic models are presented. Number of parameters needed, parameter estimation and analytical tractability of traffic models are discussed. Traffic models are analyzed based on their ability to capture marginal distribution, auto-correlation structure, and ultimately predict delay and cell loss probabilities of the actual data they model. This research was supported in part by the National Science Foundation under grant NCR-939629...