The authors propose a computationally simple approximate expression for the equivalent capacity or bandwidth requirement of both individual and multiplexed connections, based on their statistical characteristics and the desired grade-of-service (GOS). The purpose of such an expression is to provide a unified metric to represent the effective bandwidth used by connections and the corresponding effective load of network links. These link metrics can then be used for efficient bandwidth management, routing, and call control procedures aimed at optimizing network usage. While the methodology proposed can provide an exact approach to the computation of the equivalent capacity, the associated complexity makes it infeasible for real-time network traffic control applications. Hence, an approximation is required. The validity of the approximation developed is verified by comparison to both exact computations and simulation results. >

/pdf/equivalent-capacity-and-its-application-to-bandwidth-4fytcdoedy.pdf

Equivalent capacity and its application to bandwidth allocation in high-speed networks

Many designs for integrated service networks offer a bounded delay packet delivery service to support real-time applications. To provide bounded delay service, networks must use admission control to regulate their load. Previous work on admission control mainly focused on algorithms that compute the worst case theoretical queueing delay to guarantee an absolute delay bound for all packets. In this paper we describe a measurement-based admission control algorithm for predictive service, which allows occasional delay violations. We have tested our algorithm through simulations on a wide variety of network topologies and driven with various source models, including some that exhibit long-range dependence, both in themselves and in their aggregation. Our simulation results suggest that, at least for the scenarios studied here, the measurement-based approach combined with the relaxed service commitment of predictive service enables us to achieve a high level of network utilization while still reliably meeting the delay bound.

/pdf/a-measurement-based-admission-control-algorithm-for-1fma064ciz.pdf

A measurement-based admission control algorithm for integrated services packet networks

The Asynchronous Transfer Mode (ATM) is the switching and multiplexing technique chosen by CCITT for the broadband access to the ISDN. The user-network interface offers one physical channel over which connections are multiplexed using short, fixed length packets (called cells). The ATM layer in the network performs relaying functions, with every cell carrying a label used for switching. This report is a tutorial. It aims at providing the reader with background information that is necessary to understand the debate about potential virtues and shortcomings of the ATM.

The Asynchronous Transfer Mode: a tutorial

An admission control scheme for real-time traffic requests in communication networks is disclosed. The scheme monitors tile performance of each link and predicts the traffic guarantees that can be offered when certain dummy traffic (pseudo traffic) mimicking the behavior of real-time streams is added to the existing traffic. When an application requests connection to the network, it specifies a Quality of Service (QoS) bound that it desires. A routing subsystem of tile network makes use of the predicted values while attempting to find a path over which tile application's QoS bound requirement will be satisfied. After the route is chosen, connection setup will involve verifying at each node along the route whether sufficient resources exist to support tile new connection. This explicit verification is likely to be necessary because the routing subsystem does not normally have up-to-date global state information. After admitting tile application, the network provides a soft guarantee that the QoS bound provided to it will be adhered to for most of the time. On the other hand, if no path through the network satisfies the QoS hound, tile application is denied connection to the network.

Flexible scheme for admission control of multimedia streams on integrated networks

In future wireless multimedia networks, there will be a mixture of different traffic classes which have their own maximum tolerable bit error rate (BER) requirements. In this paper, a novel medium access control (MAC) protocol called wireless multimedia access control protocol with BER scheduling (in short form, WISPER) for CDMA-based systems is proposed. WISPER utilizes the novel idea of scheduling the transmission of multimedia packets according to their BER requirements. The scheduler assigns priorities to the packets, and performs an iterative procedure to determine a good accommodation of the highest-priority packets in the slots of a frame so that packets with equal or similar BER requirements are transmitted in the same slots. The proposed WISPER protocol has been validated using a software emulator on the cellular environment. Performance evaluation results based on the implementation are also included.

A slotted CDMA protocol with BER scheduling for wireless multimedia networks

The concept of equivalent bandwidth of bursty traffic sources is investigated. The equivalent bandwidth measure depends on the source characteristics, the grade of service, and the interaction with the other type of traffic multiplexed on the same data link. The behavior of the statistical multiplexer is analyzed with inputs consisting of homogeneous bursty sources. The values of the equivalent bandwidth are analyzed for fixed statistical characteristics of the bursty sources and a fixed grade of service, and as a function of a background stream load. By assuming that the peak rate of the sources is sufficiently lower than the link rate (e.g. 1/15) and that the average burst length is greater than the queue length, the equivalent bandwidth independence of the average burst length and of the queue length is shown by simulation. The superposition of two heterogeneous bursty classes is analyzed. >

http://ieeexplore.ieee.org/iel2/123/3395/00117194.pdf

On bandwidth allocation to bursty virtual connections in ATM networks

The suggested class-related rule (CRR) was tested for mixes of two and three different bursty service classes. Performance was evaluated at the ATM cell level in terms of effective assigned bandwidth with respect to the peak bandwidth allocation. The gains obtained using the CRR are significant for the various bursty classes investigated with burstiness values b>or=5. It was found that the CRR is a conservative approach, based on the extensive simulation of a wide range of cases and of the assumed traffic and multiplexing models. Performance of the CRR was then evaluated at the virtual call level in terms of call blocking probability in a group of ATM trunks with respect to the peak bandwidth allocation. It was found that the CRR gain at the virtual call level is greater than the CRR gain at the ATM cell level, thus increasing the attractiveness of using statistical bandwidth assignment. All the results reported require further careful verifications, in particular to test other types of service classes and cases of multiple streams mixed with bursty and variable-bit-rate traffic sources. >

Bandwidth assignment and virtual call blocking in ATM networks

Congestion control mechanisms are required in ATM networks to fairly and efficiently allocate network resources to the various service types. Bandwidth allocation is a preventive control mechanism. On the basis of the traffic description given by the source, of the traffic load along the connection route and of the cell performance requirements, this mechanism gives a criterion to accept or reject a virtual connection request. A bandwidth called ‘equivalent bandwidth’ is allocated in order to ensure the cell transport objectives.



This paper proposes a definition of equivalent bandwidth for stream, bursty data and variable bit rate video (VBRV) sources. The equivalent bandwidth values are evaluated by simulation considering both homogeneous and heterogeneous sources. In the heterogeneous mixes with different cell loss probability requirements the selective discarding mechanism permits a more efficient utilization of the network resources.

Bandwidth allocation and selective discarding for variable bit rate video and bursty data calls in ATM networks

The paper analyses the problem of Medium Access Control in a Passive Optical Network, used to concentrate ATM user traffic. Feasible solutions for the problem are investigated and two protocols to resolve contention on the shared medium are presented, the single cell and the multiple cell frame based. The performance of the two protocols is evaluated, obtaining indications about the transfer delay introduced by the protocols and the amount of buffering needed at the user side.

Performance of shared medium access protocols for ATM traffic concentration

The authors focus on the application of adaptive end-to-end windowing as a reactive congestion control mechanism for data communications over asynchronous transfer mode (ATM) networks. In particular, an adaptive window mechanism including go-back (n) retransmission is proposed. The performance achievable by its application was evaluated by simulation under a particular network model assumption. The gain possible in terms of improvement of cell acceptance is estimated with different source traffic descriptors and network parameter values. Some comments on the interaction between the window mechanism and other congestion control mechanisms for ATM networks are given. >

Tiziana Toniatti

Papers

On bandwidth allocation to bursty virtual connections in ATM networks

Bandwidth assignment and virtual call blocking in ATM networks

Bandwidth allocation and selective discarding for variable bit rate video and bursty data calls in ATM networks

Performance of shared medium access protocols for ATM traffic concentration

Performance simulation of end-to-end windowing in ATM networks