DQRUMA (Distributed-Queueing Request Update Multiple Access) protocol has been proposed as an access protocol for the wireless ATM Local Area Networks. DQRUMA protocol is useful to transmit fixed-length packets (e. g. ATM cells). However, it cannot be applied to multimedia environment because it does not include any access control policy for multimedia traffic. In the paper, we propose a slot assignment scheme of DQRUMA protocol in wireless ATM LAN which supports integrated multimedia traffic with different service requirements. In this scheme we can allocate network resources according to the service requirements of each medium because the base station assigns Transmit-Permission flexibly according to the features of each medium.

Originally intended for application to B-ISDN, which is carrier oriented, ATM technology has been actively studied for application to LAN based environments since the beginning of the 1990s. One of the most notable things in LAN area is development of a rich set of application services. A number of technical specifications for major application services have been developed, which include LAN Emulation, IP over ATM, Multi-protocol over ATM, Voice and Telephony over ATM, as well as Native ATM services such as MPEG2 over ATM. Development of these new services raises new challenges related to traffic management. Keeping pace with the development, a number of traffic control mechanisms have also been developed to maximize the performance of these services. Traffic control and management techniques, however, are still in the early stage of their learning curve. Network engineers are facing challenging problems related to traffic management. This paper reviews major service-related technologies and discusses traffic management issues associated with these services. Especially, it describes the real world traffic management as practiced by average network engineers with state-of-the-art products. Although the thechnology developments have advanced through many research works, there seems to be a considerable gaps between the practice and principles. This paper discusses the traffic issues of ATM LAN from this perspective and points out some challenges for the future. Most of the difficulties in handling traffic issues stems from the differences in implementation details. To alleviate this difficulty, the introduction of a unified node model which describes the traffic handling capability of ATM nodes in sufficient detail is suggested.

We propose burst based bandwidth reservation method called FRP (Fast Reservation Protocol) in ATM LAN with general topology, and evaluate its performance. In FRP, the bandwidth is allocated on each link on burst basis, not on call basis. This enables an effective use of network resources when it is applied to highly bursty traffic, which can be typically found in data communications. The problem of FRP is that VCs traversing the different number of links experience different blocking probabilities as can be found in the conventional circuit-switching networks. In this paper, we treat a fairness issue in FRP-based ATM local area networks. The Max-Min flow control is adopted as the fair bandwidth allocation method to accomplish the fairness in the throughput. However, the original Max-Min flow control works in a centralized fashion, which is not desirable in the FRP-based ATM LAN. We therefore propose a "semi"-distributed Max-Min flow control suitable to FRP, in which each switch maintains its own local information about bandwidth usage of the connected links. Through simulation experiments, we show that the proposed semi-distributed Max-Min flow control can achieve the fairness among VCs as the original Max-Min flow control when the propagation delays are not large and the number of VCs is not so much.

Asynchronous Transfer Mode (ATM) shared buffer switches have numerous advantages, but have the principal disadvantage that all switch traffic must pass through the bottleneck of a single memory. To achieve the most efficient usage of this bottleneck, the shared buffer is made blockable, resulting in a switch architecture that we call "throttled-buffer", which has several advantageous properties. Shared buffer efficiency is maximized while decreasing both capacity and power requirements. Asynchronous operation is possible, whereby peak link data rates are allowed to approach the aggregate switch rate. Multicasting is also efficiently supported. The architecture and operation of this low-cost switch are described in detail.