Editorial introduction

There is a rather special relationship between queueing and communication systems, with a long and distinguished history of mutually beneficial interactions. For almost three quarters of a century, the field of communications has provided both a stimulus to the development of, and a rewarding application area for queueing theory methods. This association is just as strong now, and is amply illustrated in the present issue. The collection of papers in this issue mirrors trends and points of debate in today's communications technology. For instance, communications systems have traditionally been grouped into two broad categories, giving rise to different sets of problems: these were the voice carrying networks using circuit-switching technology, and the data carrying ones employing packet-switching. Although that distinction still exists, and will continue to exist for some time, the boundary between the two categories is being dismantled. The trend of recent and current developments is toward ISDN (Integrated Services Digital Network) and ATM (Asynchronous Transfer Mode) systems, which allow a number of widely disparate traffic streams to share the same broadband channel. This trend is attended by yet another set of interesting problems which require and are receiving the attention of network designers. Another theme of considerable topical interest reflected in several papers here is the scheduling of service to several disparate users to achieve efficiency, of course, but also fairness. Efficiency encompasses not only throughput and delay, traditional yardsticks, but also the amount of memory required, a factor of some potential importance in high speed commtmications with many users. The scheduler is required to protect well-behaved users from those that demand a disproportionate share of resources.