In this paper, we address issues related to providing guaranteed real-time communication in ATM local area networks. We concentrate on output link scheduling because it plays a critical role in meeting message deadlines. We are particularly interested in the weighted round robin scheduling policy because of its simple design and implementation. To use weighted round robin scheduling for hard real-time applications, the weights must be properly allocated to each of the connections. We propose and analyze two weight allocation schemes. The first scheme is heuristic, and is easy to understand and implement. The second scheme is optimal. That is, it can always guarantee a set of hard real-time connections whenever it is possible to do so. We evaluate and compare the system performance in terms of its admission probability-the probability that deadlines of all connections in a randomly chosen connection set can be met. We find that the optimal weight allocation scheme indeed performs the best. However, the heuristic scheme performs closely to the optimal scheme over a wide range of loading conditions.

We study the effect of the output link scheduling discipline of an ATM switch on the ability of an ATM LAN to admit real-time connections. Three output link scheduling policies are studied: first come first served (FCFS), round robin (RR), and packet-by-packet generalized processor sharing (PGPS). We derive connection admission criteria for the three scheduling policies. To evaluate the performance of the three scheduling policies, we introduce the metric of admission probability. The admission probability gives the probability that a randomly chosen set of real-time connections will be admitted into the network. The admission probability allows system designers to study the performance of different scheduling policies over a wide range of network loads. We observe that the performance of the three scheduling policies is sensitive to message deadlines. When the deadlines are small, PGPS outperforms both RR and FCFS, and RR outperforms FCFS. When the deadlines are large, all three scheduling policies perform the same. We also note that although PGPS is better than RR and FCFS most of the time, its improved performance is achieved at the cost of high implementation complexity and run time overheads.>

Issues related to guaranteeing synchronous messages with arbitrary deadline constraints in a fiber distributed data interface (FDDI) network are addressed. It is shown that several network parameters must be set carefully if message deadlines are to be satisfied. Message deadlines can only be met if sufficient synchronous bandwidth is allocated to each mode. Thus, proper synchronous bandwidth allocation is essential if deadlines are to be guaranteed. The target token rotation time (TTRT) determines both the speed of token circulation and the network utilization available to user applications. TTRT should also be chosen carefully to ensure that the token circulates fast enough while maintaining a high available utilization. Sufficient buffer space must be provided for outgoing messages, otherwise messages could be lost due to buffer overflow. An integrated method for allocating the synchronous bandwidth and selecting TTRT so that the time constraints of synchronous messages with arbitrary deadlines are guaranteed to be met is proposed and analyzed.

We study the performance of FDDI networks in terms of their guarantee probability, i.e., the probability that a set of synchronous messages are guaranteed to meet their deadlines. Traditional techniques such as queuing analysis cannot be directly used to derive the guarantee probability. To counter this problem, we develop a new geometric model of schedulability. Based on this model, we obtain a numerical method to compute the exact values of the guarantee probability. A closed-form approximation for the guarantee probability is also derived, and is shown to be relatively accurate and computationally efficient. The network performance is then systematically examined in terms of the guarantee probability. We find that there is a high probability that a randomly chosen message set can be guaranteed even when the real-time traffic is increased beyond the worst case achievable utilization bound. Hence, FDDI networks are applicable for real-time applications in a wide range of loading conditions.>

Recent developments in real-time communications in local-area networks are reviewed. Issues relating to the transmission of both synchronous and asynchronous messages are examined. Priority-driven protocols, global reservation protocols, and protocols with bounded access times are discussed.>

Version selection schemes for hard real-time communications in multiaccess networks are studied. Two classes of version selection schemes are proposed: those that are based on message time constraints and those that are based on the network traffic. The authors discuss the design and implementation of both classes of version selection schemes used in conjunction with a window-based media access protocol. Integrated version selection schemes that combine both the timing- and the traffic-based approaches are discussed. The simulation results show that as the network load increases, the version selection schemes are effective at reducing the message loss by trading off the proportion of long version messages sent.>

A classification scheme is proposed which divides real-time messages into three different classes according to the impact on system performance when the deadline of a message is missed. A study is made of several multiaccess protocols that take into account the time constraints of messages as well as their classifications and that are able to given an early notice when an important message will miss its deadline. The performances of these protocols are evaluated and compared through simulation.>