The delivery delay in a point-to-point packet switching network is difficult to control due to the contention among randomly-arriving packets at each node and multihops a packet must travel between its source and destination. Despite this difficulty, there are an increasing number of applications that require packets to be delivered reliably within prespecified delay bounds. This paper shows how this can be achieved by using real-time channels which make "soft" reservation of network resources to ensure the timely delivery of real-time packets. We first present theoretical results and detailed procedures for the establishment of real-time channels and then show how the basic real-time channels can be enhanced to be fault-tolerant using the multiple disjoint paths between a pair of communicating nodes. The contribution of the former is a tighter schedulability condition which makes more efficient use of network resources than any other existing approaches, and that of the latter is a significant improvement in fault tolerance over the basic real-time channel, which is inherently susceptible to component failures.

We propose a flow-control scheme that combines the merits of credit- and rate-based flow-control schemes by applying direct control over both bandwidth and buffer resources. The goal of the proposed scheme is to design an optimal rate-control policy for a given finite buffer capacity that maximizes the average throughput and bounds the end-to-end delay. By applying higher-order rate control, the proposed scheme not only makes the rate process converge to the neighborhood of the link bandwidth, but also confines the queue-length fluctuation to a regime bounded by the buffer capacity (thus guaranteeing lossless transmission). Using the fluid approximation method, we model the proposed flow-control scheme and study the system dynamic behavior for ABR (available bit rate) service under the most stressful traffic condition. We derive the expressions for queue build-ups and average throughput in both transient and equilibrium states. The analytical results have shown the proposed scheme to be stable and efficient in that the source rate and bottleneck queue length rapidly converge to the designated operating region. Also presented, are examples showing that the proposed scheme outperforms the other existing schemes.

Synchronous messages are usually generated periodically and each of them is required to be transmitted before the generation of the next message. Due to the inherent deficiency in its medium access control (MAC) protocol, an FDDI token ring can use at most one half of its ring bandwidth to transmit such synchronous traffic. This deficiency greatly reduces the FDDI's capability of supporting multimedia applications like real-time voice/video transmissions. In this paper, we show how a few simple modifications to the FDDI's MAC protocol can remove this deficiency and double a ring's ability of supporting synchronous traffic. The modified protocol, called FDDI-M, can also achieve a higher throughput for asynchronous traffic than the standard FDDI and the FDDI-II, thus making it useful even for those networks without heavy synchronous traffic.

Fault-tolerant, real-time communication in distributed systems is very important yet difficult to achieve. Traditional protocols like the TCP/IP achieve reliable communication through acknowledgment and retransmission schemes, where one achieves the reliability at the cost of performance, In this paper, we discuss how both the timeliness and fault-tolerance of communication can be achieved by using the concept of real-time channel and exploring the inherent spatial redundancy of a given network topology, Specifically, we show how isolated failure immune real-time channels can be established in wrapped hexagonal mesh networks, thus ensuring timely delivery of messages in the presence of network component failures as long as the failures are isolated. This kind of fault-tolerance cannot be achieved with other commonly-known topologies like rings, rectangular meshes, and hypercubes. The proposed approach is to be implemented in an experimental distributed real-time system, called HARTS, whose construction is underway.>

It is well known that an FDDI token ring network provides a guaranteed throughput for synchronous messages and a bounded medium access delay for each node/station. However, this fact alone cannot effectively support many real-time applications that require the timely delivery of each critical message. The reason for this is that the FDDI guarantees a medium access delay bound to nodes, but not to messages themselves. The message-delivery delays may exceed the medium-access delay bound even if a node transmits synchronous messages at a rate not greater than the guaranteed throughput. We solve this problem by developing a synchronous bandwidth allocation (SEA) scheme which calculates the synchronous bandwidth necessary for each application to satisfy its message-delivery delay requirement. The result obtained in this paper is essential for effective use of the FDDI token ring networks in supporting such real-time communication as digital video/audio transmissions, and distributed control/monitoring.

Synchronous messages are usually generated periodically and each of them is required to be transmitted before the generation of the next message. Due to the inherent deficiency in its Medium Access Control (MAC) protocol, an FDDI token ring can use at most one half of its ring bandwidth to transmit such synchronous traffic. This deficiency greatly reduces the FDDI's capability of supporting multimedia applications like real-time voice/video transmissions. In this paper, we show how a few simple modifications to the FDDI's MAC protocol can remove this deficiency and double a ring's ability of supporting synchronous traffic. The modified protocol, called FDDI-M, preserves all other good features of an FDDI network and can also achieve a higher throughput for asynchronous traffic than the standard FDDI and the FDDI-II, thus making it useful even for those networks without heavy synchronous traffic.

The asynchronous transfer mode (ATM) is expected to be the multiplexing and switching technique for the future broadband integrated service digital networks (BISDNs) which can transport almost all types of traffic including bursty data traffic and continuous voice/video frames. However, this capability cannot be realized without a proper congestion-control scheme. This paper discusses the suitability of the real-time channel approach for congestion control in ATM networks. A main contribution of the paper is the development of a new message-transmission scheme for real-time channels which does not require each cell to carry deadline information which is difficult to do with the standard ATM cell format. This makes our scheme compatible with, and implementable in, ATM networks. The new scheme also has a number of advantages over existing ones. The achievable quality of service with the proposed approach would make ATM networks capable of supporting both real-time services and dynamic bandwidth sharing which have been promised but not yet fully realized.>

There are numerous applications which require packets to be delivered within pre-specified delay bounds in point-to-point packet-switched networks. To meet this requirement, we define a real-time channel as a unidirectional connection between two nodes in such a network that guarantees every packet to be delivered before a user-defined, end-to-end deadline. The goal of this paper is to lay a mathematical basis for the problem of establishing real-time channels by (i) deriving a necessary and sufficient condition for the schedulability of a set of channels over a link, and (ii) developing an efficient method for calculating the minimum delay bound over a link for each channel. Given the traffic characteristics of a channel, our results can be used to check whether or not every packet will be delivered within a pre-specified delay bound. The results are also applicable to a wide variety of real-time task scheduling problems.>

It is well-known that an FDDI token ring network provides a guaranteed throughput for synchronous messages and a bounded medium access delay for each node#station. However, this fact alone cannot e#ectively support many real-time applications that require the timely delivery of each critical message. The reason for this is that the FDDI guarantees a medium access delay bound to nodes, but not to messages themselves. The message-delivery delays may exceed the medium-access delay bound even if a node transmits synchronous messages at a rate not greater than the guaranteed throughput. We solve this problem by developing a synchronous bandwidth allocation #SBA# scheme which calculates the synchronous bandwidth necessary for each application to satisfy its message-delivery delay requirement. The result obtained in this paper is essential for e#ective use of the FDDI token ring networks in supporting such real-time communications as digital video#audio transmissions and distributed control#monitoring.

It is well known that some time-token medium access protocols for local area networks (LANs) like the IEEE 802.4 token bus and the FDDI token ring can guarantee the medium access delay for time-constrained packets. However, a problem which has been largely overlooked is how these protocols can be made to provide a maximum throughput for nontime-constrained packets while guaranteeing the delay bound of time-constrained packets. The authors first show how the parameters of the IEEE 802.4 token bus and the FDDI token ring can be set to solve the above problem. Then, they design a new timer mechanism for the timed-token protocols which provides the highest guaranteed throughput of nontime-constrained packets among a set of medium access protocols called the token passing protocol, to which most of the existing non-contention LAN protocols belong. They present numerical examples to compare different protocols, all of which have shown the superiority of the proposed protocol to the others.>

The feasibility of using buffered transmission, which has the advantages of eliminating the token passing overhead and relieving the network interface from the complex token handling operations, is considered. The concept of a real-time channel (RTC) and the RTC establishment procedures are reviewed. Its performance is then compared with the FDDI timed token protocol. Simulation results show that the buffered transmission in local area ring networks equipped with real-time channels enhances both the networks' throughput and the ability to support heterogeneous real-time traffic. The implementation of the network interface is also examined with an example design showing the feasibility of real-time channels in high-speed local area networks.>

Collision-free multirobot motion planning can be achieved in two steps: path planning and trajectory planning. Path planning finds for the robots geometric paths to avoid collision with static obstacles and trajectory planning determines how fast each robot must move along its geometric path to avoid collision with other moving robots. For a dual-robot system, a simple trajectory planning strategy is to let each robot move along its path as fast as possible and delay one robot at its initial position to avoid collision with the other robot. A sufficient condition under which this simple (thus easy to implement) strategy for dual-robot systems can achieve time optimality as well as collision avoidance, i.e., the two robots reach their final positions without colliding with each other in the minimum amount of time, is derived. A demonstrative example is presented, showing how this strategy can be used for loading and unloading applications.>

The problem of scheduling job operations in an automatic assembly line used for manufacturing a small to medium volume of mixed workparts is addressed. An assembly line model is discussed. This model differs from the classical flow shop model in the following three aspects: there are no buffers at machine stations; constraints associated with the material transport system are included; and, for each batch of production, workparts are distinguished in groups, rather than individually. An optimal algorithm that requires very little computation is derived first by minimizing the total finish time for two machine assembly lines. This result is then generalized to the problem of scheduling an assembly line with m>2 machines processing single-operation jobs. In order to reduce the computational complexity of the latter problem, heuristic algorithms are proposed and shown to work quite well for all the cases considered. A heuristic solution to the problem of scheduling an assembly line with m>2 machines processing multioperation jobs is discussed.>

The problem of scheduling job operations in an automatic assembly line used for manufacturing a small to medium volume of mixed workparts is addressed. The assembly line model differs from the classical flow shop model in the following three aspects: (1) there are no buffers at machine stations, (2) constraints associated with the material transport system are included, and (3) for each batch production, workparts are distinguished in groups, rather than individually. An optimal algorithm that requires very little computation is derived by minimizing the total finish time for two machine assembly lines. This result is generalized to the problem of scheduling an assembly line with m>2 machines processing single-operation jobs. In order to reduce the computational complexity of the latter problem, heuristic algorithms are also proposed and shown to work quite well for all the cases considered. The solution to the problem of scheduling an assembly line with m>2 machines processing multioperation jobs is discussed.>