Abstract

This dissertation examines issues concerning real-time communication in ATM LANs such as those used in Hard Real-Time (HRT) systems. Applications in such systems have stringent quality of service (QoS) requirements. Support of hard real-time applications in ATM LANs requires efficient traffic management functions. This work studies two such functions: connection admission control (CAC) algorithm and input traffic regulation. In hard real-time systems, the QoS requirements are specified in terms of end-to-end cell deadlines and no cell loss due to buffer overflow. A CAC algorithm must admit a new connection only if its QoS requirements can be met without violating that of the previously admitted connections. Hence, computation of the worst case end-to-end cell delays and buffer requirements are pivotal issues in CAC design. Arbitrary topology of the network complicates the problem as it might lead to cyclic dependencies among various connections, making the delay and buffer analysis difficult. The first major contribution of this work is the development of a general framework to efficiently analyze an ATM LAN. The efficiency of the end-to-end delay and buffer analysis depends on the effectiveness of the traffic descriptors used in the network. We propose a comprehensive traffic description function, $\Gamma$(I). $\Gamma$(I) provides adequate information about the worst case traffic behavior of connections anywhere in the network and its use simplifies the analysis of ATM LANS, even those which do not support internal traffic restoration mechanisms. Further, simple approximations of $\Gamma$(I), proposed in this work, alleviate the space and time constraints imposed by the network on the CAC. We use the proposed analysis methodology to design an effective and efficient CAC algorithm for ATM LANs with arbitrary topology. The second major contribution of this work is the use of traffic regulation as a technique to meet the delay requirements of connections. This is the first work which uses leaky bucket parameter selection as a means to guarantee the QoS of HRT connections. Leaky buckets provide simple and user-programmable means of traffic regulation. We design and analyze an efficient optimal algorithm for selecting the burst parameters of leaky buckets to meet connections' deadlines.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call