NETWORK PROGRAMMING METHODS FOR LOSS NETWORKS

This paper describes how some of the insights available from the stoch astic analysis of dynamic routing may be incorporated into the classic al mathematical programming approach to the design of networks, In par ticular, we present the results of a number of numerical investigation s into network architectures for circuit-switched communication networ ks. Our investigations use recent theoretical results integrating netw ork dow optimization and Markov decision processes to provide performa nce bounds for dynamic routing strategies. Following a tutorial introd uction of the above mentioned topics we develop a sequence of network examples, Our first examples are familiar ones, such as symmetric full y connected networks and networks with moderate amounts of asymmetry, and we describe how network programming methods complement earlier wor k on dynamic routing. We then consider a variety of example networks w hich have a more sparse collection of links. These examples indicate t he potential applicability of the methods to a variety of areas, inclu ding studies of the design, performance and resilience of future commu nication networks.