Virtual Path Layout Design on ATM Networks 1

The ATM standard speciies two types of connections-Virtual Paths (VPs) and Virtual Channels (VCs). A VP consists of an aggregated grouping of VCs. A VC consists of a concatenation of VPs. The VP concept was motivated by the desire to reduce the setup and switching costs of VCs, which are established on demand (or per call), by using existing VPs. In an ATM network, VC establishment is based on the layout of existing VPs. Thus it is important that VPs are constructed such that they can support the eecient setup of connection requests (VCs) among all network user pairs. Ideally, the most desirable VP-layout is a VP between each source-destination pair. However this is not feasible due to the limit in the number of incoming VPs each switch can support, and the high management overhead associated with supporting a large number of VPs. Other constraints on the VP-layout problem include the physical topology of the network (which include the physical link capacities), and the delay bounds between pairs of end user nodes (which is a traac class speciic constraint). This paper examines the eecient layout of VPs in an ATM network. The ATM network consists of ATM switches and their attached network end users, which may be gateways, routers, and hosts. The physical topology, the ooered traac, and call setup matrices of the network end users are assummed to be given. The problem is formulated as a ow-based optimization problem. A heuristic approach is presented which (i) establishes VPs according to physical network-speciic and application-speciic constraints and a cost function, (ii) provides multipaths between each source-destination user pair to minimize the call blocking probability and to increase network resilience, and (iii) uses a novel VP combining process which is guaranteed to always satisfy the switching constraints. Simulation results are presented for the proposed VP planning policy. Guidelines for the design of robust VP layouts and the eecient establishment of VCs are also presented. (ii) a grant from Bellcore. 2 This study was undertaken while Dr. Ahn was a PhD candidate at the University of Minnesota.