Comparison of capacity efficiency of DCS network restoration routing techniques

Abstract

A crucial issue in the management of telecommunications networks is restoration after a network failure. We compare the capacity efficiency (the amount of traffic restored over fixed capacity) of several types of restoration methods for Digital Crossconnect Systems (DCSs) under both node or link failures in metropolitan networks. These restoration methods can be broadly categorized based on the type of control of the restoration process and the type of procedure for rerouting traffic around the failure. It is easily shown that point-to-point methods (where traffic is rerouted end-to-end) restore a higher percent of traffic than patch methods (where traffic is rerouted only around the failed segment). However, quantification of the difference in efficiency among the methods is important to fully evaluate other trade-offs, such as cost and speed of restoration. To evaluate the efficiency difference, we generated random network traffic distributions by simulation and then failed links and nodes for each simulation sample. Statistics were generated on the expected amount of restored traffic for each restoration method. In particular, we found that the efficiency advantage of point-to-point methods over patch methods is greater for link failure than for node failure. Also, the difference was not statistically significant in low levels of network congestion (i.e., networks with large amounts of excess capacity) and became more significant as network congestion increased.