No abstract

No abstract

Several self-healing protocols utilizing virtual paths have been proposed in the relevant literature. Those which work in a mesh topology function according to three main rerouting strategies (though specific flooding administrations differ): local rerouting, source-destination rerouting, and local-destination rerouting. Most performance studies of self-healing protocols have considered restoration time as sole performance metric. This would have to be within the 2s threshold in order to guarantee service continuity. This one-sided metric needs to be completed. In this paper, I propose an extended performance(or goodness) metric framework in order to catch more performance aspects. In addition, this analysis uses survivability functions to measure the performance of rerouting strategies.

Since high availability requirements are expected to increase in the B-ISDN era, adequate methods are needed, which allow network designers and operators to outline the relation between available network resources and actual traffic demands on one hand and the reachable availability (performance index) for either the whole network or individual virtual path connections on the other hand. This paper suggests a simple and thus fast method for deriving the symbolic expression of the performance index in a compact form, from which a "capacity related reliability polynom" may easily he derived. Since the capacity of several subnetworks must be computed taking re-routing capability into account, an efficient procedure for capacity determination is also suggested. An example illustrates the procedure and particularly points out the effect of both resource sharing and VP bandwidth control schemes on the performance index.

No abstract