New reliability model and its application to assess the performance of sensor networks

摘要

In this paper we introduce a new network reliability model that can be particularly applied to evaluate performance objectives of wireless sensor networks. In the classical reliability theory a communication network can be modeled as directed graph G - (V, E), composed of a set of nodes V, and a set of directed links E, respectively. Given that the links of the network underlying graph fail independently with known probabilities (nodes are perfectly reliable), and given a set K of terminal nodes (or participating nodes) and a terminal node s of K, the K-terminal-to-sink reliability measure, RKs(G), is the probability of the event that the surviving links span a sub-digraph of G such that for each node u of K there exists a operational directed path from u to s. In this paper we study a combinatorial property of graphs called the domination invariant which has been applied within the context of the classical reliability to efficiently compute the reliability of communication networks. Moreover we model wireless networks as random digraphs using current results in Information Theory and we discuss how the K-terminal-to-sink reliability could be applied to tackle several optimization as well as design problems in sensor networks.