摘要：AbstractIn this paper a distributed Received Signal Strength (RSS) minimization algorithm is proposed that guarantees strong connectivity of the network topology while minimizing the RSS of the network received at a given eavesdropper’s location. The proposed algorithm is composed of multiple rounds of maximum consensus network communications implementing a distributed greedy solution of the problem. The proposed RSS minimization algorithm is distributed in the sense that nodes do not assume, estimate or communicate any network connectivity knowledge such as a routing table, the Laplacian matrix, neighbour lists or the total number of nodes. The proposed algorithm assumes that the initial network topology is strongly connected and that each agent knows its own location and that of the eavesdropper. We provide an extension of the proposed algorithm for dealing with multiple and moving eavesdropper. In this case, we also propose a heuristic for increasing nodal transmit power to effectively reshape the network topology according to the closer eavesdropper. Performance of the proposed algorithm is demonstrated in simulations.
关键词：KeywordsMulti-agent systemsDistributed optimisation for large-scale systemsSensor networks
