Ranking node influence according to dynamic network load

Abstract

When a mobile network suffering attacks, ranking and then protecting influential nodes to ensure the security of the network are critical. Many methods have been proposed to evaluate nodes influences. However, most of such methods are based on network topology analysis through graph theory only. As a result, the most influential node chosen by such methods may be a rarely used one for packets transmissions. We devise NodeRank, Direct Principal Component Ranking (DPCR) and Comprehensive Principal Component Ranking (CPCR) algorithms to rank node influence according to dynamic network load changes, where NodeRank is a single-dimensional method with higher efficiency; DPCR and CPCR are multi-dimensional methods with higher accuracy. DPCR applies the rule that the first principle component takes all in the measurement; CPCR applies a moderate rule that uses linear-weight comprehensive evaluation which takes the importance from each principal component in the measurement. Simulation results show that our proposed schemes outperform the known methods such as Degree Centrality (DC), Betweenness Centrality (BC), Closeness Centrality (CC) and Eigenvalue Centrality (EC) for node ranking. To the best of our knowledge, DPCR and CPCR are the first known schemes that can rank node influence with respect to the direction of packets transmissions in networks.