IDS Based On Multi-objects Game Theory

Security has become the most important issue in wireless sensor networks (WSNs), because of the open-access of wireless link, the limited survival circle, dynamic topology changing. The traditional security strategy based on firewall takes a little of effects because its delay on probing the attack. While the IDS, as a initiative way, could protect the certain nodes before attack and minimize the effect the attack caused.This article emphasizes on predicting which nodes will be attacked. It combines the game theory and traditional IDS, which views the network's defender and attacker as two"rational"gamblers. They would pay certain cost and get the behalf, no matter their attacking or defending. In this way, they would gather to Nash Equilibrium finally, which makes the predict feasible. The nodes Nash Equilibrium involves is just the targets the IDS is going to protect.Whether to attack a node or not is determined by many factors. To value only one factor is unilateral. This article builds up a multi-objects game model, which combines factors together and introduces four solutions(Pareto-Nash Equilibrium,Maximum,Minimum,Weight Measure) to resolving the multi Nash Equilibrium issue. At the end, a simulation experiment on the mobile-network simulation platform– GloMoSim is made to analyze the four solution's advantages,disadvantages and the scenes they fits.