论文部分内容阅读
物联网中攻击者由于知道和用户相同的安全测试信息,能够得知哪些产消者更受用户青睐并集中攻击。提出了一种新的非协作、双玩家的决策支持方法—产消者选择博弈(Prosumers Selection Game,PSG),PSG采用纳什均衡确定了最优化用户报酬的产消者子集。该方法使用纳什玩家选择(Nash Prosumer Selection,NPS)得到产消者子集概率向量的博弈解,并且用户将面临最小期望损害。此外,用户使用NPS选择产消者,依然有非0概率选择出最不安全的产消者。最后将PSG方法中的NPS和两种不同的启发式选择算法进行仿真比较,结果表明NPS在降低安全风险方面能够获得约38%的性能提升。
In the Internet of things, attackers know the same security testing information as the users, and can know which consumers are more favored by users and concentrate on attacks. A new non-cooperative and dual-player decision support method called Prosumers Selection Game (PSG) is proposed. PSG uses Nash equilibrium to determine the subset of consumers who optimize user compensation. The method uses Nash Prosumer Selection (NPS) to obtain the game solution of the pro-consumer subset probability vector, and the user will face minimum expected damage. In addition, users use NPS to select consumers and still have a non-zero probability of choosing the most unsafe consumers. Finally, the NPS in the PSG method is compared with two different heuristic selection algorithms. The results show that NPS can achieve about 38% performance improvement in reducing the security risk.