基于分布式梯度算法的WSN隐私保护技术研究

数据隐私保护技术是WSN领域的研究热点之一,针对数据隐私保护问题提出了一种基于分布式梯度算法的密钥管理策略.把网络拓扑结构抽象为有向图,每个节点都有各自的目标函数,密钥采用异步更新方式.更新过程中,每个节点的梯度值由目标函数给出,通过分布式优化算法求得全局目标函数的最优解,以此来计算通信密钥.随机因子依据数据与梯度的差值自适应调整作动态变化,攻击者无法获取随机因子及相关参数,从而达到隐私保护的目的.论文从隐密性、收敛性、有效性3个方面验证分析了该算法的优越性.

Research on privacy preserving technology based on distributed gradient algorithm in WSN

Data privacy protection technology was one of the research hotspot issues in the field of WSN. This paper proposed a key management strategy based on distributed gradient algorithm for data privacy protection. The network topology was abstracted as a directed graph,in which each node had its own objective function,and the private key was updated in an asynchronous way. In the process of updating,the gradient value of each node was given by the objective function. And the communication key could be calculated this way in which the optimal solution of the global objective function was obtained through the distributed optimization algorithm. The purpose of privacy protec-tion was achieved because the random factors could be adjusted dynamically according to the difference between the data and the gradient so that the attacker cannot receive random factors and relevant parameters. The superiority of the proposed algorithm in this paper is verified by three aspects:privacy,convergence and validity.