无线传感器网络中不依赖MAC认证的虚假数据过滤算法

False data filtering algorithm without relying on MAC authentication for wireless sensor networks

依据传感器网络分簇协作聚合的特点,提出了一种多重半脆弱水印算法,水印的生成、嵌入和提取是由簇内选取的多个证人节点共同完成,具有节约能量和抵御单个节点被俘获而泄密的特点。提出了一种基于多重半脆弱水印的虚假数据过滤算法,利用水印对数据包进行相关认证,无需依赖MAC认证。分析和仿真表明,新的算法不仅过滤虚假数据的能力强、通信开销低,且具有"顽健但脆弱"的特性,能支持网内有损数据处理和抵抗一定的噪声干扰。

A multiple semi-fragile watermarks algorithms was proposed, considering the characteristics of clustering collaboration aggregation in WSN. The generation of watermarks, embeding and extraction were completed by not a single node, but a fixed number of witnesses nodes which selected within the same clusters. This algorithm could against leakage of keys when a single node was captured, and it could save energy costs also. On this basis, a false data filtering algorithm was proposed based on multiple semi-fragile watermarks, this algorithm could authenticate the accuracy, completeness and fresh using watermarks embedded in data, without relying on MAC authentication. It provided a comparison between the scheme and previously proposed schemes. The analysis and simulation results indicate significant improvement in probability of filtering false data, energy saving and having robust, yet fragile characteristics, which support lossy in-network data processing and resist of the noise in certain.