Research Of Data Security In Wireless Sensor Networks Based On Dual Digital Watermarking

Wireless sensor networks have been widely applied in industry, agriculture, military field and environmental monitoring. Its data security are becoming prominent because of the WSNs characteristics. Most of existing solutions don’t take WSNs characteristics into consideration and can hardly be used in real networks. Digital watermarking is suitable for protecting data security in WSNs because of its concealment, robustness and safety. However, most of existing watermarking methods for data security protection in wireless sensor networks do not give sufficient weight to the various conditions of the real networks. Firstly, among existing watermarking methods, the problem of packet loss has not been considered. Most of methods treat packet loss as anomaly attacks. Secondly, these methods do not have great anti-attack capability because of the simplistic watermarking. To solve these problems, the main contributions in this paper is listed as follows:(1) This paper proposes a packet loss tolerated method for data integrity protection in wireless sensor networks using dual digital watermarking. Sensory data collected by nodes generate digital watermarking W1 and W2 for integrity detection. Watermarking will be stored in sensory data. At the base station side, a watermarking algorithm is designed to extract the dual digital watermarking information. W1 is tamper detection watermarking to ensure that data have not been tampered or forged in transit, and W2 is packet loss detection watermarking to ensure that the network has the prevention of replay attack and deletion attack. This plan mainly considers a packet loss tolerated method for data integrity protection in wireless sensor networks based on dual digital watermarking.(2) This paper proposes a packet loss tolerated method for data copyright protection in wireless sensor networks using dual digital watermarking. An intact watermarking can hardly be stored in a single packet because of the limited embedding space. So, an intact watermarking is split into multiple parts and be embedded in multiple packages. In a complete embedding process, each part of copyright watermarking is embedded into multiply positions in different packages. At the base station side, a watermarking algorithm is designed to extract the watermarking information. In order to prevent non-intact extracted watermarking, which may be caused by packet loss, packet loss detection watermarking is embedded into every package. This paper uses packet loss detection watermarking to analyze package loss situation, and finds missing parts of copyright watermarking from successful transaction packages. The length of the watermarking is no longer limited by embedding space. These methods can effectively protect copyright and authentication source even if package loss rate is biggish.