Research On Security Denfense Technology Of WSNs In Advanced Metering Infrastructure

Smart Grid is a new energy management paradigm. The development of smart grid has been included in the grand strategies of our country. Advanced Metering Infrastructure(AMI) is the first step to implement smart grid. The most crucial character of AMI is its two-way data exchange between user and supplier via secure and reliable network and thus an effective implementation of an interactive consumption of electricity can be realised. The low-cost WSNs is a commonly used technology in the wireless communication of AMI. Given the important position of AMI in a smart grid, the guarantee of security of WSNs communication is essential. Security defense technology is a commonly applied method in WSNs communication. The passive defense techniques such as key management and encryption are the first line of defense in the information security. And the active defense techniques such as the intrusion detection is the second line of defense. Therefore, this paper conducts an in-depth study about the defense technology in respect of the WSNs key management, secure data aggregation, trust management and intrusion.A key management scheme based on the improved ECDH method is proposed. The scheme is specifically designed for the network structure and information transmission paradigm of WSNs in AMI. It solves the key establishment, distribution and update issues in WSNs’ s unicast and multicast communication. In order to solve ECDH’s problems of middleman attack and lack of data integrity verification algorithm, a distribution and updating method of the node’s public and private key which is resistant to the middle man attack is proposed. And the data integrity verification is achieved by the authentication function based on the hash algorithm. The performance analysis shows that the scheme can ensure the security of the communication key, information’s integrity verification and resistance of attack are achieved. The scheme also has good performance in time and storage consumption.In order to reduce the resource consumption in data transmission of the smart meters in AMI and ensure the security of data transmission, a symmetric homomorphic encryption based concealed data aggregation scheme(HE CDA) is proposed. This scheme can ensure the confidentiality of data aggregation by additively homomorphic encryption algorithm and realize the integrity verification in data transmission by the Chinese Remainder Theorem(CRT). In order to obtain the data of each user’s intelligent meter so that the generation of customer billing, power quality analysis and optimization and other functions of the AMI can be achieved, a data recoverable concealed aggregation scheme(ERCDA) is proposed. Firstly, the plaintext data is encoded using a coding method in the scheme. Then the additively homomorphic encryption algorithm is used to encrypt the encoded data. And in the final decryption all the plaintext are reconstructed. The performance analysis shows that the two schemes have better data confidentiality, integrity and freshness protection. And both the computing cost and communication cost are improved.An effective trust management scheme is proposed in order to show the network state of each node in AMI. The trust value computing of each node is based on optimized Beta distribution theory. Firstly, the trust property of typical network attacks in WSNs and its main characteristics are considered to be defined. Then the trust attributes are collected by using an adaptive method, and the singularity of trust property is judged based on the Mahalanobis distance. Finally, punitive measure is used when the trust property continuous abnormal and alternating abnormal in the trust value calculation. The simulation analysis shows that the scheme can reflect the network state of working nodes more accurately than the RFSN scheme.Because of the complexity of the working environment of AMI where exists severe electromagnetic interference and other wireless communication interference, node trust values reduction cannot be attributed to interference or attack. Thus, a hierarchical intrusion detection scheme based on abnormal node’s trust value is proposed through the analysis on the node trust values in AMI with time and spatial correlation. The scheme includes two level detections, trust value abnormity and sequence abnormity detection. To detect the abnormity of value trust, an algorithm according to the transductive confidence machines and K nearest neighbor algorithm(TCM-KNN) is proposed. The simulation analysis shows that, the scheme has better detection performance(including higher detection rate and lower false-positive rate).