The Study Of Trust Management System In Wireless Sensor Networks

Wireless Sensor Network (Wireless Sensor Network, WSN) is a new kind of Network. It has been used in many environments, such as, military, environment, traffic maintenance, health care, and many other areas. With the development of the network, its security problem is becoming more and more important. WSN node is faced with malicious packet discarding, eavesdropping threats, so an effective security tools is necessary. The Trust Management System is put forward under this background. It is able to effectively respond to attacks and ensure the security of network operation. However, due to the limit of hardware and software, complex calculation cannot be processed in wireless sensors. So traditional trust management systems cannot be adopted in wireless sensor networks. For example, public-key cryptographic, authentication and other mechanisms cannot detect the node physical capture, the malicious or selfish nodes. They are insufficient for secure routing of message from source to base station in WSN. So, in the wireless sensor network, many researches are trying to design and develop a light weight trust management system, which takes a little resources of the node in calculation and management of trust between/among the nodes. The trust management of the WSN should be as simple as possible, detect the different attacks easily, and mange&update trust values.In this paper, we present a novel multivalued trust level routing protocol, called MTR (Multivalued Trust level Routing), which can find malicious nodes and improve the ability of resisting malicious sensor nodes attack. The algorithm is mainly composed of two parts:trust evaluation and fuzzy routing selection. In trust evaluation part, Sponsor node calculates trust value based on the interacting record of target node. In fuzzy routing selection, Sponsor node employ Fuzzy Mathematic Theory to select the routing node based on the information of trust value and topological structure. By this means, it can guarantee the security of the message sending to Base Station and avoid the energy consume by using flood broadcast to find route. At the end, we used OMNet to stimulate MTR protocol, and analysis the stimulation result.In this paper, we extend our previous work and give a new protocol, which is called RDSC (a reliable data aggregation protocol based on cross sensing in wireless sensor network). Compared with MTR, mainly used in the WSN which does not contain data aggregation, RDSC is suit to the WSN which is based on cluster and have data aggregation. In this protocol, cluster head node and normal node (contains in the same cluster with cluster head node) can evaluate trust, find malicious nodes, and guarantee the security of the network. RDSC mainly contain two parts:cluster head node evaluate the trust of normal node, normal node evaluate the trust of cluster head node. When cluster head node evaluate the trust of normal node, in a cluster, cluster head node based on the cross sensing, analysis the degree of association in different normal nodes, then give the trust evaluating of normal node, find malicious nodes. When the normal node evaluate the trust of cluster head node, if the normal node detect that the aggregation data did not contain the data of this normal node’s data, then it will interact this information with other normal node, if there are many nodes (e.g.30%of total number of nodes in cluster), normal node can marked the cluster head node with unusual cluster head nodes. At the end, we used OMNet to stimulate MTR protocol, and analysis the stimulation result.