Research On Trust Evaluation Method Based On Cloud Model Under Clustered Wireless Sensor Networks

Wireless Sensor Networks(WSNs)is an important part of Internet of things and has been widely applied to various fields.However,the inherent characteristics of sensor nodes constrained by price and volume limit its communication,computation,storage,and battery power,and due to its open working environment of unmanned surveillance or hostile areas,WSNs is prone to sudden accidents and suffers from attacks of malicious nodes,which lead to networks damaged.The encryption-decryption techniques devised for the traditional networks have not been adapted to wireless sensor networks because of their complex algorithm and computational overhead;In addition,they are mainly used for resisting the external attack,but attacks raised from internal malicious nodes are hard to be defended.In order to save resources and solve internal attacks of WSNs,a cloud model trust evaluation method based on clustered structure of WSNs is designed on the basis of researching and analyzing the existing methods that aim at the problems of limited resources and internal network security.The research results achieved in this paper are as follows:(1)An energy-efficient and balanced unequal clustering nodes scheduling algorithm is proposed.In the cluster formation stage,the candidate cluster heads form heterogeneous competition scopes and builds different sizes according to many factors,such as the distance between candidate cluster head and the sink node,the regional density of node,as well as the residual energy of node,which can balance the energy consumption between clusters.At the same time,after clusters formation,finding redundant nodes in the cluster by judging intersection angle between central node and its perceptive nodes,and scheduling its working status.It can decrease the number of working nodes in the network and improve the energy utilization.(2)Considering wireless sensor networks is vulnerable to internal attacks,a trust evaluation method based on cloud model under the background of clustering structure of WSNs is proposed,which effectively solves the uncertainty of trust relationship and improves the accuracy of trust evaluation.Firstly,the algorithm considers multi-factors including communication factor,message factor and energy factor.Building mathematical model for each trust factor and combining with cloud generator to get factor trust clouds.Secondly,immediate trust cloud is calculated by assigning adjustive weights for each factor trust cloud and combining them in order to achieve better trust evaluation.Furthermore,recommendation trust cloud and immediate trust cloud are synthesized according to revised time sensitive factor to get final trust cloud.In addition,the final trust cloud of sensor node is converted to trust grade by trust cloud decision-making,which can accommodate the fuzziness and randomness of trust and ensure the security of networks.(3)Carrying out simulation experiments to the proposed methods on network simulation platform OMNeT++.Various validation and contrast experiments show that the proposed unequal clustering nodes scheduling algorithm can save and balance energy consumption effectively and prolong the network lifetime.At the same time,the trust evaluation method under the clustering structure has accuracy and comprehensiveness in the aspect of evaluating sensor node's trust.It not only can identify malicious nodes faster and resist various attacks,but also has higher tolerance to abnormal environments.