Trusted Network Coding In Green Wireless Networks

Network Coding has been proved to be an important theory to improve the networkperformance, which can meet the high energy efficiency demand of Green WirelessNetworks. However, the potential misbehaved nodes in Green Wireless Networks bringchallenges for the security and efficiency of network coding. This thesis focuses on how toavoid the influence of the misbehaved nodes and ensure the performance of networkcoding.The thesis first analyzes the misbehaviors-spreading phenomenon and efficiencybottleneck problem of network coding in wireless networks and proposes to use trustscheme to solve these problems. Trust scheme is a distributed network security mechanismwhich is an effective supplementary for traditional security mechanisms. In trust scheme,network entities make deputy judgment for neighbor nodes according to the observation oftheir behaviors and status, and provide reference for future interactions. The thesis takesconsideration of the reliability of coding and decoding procedures as the trust propertiesand proposes a coding aware trust model. The coding aware model considers three trustproperties: transmission trust, encoding trust and decoding trust. In the model, nodes’synthetical trust vector is generated by merging the direct trust and indirectrecommendation trust, which provides reference for trusted network coding.This thesis proposes Trusted Coding Aware Routing (TCAR) protocol based on theproposed coding aware trust model. TCAR protocol makes routing selections according totransmission trust and encoding trust and chooses highly trusted routes to transmit data,which ensures the high reliability of encoding nodes and creates more codingopportunities. TCAR protocol contains a trusted coding algorithm “TrustCode” whichconsiders the decoding trust of neighbor nodes as the reference of making codingdecisions, which enhances the coding efficiency of network coding.This thesis made simulations of TCAR and compared with a typical network codingprotocol COPE. Simulation results showed that TCAR successfully avoided the influenceof misbehaved nodes, and make large improvement of performance compared with COPE.