Research On Secrecy Capacity Of Inhomogeneous Wireless Networks And Secure Transmission Strategy

Ad hoc network is a distributed and self-organized network.Nodes communicate with each other over wireless channels.The transmission is performed using multi-hop mode with intermediate nodes as relays.It is widely deployed in military communications,disaster emergency communications and sensor networks.Compared with traditional cellular networks,the circumstance of transmission in Ad hoc network is much more complicated.In the cellular network,the capacity of tranmssion can be calculated by Shannon capacity formula,while in the Ad hoc network,the calculation of an exactly formula of tranmsssion rate is very difficulty.In 2000,an approach based on scaling law was proposed to analyze the capacity of large scale wireless Ad hoc network.The scaling law reflects the relation between rate and the number of node,when the number of node goes to infinity.Due to the variety of network model and the feasible of computation,most of studies assumed the node was uniformly and statically distributed in the network.However,in the realistic network deployment,node distribution often demonstrates inhomogeneity and forms into clusters.In this statical inhomogeneous network,due to the inhomogeneity of node distribution,the traditional derivation of uniform network capacity is no longer valid.Additionally,due to the broadcast characteristics of the wireless channel,the transmission of information can be easily intercepted by the eavesdropper,which results in insecure information transmission.Aiming at solving these problems,this dissertation,taking secure information transmission and inhomogeneous network distribution into consideration,investigates the network capacity while guaranteeing the information security,which we call secrecy capacity or secrecy throughput.The main contents of this dissertation are as follows.Firstly,we proposed an artificial noise strategy to guarantee the secure transmission of information.The strategy suppresses the channel of eavesdropper by using extra energy to generate artificial noise,guaranteeing that the channel of legitimate node is better than that of the eavesdropper.In this way,we can achieve secure transmission of information.For the cluster structure of inhomogeneous network model,the information transmission is performed in two steps: intra-cluster transmission and inter-cluster transmission.For the intra-cluster transmission,we proposed a circular percolation model in each cluster based on percolation theory.Based on this model,we proved that each cluster can build a certain number of information highways.While in the case of inter-cluster transmission,we constructed a highway system similar to that of the uniform network according to the minimum density of network.Combined with the proposed artificial noise strategy,we derived the secrecy throughput of intra-cluster transmission and inter-cluster transmission,respectively.The results of asymptotic analysis showed that the secrecy throughput of intra-cluster transmission is higher than that of the inter-cluster transmission.Compared with the secrecy capacity of uniform network,we found that it is lower than that of intra-cluster transmission,but greater than that of the inter-cluster transmission.Secondly,to avoid energy consumption of artificial noise generation,we proposed a secrecy zone strategy to guarantee the secure transmission of information.According to the finite independent set,if the density of eavesdropper is less than a certain threshold,we can also use the percolation theory to build highway system for intra-cluster and inter-cluster transmission in inhomogeneous networks.Given a routing scheme of transmission,we derived the secrecy throughput of intra-cluster and inter-cluster transmission,as well as their corresponding tolerable density of eavesdropper.The results of asymptotic analysis showed that,comparing the artificial noise strategy with the secrecy zone strategy,we found that secrecy zone strategies can achieve the same order of secrecy throughput capacity.However,the tolerable number of eavesdropper is decreased.Finally,in the derivation of the securecy throughput,we only give a simple routing scheme,and do not considered the resource allocation of network.To strengthen this part,under the cases of infinite coding block and finite coding block,we proposed a control algorithm for secrecy transmission in multi-hop wireless networks,respectively.Such that the information secure transmission is guaranteed.The algorithm aims to maximize the utility function of network security throughput.Using the theory of stochastic network optimization,the network control algorithm can be simply decomposed into sub-algorithms: flow control,end-to-end coding,scheduling and resource allocation.Using Lyapunov drift and penalty analysis,we proved that the proposed control algorithm can arbitrarily approach the optimal value of the objective function.However,as the objective function approaches the optimal value,the transmission delay is also increased.The research of secrecy transmission in inhomogeneous wireless network not only complements the research of network capacity problem,but also provides insightful guidelines for the planning and deployment of wireless network.