Research Of Secure Network Coding Against Pollution Attacks Based On Robust Network

Network coding is a new information exchange technology combined routing andcoding, which allows the network information flow approach the upper bound fixed by themax-flow min-cut theorem. The encoding operation improves the throughput of thenetwork while also makes network coding is very vulnerable to pollution attacks. Anymaliciously modified data fused with other data by encoding, which generated pollutiondata and distributed to all downlinks. If the downstream node receives the pollution dataand use it for coding directly without detecting will result in a new round of pollution,eventually cause the entire system to a standstill.In order to solve the problem of pollution attacks in network coding, firstly, wepropose to build robust network to enhance the anti-pollution attack capability for networkcoding. Improve the success rate of information transmission in the face of node failuresand pollution attack through proper redundancy design. Based on the robust networkmodels already existing, the robust network judgment method is given and the scheme ofrobust network for network coding is proposed. Secondly, we study schemes of pollutionattacks detecting in network coding, discuss the pollution attacks detecting schemes baseon linear space signature and information theory, improve the pollution attacks detectionalgorithm based on information theory, analysis the computational overhead of twoschemes in different file sizes and data block size case, propose a combination schemeagainst pollution attack.Theory and simulation results show the effectiveness of the robust network. Byanalyzing the computational overhead of the combination scheme show that the schemeimproves the performance to some extent.