Analysis Of The Secrecy Performance Of Broadcast Channels With An Eavesdropper

The emergence of wireless communication enriches means of communication between people. However, the broadcast nature of wireless networks makes communication security very difficult to be effectively protected. Any user can receive the signals sent by the sender, accordingly, the data is very likely to be eavesdropped by the wire-tapper. Therefore, the study on confidentiality and security of communication will become more and more significant. This paper analyzes the issue of secure communication from the perspective of the information-theoretical security. Compared with the traditional encrypted methods based on the computational complexity, information-theoretical security can withstand the wire-tapper with unlimited computing power. Thus, it is widely considered as the most strict security concept.This paper focuses on the study of the achievable rate regions on broadcast channels. For broadcast channels without the wire-tapper, scholars have studied and achieved large amounts of meaningful results. For broadcast channels with the eavesdropper, an inner bound has been just obtained, while the outer bound is still not achieved. Based on the summarization of existing conclusions, the paper emphatically studies information-theoretical security issues of broadcast channels with an eavesdropper. First of all, the paper analyzes the achievable secrecy rate regions on broadcast channels with an eavesdropper, then compares them with the existed results. Secondly, according to the definition on the achievable secrecy rates and the equivocation rates, the paper proposes and proves an outer bound on the achievable secrecy rate regions of broadcast channels with an eavesdropper. And the outer bound is shown to be tighter than the cut-set bound based on max-flow min-cut theorem, which means that the outer bound makes sense. Then the comparisons between this outer bound and the existed conclusions are performed, which show that the outer bound is tighter than the existed conclusions without considering the wire-tapper. Finally, this paper presents an issue to be further studied, i.e. people can study the security of broadcast channels with an eavesdropper when the two legitimate users are cooperative.