Security-Reliability Tradeoff Analysis of Artificial Noise Aided Two-Way Opportunistic Relay Selection

In this paper, we investigate the physical-layer security of cooperative communications relying on multiple two-way relays using the decode-and-forward (DF) protocol in the presence of an eavesdropper, where the eavesdropper appears to tap the transmissions of both the source and of the relay. The design tradeoff to be resolved is that the throughput is improved by invoking two-way relaying, but the secrecy of wireless transmissions may be degraded, since the eavesdropper may overhear the signals transmitted by both the source and relay nodes. We conceive an artificial noise aided two-way opportunistic relay selection (ANaTWORS) scheme for enhancing the security of the pair of source nodes communicating with the assistance of multiple two-way relays. Furthermore, we analyze both the outage probability and intercept probability of the proposed ANaTWORS scheme, where the security and reliability are characterized in terms of the intercept probability and the security outage probability. For comparison, we also provide the security-reliability tradeoff (SRT) analysis of both the traditional direct transmission and of the one-way relaying schemes. It is shown that the proposed ANaTWORS scheme outperforms both the conventional direct transmission, as well as the one-way relay methods in terms of its SRT. More specifically, in the low main-user-to-eavesdropper ratio (MUER) region, the proposed ANaTWORS scheme is capable of guaranteeing secure transmissions, whereas no SRT gain is achieved by conventional one-way relaying. In fact, the one-way relaying scheme may even be inferior to the traditional direct transmission scheme in terms of its SRT.

I. Introduction

Cooperative relaying has attracted substantial research interests from both the academic and industrial community, since it is capable of mitigating both the shadowing and fast-fading effects of wireless channels. There are two popular relaying protocols, namely the amplify-and-forward (AF) [1], [2] as well as the decode-and-forward (DF) [3] , [4]. In the case of AF relaying, the selected relay multiplies its received signals by a gain factor and then forward them to the destination [1] , [2]. By contrast, the DF relay decodes its received signals and then the selected relay forward its decoded signal to the destination [3], [4]. Additionally, in [5], both AF and DF relaying schemes are investigated. In general, closer to the source, DF relaying has a high probability of successful decoding and flawless retransmission from the relay to the destination from a reduced distance  [6]. By contrast, close to the destination the DF relay has just as bad reception as the destination itself, hence it often inflicts error propagation. Fortunately in the vicinity of the destination AF relying tends to outperform DF relaying [6]. Additionally, [7] also shows that adaptive DF outperforms AF in terms of its frame error rate (FER).