Combating Unknown Eavesdroppers by using Multipath Wireless Receptions

Unlike many existing studies on physical layer security which use multi-antenna techniques to improve the transmission security, this paper exploits the multipath wireless channel to achieve secure transmission. Specifically, we consider a time-reversal transmission system where the signal waveform is designed as the conjugated time reversed counterpart of the wireless multipath channel. Thus, the wireless channel acts as a matched filter that can boost the signal power at the intended receiver. To further improve the transmission security, we also inject a time-domain artificial noise which causes no interference to the intended receiver's signal detection. As for the eavesdroppers, we assume that their specific number and locations are unknown and use the homogenous Poisson point process (HPPP) to model the distribution of them. First, we study the average achievable rate of the intended user and a given eavesdropper, respectively. Then, based on the HPPP model, we can obtain the secrecy outage probability under a given density of eavesdropper. With a given requirement on the secrecy outage probability, we can finally obtain the optimal energy allocation parameter and the corresponding maximum secure transmission rate. Numerical results are presented to show that the proposed scheme can guarantee the transmission security under various system parameters.