论文部分内容阅读
Physical layer(PHY) security is recently proved to enable improving the security of wireless communication networks. In downlink frequency division duplex(FDD) cloud radio access network(C-RAN), the performance of PHY security highly relies on the channel state information(CSI) which is usually acquired through the codebook-quantization-based technique at the transceiver. However, the conventional quantization method aggravates the leakage of privacy information in C-RAN under the eavesdropping environment. In this paper, a novel channel quantization method is investigated to improve the secrecy-rate performance of C-RAN by exploiting the high-dimension space geometry. Based on this method, it is proved that when the statistical distribution of the channel matrices of both the legitimate user and the eavesdropper is exploited, a win-win situation can be created where secrecy-rate gains are improved without sacrificing beamforming gains from the point of view of ergodic rate. Particularly, a secrecy-oriented criterion is devised to implement the proposed method for generating codebooks. Then a weighted Voronoi diagram(WVD) is formulated on the complex Grassmann manifold and finally, a vector quantization based algorithm is proposed to build up novel quantization codebooks iteratively. Simulation results further validate the superiority of our proposed codebooks over conventional codebooks in C-RAN systems.
Physical layer (PHY) security is recently proved to enable improving the security of wireless communication networks. In downlink frequency division duplex (FDD) cloud radio access network (C-RAN), the performance of PHY security highly relies on the channel state information CSI) which is usually acquired through the codebook-quantization-based technique at the transceiver. However, the conventional quantization method aggravates the leakage of privacy information in C-RAN under the eavesdropping environment. In this paper, a novel channel quantization method is to improve the secrecy-rate performance of C-RAN by exploiting the high-dimension space geometry. Based on this method, it is proved that when the statistical distribution of the channel matrices of both the legitimate user and the eavesdropper is exploited, a win -win situation can be created where secrecy-rate gains are improved without sacrificing beamforming gains from the point of view of ergodic rate. ly, a secrecy-oriented criterion is devised to implement the proposed method for generating codebooks. Then a weighted Voronoi diagram (WVD) is formulated on the complex Grassmann manifold and finally, a vector quantization based algorithm is proposed to build up novel data . Simulation results even validate the superiority of our proposed codebooks over conventional codebooks in C-RAN systems.