Robust Secure Beamforming for Wireless Powered Full-Duplex Systems With Self-Energy Recycling

• Published in: IEEE transactions on vehicular technology Vol. 66; no. 11; pp. 10055 - 10069
• Main Authors: Wu, Wei, Wang, Baoyun, Zeng, Yong, Zhang, Haiyang, Yang, Zhenxing, Deng, Zhixiang
• Format: Journal Article
• Language: English
• Published: New York IEEE 01.11.2017; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Array signal processing; Beamforming; Communication system security; Communications systems; Covariance matrices; Covariance matrix; Eavesdropping; Energy harvesting; full-duplex (FD); Optimization; physical (PHY) layer security; robust beamforming; Robustness (mathematics); Transmitting antennas; Wireless communication; Wireless communications; Wireless powered communication (WPC)
• ISSN: 0018-9545; 1939-9359
• DOI: 10.1109/TVT.2017.2744982

In this paper, we study a multiuser wireless powered communication system, where an energy-constrained full-duplex information transmitter (IT), powered by wireless energy from a dedicated energy transmitter (ET), intends to send confidential information to the information receiver (IR) in the presence of multiple idle users that could be the potential eavesdroppers. In the practical scenario of imperfect channel state information and assuming that the idle users need to harvest energy from the ET, we aim to maximize the worst-case secrecy rate at the IR by jointly optimizing the transmit covariance matrix at the ET as well as the information beamforming and artificial noise covariance at the IT, subject to their individual transmit power constraints and the minimum required power transferred to the idle users. We employ the semidefinite relaxation (SDR) and extended S-procedure approaches to transform the original nonconvex optimization problem into convex problem, which can be efficiently solved by solving a sequence of semidefinite programs. Furthermore, we show that the SDR is tight since there always exists a rank-one optimal solution. For performance comparison, two heuristic schemes for ease of implementation are also developed. Numerical results are presented to show the effectiveness of our proposed schemes.