TY - JOUR
T1 - Confidential broadcasting via linear precoding in non-homogeneous MIMO multiuser networks
AU - Yang, Nan
AU - Geraci, Giovanni
AU - Yuan, Jinhong
AU - Malaney, Robert
PY - 2014/7
Y1 - 2014/7
N2 - We propose linear precoding with power control to achieve confidential broadcasting in multi-input-multi-output multiuser networks such that the base station (BS) with Nt antennas securely broadcasts messages to K users with Nr antennas each. We focus on the practical non-homogeneous scenario where the distances between the BS and the users are not equal. We first design a linear precoder based on regularized channel inversion, and derive new channel-independent expressions for the achievable secrecy sum-rate in the large system regime. With the aid of these expressions, we examine the impact of user dispersion, Nt, and K on the secrecy sum-rate. We then propose a power reduction strategy and power allocation algorithms to increase the secrecy sum-rate. We demonstrate that our power reduction strategy increases the secrecy sum-rate at high signal-to-noise ratios. We also show the secrecy sum-rate advantage of optimal power allocation over equal power allocation. Furthermore, we consider channel correlation and derive an easy-to-compute expression for the secrecy sum-rate to examine its impact on the secrecy performance.
AB - We propose linear precoding with power control to achieve confidential broadcasting in multi-input-multi-output multiuser networks such that the base station (BS) with Nt antennas securely broadcasts messages to K users with Nr antennas each. We focus on the practical non-homogeneous scenario where the distances between the BS and the users are not equal. We first design a linear precoder based on regularized channel inversion, and derive new channel-independent expressions for the achievable secrecy sum-rate in the large system regime. With the aid of these expressions, we examine the impact of user dispersion, Nt, and K on the secrecy sum-rate. We then propose a power reduction strategy and power allocation algorithms to increase the secrecy sum-rate. We demonstrate that our power reduction strategy increases the secrecy sum-rate at high signal-to-noise ratios. We also show the secrecy sum-rate advantage of optimal power allocation over equal power allocation. Furthermore, we consider channel correlation and derive an easy-to-compute expression for the secrecy sum-rate to examine its impact on the secrecy performance.
KW - Multi-input-multi-output
KW - multiuser networks
KW - physical layer security
KW - regularized channel inversion
KW - secrecy sum-rate
UR - http://www.scopus.com/inward/record.url?scp=84905157998&partnerID=8YFLogxK
U2 - 10.1109/TCOMM.2014.2329471
DO - 10.1109/TCOMM.2014.2329471
M3 - Article
SN - 1558-0857
VL - 62
SP - 2515
EP - 2530
JO - IEEE Transactions on Communications
JF - IEEE Transactions on Communications
IS - 7
M1 - 6827200
ER -