In this paper, we consider coordinated downlink transmission in a cellular system wherein each base station (BS) has multiple geographically dispersed antenna ports. Each port uses a fixed transmit power and the goal of the BSs is to collectively determine the subset of ports and the corresponding beam steering coefficients that maximize the minimum signal-to-interference-plus- noise ratio observed by the user terminals. This problem is NP-hard. To circumvent this difficulty, a two-stage polynomial-complexity technique that relies on semidefinite relaxation and Gaussian randomization is developed. It is shown that, for the considered scenarios, the port state vectors and beam steering coefficients generated by the proposed technique yield a performance comparable to that yielded by exhaustive search, but with a significantly less computational complexity. It is also shown that the proposed technique results in significant power savings when compared with other transmission strategies proposed in the literature.

Additional Metadata
Keywords beam steering optimization, Distributed antenna systems, Gaussian randomization, multi-cell coordination, port selection, remote radio heads, semidefinite relaxation
Persistent URL dx.doi.org/10.1109/TWC.2012.030512.111256
Journal IEEE Transactions on Wireless Communications
Citation
Ahmad, T. (Talha), Gohary, R, Yanikömeroǧlu, H, Al-Ahmadi, S. (Saad), & Boudreau, G. (Gary). (2012). Coordinated port selection and beam steering optimization in a multi-cell distributed antenna system using semidefinite relaxation. IEEE Transactions on Wireless Communications, 11(5), 1861–1871. doi:10.1109/TWC.2012.030512.111256