A decentralized downlink beamforming algorithm for multicell processing

Le, Tuan Anh and Nakhai, Mohammad Reza (2012) A decentralized downlink beamforming algorithm for multicell processing. IEEE Global Communications Conference (GLOBECOM) . pp. 3924-3929. ISSN 1930-529X (Accepted/In press) (doi:10.1109/GLOCOM.2012.6503729)

Full text is not in this repository.

Abstract

In this paper, we tackle the problem of intercell-interference in the absence of cooperation amongst base stations (BSs). The problem originates from the simultaneous transmissions of several BSs to their local users within their corresponding cells using the same frequency band. We propose a decentralized downlink beamforming scheme in cellular interference channels so that individual BSs can independently design their own beamforming vectors based on the locally attainable second order statistical channel state information at each BS. In this scheme, while each BS minimizes a combination of its total transmit power and the resulting total interference on the other vulnerable users of the adjacent cells, it assures that certain targeted signal-to-interference-plus-noise ratio (SINR) levels are achieved at its local users. Using the reciprocity of wireless network and the uplink-downlink duality, we develop a fast-converging iterative algorithm for downlink transmissions by distributed BSs. Finally, we derive a feasibility condition for the existence of a unique solution to the proposed optimization problem. This condition can be used to develop a scheduling algorithm for choosing a set of feasible users within each cell. Simulation results confirm the effectiveness of the proposed decentralized algorithm in closely following the performance curve of its centralized counterpart.

Item Type: Article
Research Areas: A. > School of Science and Technology > Computer and Communications Engineering
Item ID: 16895
Useful Links:
Depositing User: Tuan Le
Date Deposited: 08 Jun 2015 08:47
Last Modified: 13 Oct 2016 14:35
URI: https://eprints.mdx.ac.uk/id/eprint/16895

Actions (login required)

Edit Item Edit Item