Skip to main content

Distributed Transmission for Cooperative Multiuser Multiple-Input Multiple-Output (MIMO) Systems

Final Activity Report Summary - DITCOMMS (Distributed Transmission for Cooperative Multiuser Multiple-Input Multiple-Output (MIMO) Systems)

It is envisioned that next generation wireless networks will be able to provide a wide variety of services from high quality and data intensive applications such as streaming multimedia to low data rate services such as text messaging. One of the main challenges for future wireless system design is developing spectrally efficient transmission techniques that can provide high data rates with high quality of service, while utilising the limited system resources like transmission power and bandwidth as efficiently as possible. In particular, the main limiting factor for high data rate communications in wireless cellular networks is the multiple access interference from other users. In cellular networks, the service area is divided into smaller coverage areas called cells, and communication within a cell is managed by the base station located in that cell. In traditional cellular networks, multiple access interference is mitigated by frequency planning, where adjacent cells do not use the same frequency band for communication. This results in the inefficient use of the scarce system resources. Recently, multicell processing, where neighbouring base stations cooperate in the transmission / reception of data, has been attracting significant interest as a spectrally efficient method for providing high data rates by taking advantage of the spatial diversity provided by cooperative base station antennas.

In this project, cooperative transmission techniques for the downlink of a wireless cellular network are investigated, where the neighbouring base stations cooperate in the simultaneous transmission of independent data to different users. In systems with base station cooperation, one can envision the channel between the base stations and the mobile users as a multiple input multiple output (MIMO) broadcast channel, where the antennas of the virtual multiple antenna transmitter are geographically distributed. The capacity region and capacity achieving transmission technique for the MIMO broadcast channel have been derived assuming a total transmit power constraint. However, for systems with base station cooperation a total transmit power constraint is infeasible, since each base station antenna is likely to have its own power constraint. As a result, it is necessary to derive the capacity region and capacity achieving transmission techniques under individual per antenna power constraints. In this project, an iterative method is developed, that computes the optimal transmission scheme for a MIMO broadcast channel, including the optimal power allocation between the users and optimal linear precoding (downlink beamforming vectors), minimising the total transmit power under individual per antenna power constraints and individual performance targets on users. Especially, when the channels for different users are not symmetric, the proposed method is shown to have performance advantage over existing methods in the literature.

The distributed implementation of the proposed method for computing the optimal transmission scheme in a cooperative wireless cellular network is not practical, since it is a centralised algorithm in nature, requiring excessive information exchange between base stations. Therefore, it is of great practical interest to develop distributed algorithms for computing the optimal power allocation and downlink beamforming vectors that require only limited and local information exchange between base stations. By modifying the proposed centralised iterative method, a suboptimal algorithm that is fully distributed in nature is developed. Numerical results demonstrate that when the performance targets of the users are not too high, the proposed algorithm achieves near optimal performance. Consequently, one can conclude that the results obtained in this project constitute a first step in realising cooperative base station transmission techniques in next generation wireless networks for higher spectral efficiencies.