Interference Management in MIMO Wireless Networks

With the growing demands for wireless communication services to support higher bit rates, there is an ever-pressing need to increase bandwidth efficiency of communication systems through the reuse of the same spectrum by simultaneous multiple transmissions. To address this and many other issues, one of the most successful and efficient solutions is the deployment of Multiple-Input Multiple-Output (MIMO) technology. However, the multi-user (MU) MIMO systems are often designed to use non-optimized methods to deal with the interference limiting the network capacity. As a result, this proposal seeks to design signal processing strategies and wireless network protocols that manage interference in MU MIMO systems to achieve the highest network throughput.

Conventional approaches to deal with interference are (i) to avoid interference by non-overlapping transmissions in time or frequency domains, (ii) lump other transmitters' signals as noise and mitigate it through power control, or (iii) deploy multi-user decoding. In MU-MIMO systems, where the spatial signaling dimension is utilized to serve many users in parallel, there are opportunities for shaping the interference through linear pre-coding at the terminals and power control to minimize the effects of the combined interference at the destinations.

In this proposal, we seek to develop strategies to deal with interference in MU MIMO systems by integrating signal alignment for few strong transmitters and interference cancellation for non-coordinated transmissions. Specifically, by a thoughtful design of pre-processing vectors and scheduling of transmissions, we expect that the signals with relevant information will be aligned into proper spatial dimensions according to the selected coding strategies. The transmission schemes will be also optimized to reduce computational complexity and to improve time-frequency utilization. The scope of this work will cover single-hop, multi-hop, and multi-way networks. In multi-hop systems, where a source and destination communicate through one or more other nodes that act as a relay, we plan to exploit the concept of physical layer network coding to aid the operation of the wireless network with the nodes distributed in space according to the Poisson-point process. Our focus here is on medium access (user selection) opportunities for power allocation, with emphasis on the topology control and routing-level issues for inter-relay interference cancellation. We expect here to deploy for analysis the tools from stochastic geometry which computes macroscopic properties of wireless networks, by averaging over all potential geometrical patterns for the terminals. The challenge in this work will be to evaluate the gains from the deployed signal alignment techniques where all terminals may not be synchronized or fully cooperating.