Diversity coding with interference avoidance

We study performance tradeoffs for joint diversity enhancement and interference mitigation in multi-cell networks with multiple antennas at the transmitter. Our transmitter combines a space time coding block followed by a multimode beamformer. We assume that the transmitter knows the channel covariance for its own and the interfered user. The coding techniques maximize link performance via diversity gain for its own user while constraining the interference power to the interfered (or other) user. We examine strategies for choosing this interference constraint based on greedy or co-operative methods for a symmetric two cell scenario where each link interferes with the other. In the greedy algorithm, each user selfishly maximizes diversity to its own user without regard to the interference it generates to the other user. Both users account for the incoming interference and they eventually iteratively reach the Nash equilibrium (NE). In the co-operative algorithm, each user balances the diversity to its own user but co-operates by limiting the amount of interference to the other user. We show that in the co-operative case, it is possible to find Pareto dominant solutions where the users either maintain or improve their symbol error rate performance