interference on wireless network： to surpress or decoding？.pdf

Interference in Wireless Networks: To Suppress or to Decode? Xiaodong Wang Electrical Engineering Dept. Columbia University New York $ % Multi-cell Networks future networks are interference-limited ? shrinking cell sizes ? ambitious spectral efficiencies ? universal frequency reuse recent developments ? MIMO networks ? game theory ? interference alignment Columbia University 2 $ % Multi-cell Downlink Systems interference management: decode or suppress? conventional way: suppress interference ? precoding ? scheduling new look: decode interference ? interference has structure ? decoding it might be helpful Columbia University 3 Motivations We consider a downlink OFDMA network with full frequency reuse among cells. Users’ data symbols are known only to the serving access point and single-user detection is employed at each one of the single-antenna mobile terminals. Base station (BS) coordination mitigates inter-cell interference by jointly optimizing the set of beamforming vectors used to serve the co-channel terminals scheduled on each resource slot (tone). Problem formulation A set of BSs simultaneously transmit over orthogonal sub-channels Each BS is equipped with T transmit antennas and SDMA is employed to serve multiple single-antenna users on each sub- channel. Let be the set of users scheduled by BS on sub-channel . Problem formulation Received signal by user on sub-channel n: Problem formulation Problem formulation We wish to maximize a network-wide utility function across all users of all coordinated BSs and all sub-channels, by choosing the set of beam-vectors s.t. the per-base-station power constraints: Problem formulation Nonconcave problem Our objective is to develop a distributed solution where each BS updates its beam- vectors locally; and with limited information exchange among the BSs, some form of optimality can be achieved. We resort to the game theoretical tool of pricing mechanisms. Utili