Mixed Timescale Cross-Layer Optimization for Multi-Antenna D2D Networks

A mixed timescale cross-layer resource optimization scheme is proposed for a multi- antenna device-to-device (D2D) network. We consider the two timescale joint optimization of beamforming, routing and flow control, where the short-term beamforming control at each transmit device is adaptive to the local real-time CSI, while the long-term routing and flow control is adaptive to the global topology and channel statistics of the D2D network. We propose a stochastic cutting plane algorithm to find the global optimal solution of the joint optimization problem. The proposed solution has low signaling overhead and does not require explicit knowledge of the channel statistics. Simulations show that the proposed solution achieves significant performance gain over several baselines.