Power-saving heterogeneous networks through optimal small-cell scheduling

Traditional macro-cell networks are experiencing an explosion of data traffic, and small-cell can efficiently solve this problem by efficiently offloading the traffic from macro-cells. Given massive small-cells deployed in each over-crowed macro-cell, their aggregate power consumption (though low for an individual) can be larger than that of a macro-cell. To reduce the total power consumption of a whole heterogeneous network (HetNet) including macro-cells and small-cells, we dynamically schedule the operating modes of all small-cells (active or sleeping) in each macro-cell, while keeping the macro-cell active to avoid any service failure in coverage. When mobile users (MUs) are homogeneously distributed in a macro-cell according to a Poisson point process (PPP), we optimally propose small-cell location-based scheduling scheme to progressively decide the states of small-cells according to their distances to the corresponding macro-cell base station. Finally, we turn to a more general case where MUs are heterogeneously distributed in different small-cells. We first prove that the optimal scheduling problem is NP-hard and then propose a location-and-coverage-based scheduling algorithm which gives a suboptimal solution in polynomial-time. Simulation results show that the performance loss of our proposed algorithm is less than 1 percentage from the perspective of network power consumption.