Binary power control for multi-cell capacity maximization

We consider the problem of optimally allocating the base station transmit powers for a wirelessmulti-cellular (N-cell) system in order to maximize the total system throughput under interference and noise impairments, and short term (minimum and peak) power constraints. Employing dynamic reuse of spectral resources, we impose the power constraints at each base station and allow for coordination between the base stations. For the two-cell case, the capacity-optimal power allocation has been previously shown to be binary [1]. We now propose to perform binary power allocation, (by simply checking the corners of the domain resulting from the power constraints), also whenN >2, and we identify two scenarios in which the optimality of binary power control can be proven also for arbitrary N. Furthermore, in the general setting for N >2, we demonstrate by simulations that a network performance with negligible loss, compared to the best non-binary scheme found by geometric programming, can be obtained.