Locally Fair Channel Allocation (Lofca) Scheme for Self-Organizing Small-Cell Base Stations

Stochastic geometry (SG) allows us to model and analyze unplanned small-cell base station (BS) deployments. SGbased studies use mainly the Universal frequency reuse (UFR) scheme in which all of the available spectrum is occupied by all of the BSs. Such an intense frequency allocation at the BSs causes high interference and low signal-to-interference-plus-noise ratio (SINR) regimes at the receivers that deteriorate the coverage probability of the network. To overcome this issue, BS coordination and frequency reuse (FR) schemes are introduced in the literature. However, their negative impacts on rate and capacity are not well-studied and mostly ignored. In addition, there are some accuracy issues in the proposed analytical models due to simplifications done in Laplace space. In this paper, first, we derive exact and approximate probability density functions (PDF) of received-signal-strength (RSS), interference, and SINR. Then, we propose a novel locally fair channel allocation (LoFCA) scheme to improve coverage probability and SINR regime while preserving capacity characteristics of the network. In LoFCA, we introduce the local fairness concept (LFC), which guides us to smartly allocate spectrum portions according to varying BS densities in the network. In denser regions, BSs coordinate and allocate lower number of subchannels to reduce interference and thus improve the coverage. Whereas in sparser regions, BSs allocate higher number of subchannels to improve capacity by increasing frequency usage per unit area. Numerical evaluations illustrate that LoFCA improves probability of coverage comparing to baseline benchmark model, and outperforms random frequency reuse schemes according to average capacity comparisons