Selection of optimal SINR threshold in fractional frequency reuse by comparing Otsu's and entropy method

Interference is a major challenge in cellular networks and has been mitigated by introducing a frequency re-use factor to alternate the frequency used by adjacent cells. Fractional Frequency Re-use is an approach that is utilized in handling interferences which occur between cells, particularly in cellular networks using long-term evolution and leads to better bandwidth utilization. The basic cell structure is split into two regions, a partial re-use region and a full re-use region. The full re-use region uses the same frequency for all the cells while the partial re-use regions adopt a re-use factor to mitigate interference with neighboring cells. The Signal-to-Interference-plus-Noise-Ratio threshold is a vital parameter to be set in Fractional Frequency re-use since it sets the basis of determining the subscribers that will fall within the full re-use region and those that will fall in the partial re-use region. The bandwidth is then allocated proportionately based on the number of subscribers in respective regions using a frequency partitioning ratio. The focus of this paper develop a dynamic thresholding technique by testing the performance of two thresholding techniques which are mostly used in image processing and include Otsu's, and entropy methods The results are validated by comparing the performance metrics of throughput and fairness with a dynamic thresholding technique from literature called centralized dynamic frequency allocation technique. The cell Signal-to-Interference-plus-Noise-Ratio value obtained from subscribers in the network are used to set threshold which is periodically updated as the subscribers continuously report their signal values. Simulation results show that Otsu's method outperforms entropy method in terms of throughput averagely by 29% and 23.3% in terms of fairness. Otsu outperforms centralized dynamic frequency allocation technique averagely by 39% in throughput and 32.9% in fairness.