Distributed resource allocation for inter cell interference mitigation in irregular geometry multicell networks

Abstract

Extensive increase in mobile broadband applications and proliferation of smart phones and gadgets require higher data rates of wireless cellular networks. However, limited frequency spectrum has led to aggressive frequency reuse to improve network capacity at the expense of increased Inter Cell Interference (ICI). Fractional Frequency Reuse (FFR) has been acknowledged as an effective ICI mitigation scheme but in irregular geometric multicellular network, ICI mitigation poses a very challenging issue. The thesis developed a decentralized ICI mitigation scheme to improve both spectral and energy efficiency in irregular geometric multicellular networks. ICI mitigation was realized through Distributed Resource Allocation (DRA) deployed at the cell level and region level of an irregular geometric cell. The irregular geometric cell consists of a minimum of four regions comprising three sectors and a central region. DRA at the cell level is defined as Multi Sector DRA (MSDRA), and at the region level is defined as Distributed Channel Selection and Power Allocation (DCSPA). MSDRA allocates discrete power to every region in a cell based on Game Theory and Regret Learning Process with correlated equilibrium as the optimum decision level. The DCSPA allocates power to every channel in a region based on non-coalesce liquid droplet phenomena by selecting optimum channels in a region and reserving appropriate power for the selected channels. The performance was evaluated through simulation in terms of data rate, spectral efficiency and energy efficiency. The results showed that MSDRA significantly improved cell data rate by 58.64% and 37.92% in comparision to Generalized FFR and Fractional Frequency Reuse-3 (FFR-3) schemes, respectively. The performance of MSDRA at the cell level showed that its spectral and energy efficiency improved 32% and 22%, respectively in comparison to FFR-3. When the number of sectors increased from three to four, data rate was improved by 30.26% and for three to six sectors, it was improved by 56.32%. The DCSPA further improved data rate by 41.07% when compared with Geometric Water Filling, and 86.46% in comparison to Asynchronous Iterative Water Filling. The DCSPA enhanced data rate achieved in MSDRA by 15.6%. Overall, DRA has shown to have significant improvement in data rate by 53.6%, and spectral efficiency by 38.10% as compared to FFR-3. As a conclusion, the DRA scheme is a potential candidate for Long Term Evaluation – Advanced, Fifth Generation networks and can be deployed in future heterogeneous irregular geometric multicellular Orthogonal Frequency Division Multiple Access networks.