Doctorate thesis defense of Safa ESSASSI

Committee

Abstract

In Long-Term Evolution (LTE) system, multi-access interference coordination has been planned to optimize the exploitation of radio resources, thus maximizing service quality, coverage and capacity. Many techniques aim to minimize inter-cell interference in LTE, such as fractional power control and spatial antenna techniques, including MIMO and SDMA. In LTE, the access technique rests on the OFDM modulation technique. The allocation of resources takes place both in time and frequency. These resources are allocated to user equipment according to a given frequency reuse approach. There are three most common used approaches namely frequency reuse 1, fractional reuse of static frequency with a pattern equal to 3, and adaptive fractional frequency reuse.

In this thesis, we were interested in the problem of optimization of resource block (RB) allocation and power control in the Uplink of LTE system. The UEs can be classified into two groups. First one concerns the users that produce high interference level. The second one depicts the users that cause low interference level. Each group compensates for the path loss differently from the other. Indeed, we are trying to approach the power control of users who do not cause much interference towards the total compensation of the path loss. This mechanism is based on the fractional compensation of the pathloss, also known as fractional power control, which aim at the minimization of the inter-cell interference. In the framework of this thesis, we studied and proposed a new solution for RB allocation to further reduce the interference level. This solution is based on the use of genetic algorithm (GA). We proposed three approaches for RB allocation optimization: (i) distributed, (ii) partially centralized, and (iii) centralized approaches.

Keywords

Resource Block allocation optimization, inter-cell interference, Genetic Algorithm, fractional Power Control, Cloud-RAN.