School of Engineering and Materials Science
Research Student Awards
PhD Thesis: Transmission power control in wireless networks
Author: LIAO, Rui
Supervisor(s): Henri Huijberts
We first introduce some important techniques and results involved in the development of power control algorithms and give literature review. A PI power control approach from literature is introduced. Due to lack of stability analysis, we show there are problems in the existing algorithm. We then propose a stable Proportional-integral (PI) power control algorithm. A forgetting factor is adopted to improve the transient performance. Distributed power control algorithms for systems with fixed SNIR targets might diverge when the feasibility condition is not satisfied. Multi-objective optimisation (MO) is adopted to deal with power control with variable SNIR targets. After discussing the existing MO algorithm, we propose a quadratic multiobjective-optimisation (QMO) algorithm where a quadratic objective function and the greedy methodology are adopted for the dynamics. Theoretical and simulation results of convergence of the new algorithms are given.
We also provide review of some important power control frameworks which can be used to show convergence of power control algorithms. However, the QMO algorithm does not fall into any existing framework. In order to show convergence of the QMO algorithm, we suggest a new generalised framework in this thesis.