Multiple-InputMultiple-Output (MIMO) systems promise higher data rates and better quality of service for wireless communications, by using multiple antennas at both the transmitter and receiver. However, applying MIMO technology at small portable wireless devices is faced with the problem of mutual coupling between antenna elements due to the limited space to put multiple antennas. It is shown in the literature that the mutual coupling degrades the MIMO performance. For a given channel matrix and a known mutual coupling model, using antenna impedance matching network(s) between the coupled antenna array and its load or source network is proposed by recent studies to counteract the mutual coupling effects and maximise the MIMO performance. There are two issues with the existing matching techniques. First, they employ a model based on open-circuit voltages that separates the channel matrix and the mutual coupling model. This model is not valid except for a limited types of antennas (e.g. half-wavelength dipoles). Secondly, there is no solution among existing approaches that are capable of adapting to variations of the channel matrix. This thesis focuses on the mutual coupling problem at the receiver. We first examine the most common approaches to model the mutual coupling. For instance, we compare various definitions of coupling matrix available in the literature, analyse their relationship and clarify when we can use them. The mutual coupling effects on MIMO performance metrics and impedance matching are also investigated using the conventional open-circuit voltage based model and a new method called receiving mutual impedances. Then we propose the idea of having an adaptive uncoupled impedance matching technique which tunes the antenna impedance loads to compensate the effects of the propagation channel and mutual coupling together by directly dealing with the received signals. The mutual coupling model is unknown, but it is included implicitly by using the voltages across the real parts of the antenna load impedances to estimate the total effects. Assuming identical impedance loads for all receive antennas, several optimisation techniques such as Gradient-based, Newton-Raphson, and random search methods are investigated to implement such an adaptive impedance match. We found the random search method to be a simple and robust solution in comparison to other approaches. Finally, we extend this adaptive matching technique to non-identical termination case, in which all load impedances are tuned individually. The performance of the adaptive matching networks are compared with the conventional termination scenarios such as: characteristic impedance match, and self-impedance conjugate match. Simulation results for a 3 × 3 MIMO system under different propagation scenarios show that both identical and non-identical adaptive impedance matching networks are capable of optimising the performance in the presence of strong mutual coupling and time variations of the channel. The adaptive non-identical match gives a significant improvement in the mean capacity (more than 20% compared to conventional terminations for 0.05λ element separation) at the expense of a longer convergence time compared to the identical match.

No description available.