Robust Transmit Signal Design and Channel Estimation for Multiantenna Systems

Abstract: In the development of advanced signal processing techniques, dealing with both uncertainties and computational burden is essential. Taking the uncertainties into consideration is required in order to guarantee a certain level of performance, even when the system is designed based on imperfect prior knowledge. Addressing the computational issues is required due to the great popularity of large-scale systems in recent years, where low-complexity signal processing techniques are not only desirable but required.The thesis considers robustness and complexity as two lines of analysis for problems related to parameter estimation and spatial power distribution. In particular, the robustness is realized by means of designing transmit signals using optimization frameworks which account for uncertainties. The complexity issue, on the other hand, is addressed by proposing methods to reduce the computational load for the Bayesian parameter estimation. The common theme throughout the thesis is the use of array with multiple elements which is applied to different applications: wireless communication systems, radars and hyperthermia therapy.The first part of the thesis focuses on the multiple-input multiple-output (MIMO) channel estimation problem. First, a worst-case robust design framework is introduced to deal with uncertain model parameters. This framework is developed specifically to provide robust training sequences for minimum mean square error (MMSE) channel estimation in MIMO communication systems. Second, low-complexity estimators are proposed for large-scale MIMO systems in order to be applied instead of the optimal MMSE estimator which suffers from high computational complexity.The second part of the thesis deals with spatial power distribution using a multi-element array. The distribution of power in space can be controlled in an optimal manner thanks to waveform diversity offered by the MIMO technology. Robust waveforms are designed by using a worst-case robust optimization framework which is tailored specifically for radar and hyperthermia therapy applications, taking uncertain array steering vectors into account.