Channel estimation and signal-space diversity for vector-valued transmissions

Abstract

To combat the degradation effects of the communication channel, which increase the probability of error at the receiver, researchers have turned to the spatial dimension. As a matter of fact, the spatial dimension has a two fold advantage: it can combat the effect of fading and shadowing and it also increases the channel capacity. The channel capacity determines the minimum required signal to noise ratio (SNR) to achieve error-free transmission at a given bandwidth efficiency. There is no general rule on how to reach the theoretical capacity in practice. Therefore, much research has been undertaken to find signal construction methods which can approach this bound as close as possible while employing the signal space and/or the spatial dimension. In this thesis, we concentrate on the problem of signal diversity design, channel estimation and channel estimation error effects on the communication system performance. In real situations, channel estimation is constrained by some factors like the limited training length and the efficiency of the estimation technique. Because of these factors, channel estimation errors result. In order to have a reliable transmission, the parameters describing the channel are required at the receiver. One common way to measure these parameters is to send a known training sequence just before the transmission of unknown data. The training symbols enable estimation of the channel characteristics at moderate cost. The most important aspects of this thesis are as follows: Effects of channel estimation errors on MIMO transmission, Different design criteria of signal-space diversity, Construction of the rotation matrices by using Hadamard or Fourier transform, Applying signal-space concept for MIMO transmissions, Comparison of different models for time-varying channels to be used for MIMO transmissions.