The ever increasing growth of wireless communication Systems has continued to drive the research efforts towards obtaining novel techniques by which system capacity can be increased, and at the same time maintaining high-quality of services. This, as earlier mentioned, has brought about the migration from single antenna, single input single output (SISO) Systems to deployments of multiple antennas at both ends of the wireless communication Systems. Emerging from this migration is the multiple-input multiple-output (MIMO) Systems. From the spectral efficiency angle of wireless communication is the emergence of orthogonal frequency division
multiplexing (OFDM) which finds deployment in both single antenna and multiple antenna wireless communication Systems. The concepts of MIMO and OFDM were combined with the emerging intent of exploiting the advantages of both techniques. This combination has given the development to MIMO-OFDM wireless communication Systems with the expectation of having spectrally efficient, high data rate system that is robust to frequency selective fading channels.
With the area of applications of the MIMO-OFDM system expanding very fast, the requirement for an improved functioning of the Systems is becoming very high. As a result, more research efforts are being directed towards achieving better MIMO-OFDM Systems performance.
However, one of the major challenges to either single antenna, SISO OFDM, or MIMO-OFDM communication Systems is means of providing accurate channel state information (CSI) at the receiver end of the Systems for coherent detection of the transmitted signal. If the CSI is not available at the receiver, the transmitted signal could only be demodulated and detected through a non-coherent method such as the differential demodulation technique. However, the employment of non-coherent detection method is at the expense of about 3-4 dB loss in signal-to-noise ratio (SNR) compared with using the coherent detection method [31]. In order to eliminate such a huge loss, it is imperative to develop an efficient and cost effective technique of providing channel state information at the receiver for coherent detection of the transmitted information in MIMO- OFDM wireless communication Systems.
There are different techniques by which channel state information can be obtained and these are classified as pilot-assisted (training-based), blind and decision-directed channel estimation methods. In the context of pilot-assisted channel estimation scheme, training-data that is known a priori at the receiver is transmitted along with the message data from the transmitter. These training data is then used to obtain the samples of CSI at the training data‟s locations. The CSI at the message data‟s locations are obtained from CSI at the training data‟s locations by means of interpolation techniques. The insertion of the training data within the message signal will definitely induce additional overhead and thus reducing the data throughput. In blind channel estimation method, no training data sequence is needed; instead the statistical properties of the channel and certain information about the transmitted signal are employed to obtain the CSI.
Consequently, there is saving in the bandwidth usage while employing blind channel estimation method in comparison with the training-based method. Though the blind channel estimation method has its advantage in that it has no overhead loss, unfortunately it can only be applied to slowly time-varying fading channels. This is because it will have to memorize the data record for
a long time. Thus, it can not be applied in fast-varying channel scenarios that are peculiar to mobile wireless communication Systems. Besides, blind channel estimation methods also tend to become heavier in terms of computational complexity [74].
Consequently, in this thesis we investigate the training-based channel estimation schemes for single antenna system rather than the blind channel estimation method. Our investigation leads us to develop a low complexity adaptive algorithm that is robust against both slow and fast fading channel scenarios, in comparison with other algorithms employed in literatures, to implement soft iterative channel estimator for turbo equalizer based receiver for single antenna communications Systems.
In the decision directed channel estimation method, all the detected data at the receiver are used for channel estimation. Hence, few numbers of pilots, in comparison with the pilot- assisted method, are required to commence the estimation process in the decision directed channel estimation method. The gain obtainable with this method in comparison with purely pilot-assisted scheme has motivated this research to focus on the decision directed channel estimation method for single antenna OFDM Systems and MIMO-OFDM Systems. In this thesis a faster and low complexity subspace algorithm, in comparison with other algorithms employed by some other authors in literature, is proposed for parametric estimation of the channel impulse response of SISO-OFDM and MIMO-OFDM Systems. Besides, a low complexity adaptive predictor, in comparison with other available ones in literature, is derived for implementation of the adaptive predictor module of the proposed decision directed channel estimation method for the SISO OFDM and MIMO-OFDM Systems. In addition, the low complexity adaptive algorithm we proposed for channel estimator in single antenna communication Systems is also proposed for use to implement the temporary channel transfer function estimator module of the proposed decision directed channel estimation method for the SISO-OFDM and MIMO-OFDM Systems.
In addition, iterative technique that is based on turbo principle is employed for the channel estimation schemes proposed in this thesis for the single antenna Systems and MIMO-OFDM wireless communication Systems.