Decomposite channel estimation and equalization for GMSK-based system with transmit diversity

In this paper, multi-channel estimation schemes for a GMSK-based system with transmit diversity (space-time coding) are presented. For such a system, the channel information (impulse response) is critical for both space-time decoding and equalization at the receiver. Three non-blind estimation schemes, which decompose the channel in the process, are proposed for the GMSK receiver to obtain the impulse response of each of the multipath channels (i.e. transmit antennas): oversampling deconvolution, minimum mean-square error, and joint adaptive and correlation estimation. Since the received signal is the sum of emitted GMSK signals, interference cancellation is employed to facilitate the estimation process. Three cancellation algorithms, including direct cancellation, mean-square cancellation, and iterative cancellation, combined with each channel estimation method are investigated and compared. The estimated channel information will feed to the receiver consisting of space-time decoder and equalizer to decode the symbols of interest. Two receiver architectures are investigated in this paper, where the first design is the space-time decoder followed by the equalizer, the other is in the reverse way (equalizer followed by space-time decoder). In each of the two receiver architectures, the channel estimation needs additional modification and so does the equalizer. The equalizer in the design is a maximum likelihood sequence estimation (MLSE) based on Viterbi algorithm. To prove the concept and algorithms, both simulation and hardware implementation are performed. From the experimental results, it is shown that all the channel estimation algorithms can produce acceptable impulse response for space-time decoding and equalizer, in which the joint adaptive estimation with iterative cancellation is superior to the others. It is also shown that the diversity gain of this transmit diversity system is as good as a system with the same degree of receive diversity.