Optimization and Performance Analysis of Linear Dispersive STBC Codes with Different Estimators

Abstract

The Linear Dispersion Codes, a type of STBC, have been of great importance nowadays, as they subsume the important characteristics of the STBC codes and have decoding simplicity of VBLAST codes. These codes perform well in case of both Ergodic capacity and the error probability. The salient advantage of the code is its ability to get designed for any configuration of transmit and receive antennas, codeword lengths and rate. It is for this reason that we extend our work on the optimization of these codes in the thesis work. We considered a flat fading, open loop MIMO communication scheme where the transmitter having multiple antennas encodes the desired message signal into codeword matrix using an Optimized Linear Dispersion encoder. The designing and optimization of the encoder has been inspired from the different design methods. Non vanishing determinant property has been employ-yed to check whether it obtains an optimal diversity trade-off. The codes designed have been verified to satisfy NVD property. Further, Rank and determinant criteria has been obtained for the Integer Forcing receiver and used to optimize the encoder matrix. The performance of the optimized Linear dispersive Encoder has been evaluated using the Sphere Decoder and has been compared with different estimators such as Zero Forcing, MMSE receivers. A new decoding scheme, Integer forcing equalization scheme has been combined with the Encoder to efficiently estimate the codeword matrix. Monte Carlo simulations were used to compare its performance with other linear estimators and decoders, in terms of bit error probability. It has been observed that the designed model achieves better performance as compared to conventional linear receivers and have lower complexity than that of the Sphere Decoder.