Optimum bit detection of quadrature amplitude modulation in self correlated gaussian noise channel

This paper discusses the optimum bit detection for coherent QAM Quadrature Amplitude Modulation) in the presence of colored Gaussian noise. We employ Gray-mapping for constellations, and derive the log-likelihood ratio (LLR) for each bit forming the symbol. Using LLR, we derive the average bit error probability for individual bits. We further investigate the equalization methods under colored noise to facilitate optimum detection.