Image Transmission using Variable M-QAM with Optimized Bit Power Allocation

In this paper, optimization of minimum mean square error for image transmission using a variable M-QAM system is proposed for improving both power and bandwidth efficiencies. Optimization is based on bit power allocation according to the importance of bits to the overall image quality. An iterative off-line algorithm is presented to find the optimum power allocated and modulation level to different bits. Our results show that a 16-QAM system using the proposed algorithm with optimized bit power allocation provides an improvement in power efficiency as measured by the average Eb/No of about 2 dB over that using a conventional 16-QAM. However, the improvement increases significantly with the variable M-QAM system to about 6 dB while improving the bandwidth efficiency by about 17% over the conventional 16- QAM system.