Mimo - Ofdm Adaptive Modulation System Research

This thesis introduces LS(Least square), MMSE(Minimum Mean-Square Error) channel estimation algorithms, and improved channel estimation based on FFT. It verifies feasibility of the improved algorithms by computer simulation. The improved algorithms effectively reduce the MSE of channel estimation, improve the channel estimation accuracy.In detail, using the SVD(singular value decomposition)to obtain the virtual SISO channels, this thesis puts forward optimal constant-bit-rate and constant-power distributing water-filling algorithm in MIMO-OFDM system, and displays algorithm flows for each of them. For reducing the difficulty of the water-filling algorithm in MIMO-OFDM system, this thesis presents an improved water-filling algorithm, and this thesis proves its feasibility. And then, improved water-filling algorithm is modeled with Matlab and validated that improved algorithm can achieve the same optimal solution as the traditional water-filling algorithm, while the count of iterations is significantly reduced. So this improved algorithm effectively reduces the complexity of the water-filling algorithm, saves system resources and improves real-time performance of the systemBits sent in the real system can only take integer. For application of improved water-filling algorithm in practical environment, this thesis introduces greedy bit-adjustment loading algorithm and its improved version. Greedy bit adjustment loading algorithm adjusts and optimizes the power and bit distribution after bit’s rounded, while its improved version reduces the number of comparisons and improves the efficiency of the algorithm. Finally, greedy bit adjustment loading algorithm and its improved version are simulated and verified with Matlab software. Through greedy bit adjustment loading algorithm, adaptive modulation algorithm is applied to the actual system and performance of the adaptive system after bit’s rounded is improved.