Normalized weighted dual-mode blind equalization algorithm
WU Chang-qi · Journal of Yanshan University · 2010
Constant modulus algorithm (CMA) is robust in opening the channel eye, making it suited for start-up phase. But the convergence rate is slow and the steady state error is high. Decision directed algorithm has fast convergence rate and low steady state error, but demands high correctness of threshold device, making it suited for trace phase. By virtue of adjusting the step size dynamically, normalized adaptive filtering algorithm can improve the convergence rate. Combining the advantages mentioned above, a weighted dual-mode blind equalization algorithm for 16QAM system is proposed in this paper. Comparing to switched dualmode algorithm, the equalizer is more robust and the convergence rate is faster. What's more, the new algorithm reduces the complexity of the receivers by avoiding switching over between two different modes. And it performs better in low signal to noise ratio circumstances.