Study on parameter estimation for the short length cyclic prefix OFDM signal

In order to overcome the shorticomings of traditional parameters estimation methods for the short length OFDM signal that required lots of OFDM symbols，and had the poor performance and the poor anti multipath fading capability. This paper presented a novel method to estimate the parameters for short cyclic prefix length OFDM signal. The method used the OFDM signal model to compute the model function of inverse fourier transform, and according the prior probability density of OFDM signal, we construct the maximum likelihood function of multi parameter fusion on the basis of model function and theoretically demonstrated that the joint estimation for cyclic-prefix length and useful symbol length can be achieved by detecting the lowest value of maximum likelihood function. In addition, we used dynamic particle swarm optimization algorithm(DPSO) to reduce computational complexity and shorten the estimated time. The experimental results illustrate the robustness of the proprosed algorithm to estimate the parameters for short CP OFDM signal under various circumstance, and show that the proprosed algorithm has better performance than conventional algorithm.