几种发射分集OFDM系统信道估计方法性能评估

Performance Evaluation of Several Channel Estimation Methods for the OFDM System with Transmitter Diversity

给出了3种基于发射分集的OFDM系统信道估计方法,分别从训练序列对算法的影响、估计准确度、算法复杂度和算法适用性等方面对各自性能作了分析比较,并进行了仿真。分析比较和仿真结果表明,训练序列对3种算法的影响很大,通过训练序列的正交化设计,LS和EM算法可以得到很大程度的简化;EM算法与LS算法估计精度相当,信噪比较大时,多项式模型算法估计精度要远远高于LS和EM算法;总体上来说,多项式模型算法复杂度最高,EM算法复杂度最低。

Several channel estimation methods for the OFDM system with transmitter diversity are studied, furthermore, analysis and comparisons are made in details, meanwhile, computer simulations are conducted to verify these analysis and comparisons. It is shown that training sequences have a great effect on these three channel estimation methods, as a result, both LS and EM algorithm can be simplified to a great extent with the orthogonal training sequences. Theoretic analysis and simulation results also reveal that EM and LS have the same estimation accuracy and polynomial model based method has great advantage over them when SNR is larger. However, as a whole, the polynomial model based method is of highest computational complexity while the EM is lowest.