在無線通訊系統當中，多重路徑傳播通道所造成的符元間干擾效應往往會降低符元偵測的可靠度。當通道資訊為事前已知時，可以利用等化器來消除此效應。然而在實際的應用當中，通道資訊通常是處於未知的狀態，通道參數和傳送符元需要聯合估測的情況。本論文基於上述前提，提出一種創新的方法來解決通道係數和符元估測的問題。 此方法分為三級，第一級是利用粒子群最佳化演算法進行通道係數估測，引入並列干擾消除的做法，加快通道係數估測的收斂速度；第二級是利用估測出來的通道參數結合基因演算法進行通道估測，經過擇優、交配、突變和並列干擾消除進化到下一個世代，直到收斂為止，而得到估測符元；接著使用第三級的遞迴技術提供互助機制使通道估測和符元偵測的品質進一步提升。經由電腦模擬分析的結果證明本論文所提出的方法能在運算複雜度大幅降低之情形下，提供近似於最大可能性演算法的最小位元錯誤率。 In wireless communications, inter-symbol interference (ISI) caused by channel's multipath propagation often reduces reliability of symbol detection. When channel information is known a priori, an equalizer can be employed to eliminate the ISI effect in symbol detection. However, in practical situations, channel information is usually unknown and joint estimation of channel parameters and transmitted symbols is often needed. As such, a novel scheme is proposed to handle the joint problem of channel estimation and symbol detection in this thesis. The proposed scheme involves three stages. The first stage deals with channel estimation using the particle swarm optimization (PSO) algorithm in conjunction with the parallel interference cancellation (PIC) to increase the convergence speed. The second stage detects transmitted symbols based on the genetic algorithm (GA) with PIC embedded, which uses the estimated channel parameters to carry out selection, crossover, and mutation and evolves to the next generation until convergence is achieved. The third stage offers cooperation between channel estimation and symbol detection through an iterated loop to further enhance estimation accuracy. Simulation results show that the proposed algorithm nearly achieves the lowest bit error rate offered by the maximum likelihood method, at a significant saving of operational complexity.