在本篇論文中，我們提出了一個適用於並行調適性濾波器的實數對稱限制最大事後機率演算法(CMA+RSC-MAP)，首先將先介紹如何去推導出這個在盲目演算法中最大事後機率演算法具有的共軛對稱限制的最佳權重解，接下來我們利用這個共軛對稱限制的最佳權重解去導出這篇論文所提出的演算法，並且利用調適的方式讓權重能夠趨近於這個最佳權重解。最後在模擬的結果中，我們也會呈現出CMA+RSC-MAP演算法確實擁有比目前新型的並行調適性濾波器還要更好的效能。與原本的CMA+SC-MAP演算法相比，CMA+RSC-MAP演算法確實擁有跟CMA+SC-MAP演算法一樣好的效能，並且擁有更低的運算複雜度。 In this paper, we propose a real-valued SC-MAP (RSC-MAP) algorithm for concurrent adaptive filter (CAF) applied to beamforming. We first contribute to deriving a closed-form optimal weight expression for blind MAP algorithm. A conjugate symmetric property associated with optimal blind MAP weights is further acquired. Then, we use the conjugate symmetric constraint to guide the proposed RSC-MAP algorithms to follow the optimal blind MAP expression form during adapting procedure. In the simulations, we show that the proposed RSC-MAP algorithms have better performance than the classic ones. Compared with SC-MAP, the RSC-MAP with less computational complexity has the same bit-error rate performance.
