Rayleigh信道下的支持向量機(jī)多用戶檢測(cè)方法
A support vector machine based detection method on rayleigh channel
-
摘要: 在BPSK調(diào)制的DS-CDMA中,基于支持向量機(jī)(Support Vector Machine,SVM)的多用戶檢測(cè)方法采用支持向量機(jī)的分類方法將接受向量分成+1和-1兩類,達(dá)到檢測(cè)的目的。與MMSE方法不同的是,支持向量機(jī)分類器的目的是找出一個(gè)能將訓(xùn)練向量中信號(hào)為+1和信號(hào)為-1的兩類數(shù)據(jù)分離的最佳分離超平面。從數(shù)值仿真結(jié)果可以看出,在Rayleigh信道,這種支持向量機(jī)的多用戶檢測(cè)方法與MMSE多用戶檢測(cè)器相比,輸出能達(dá)到較低的誤碼率。
-
關(guān)鍵詞:
- Rayleigh信道; 支持向量機(jī); 兩分問題; 多用戶檢測(cè)
Abstract: In DS-CDMA system with BPSK modulation, Support Vector Machine (SVM) based multiuser detection uses SVM classification method to classify received vectors into two classes. One is the received vectors of desired users symbol +1 and the other is the vectors of desired users symbol -1. So desired users symbol can be detected by this method. Different with MMSE method, SVM classifier finds the optimal separating hyperplane that separates the class of +1 and class of -1. Simulation results show that the performance of the SVM detector is better than that of MMSE detector in Rayleigh channels. -
B. Aazhang, B. Paris, G. Orsak, Neural networks for multiuser detection in CDMA communication, IEEE Trans. on Communications, 1992, 40(7), 1212-1222.[2]C. Burges, A tutorial on support vector machines for pattern recongition, Data Mining and Knowledge Discovery, 1998, 2(2), 121-167.[3]U. Madhow, M. Honing, MMSE interference suppression for direct-sequence spread-spectrum CDMA, IEEE Trans. on Communications, 1994, 42(12), 3178-3188.[4]U. Madhow, MMSE interference suppression for timing acquisition and demodulation in directsequence CDMA systems, IEEE Trans. on Communications, 1998, 46(8), 1065 1075.[5]G. Proakis, Digital Communication, Third Edition, USA: McGraw-Hill, 1995, 758-833.[6]S. Rappaport, Wireless Communications Principles and Practice, USA, Prentice-Hall, 1996, 177181.[7]S. Keerthi, S, Shevade, C. Bhattacharyya, K. Murthy, A fast iterative nearest point algorithn for support vector machine classifier design, IEEE Trans. on Neural Networks, 2000, 11 (1), 124-136. -
計(jì)量
- 文章訪問數(shù): 2020
- HTML全文瀏覽量: 97
- PDF下載量: 576
- 被引次數(shù): 0
下載:
