基于隐马尔科夫模型的CPM信号盲Turbo均衡算法

doi:10.11959/j.issn.1000-436x.2015071

摘要/Abstract

摘要：

针对高阶连续相位调制信号(CPM)盲均衡中存在的均衡性能较差以及不容易收敛等问题，从隐马尔科夫模型(HMM)的角度出发，建立一种新的EM-SOVA信道盲均衡方法，并结合Turbo均衡思想，提出了一种适用于高阶CPM信号的盲Turbo均衡算法。该算法通过将SOVA内嵌到EM算法的迭代过程中，有效改善了信道均衡效果，同时使用Turbo均衡的软信息迭代处理来进一步提高低信噪比条件下系统性能。理论分析和仿真结果表明，所提算法具有良好的盲均衡性能以及收敛性。

关键词: 连续相位调制, 隐马尔科夫模型, 软输出维特比算法, 期望最大化算法, Turbo均衡

Abstract:

To solve the issues of the poor performance and convergence difficulties in the blind equalization of high order continuous phase modulation (CPM) signals, a new EM-SOVA channel blind equalization method was developed from the perspective of hidden Markov model(HMM). A novel blind Turbo equalization algorithm for high order CPM signals was proposed based on the combination of this method and Turbo equalization. The algorithm improved the equalization performance by applying the soft-output Viterbi algorithm(SOVA) within an EM iteration. Simultaneously, a soft infor-mation iterative process was used for further improvement of the system performance at a low signal-to-noise ratio. The analysis and simulation results show that the proposed algorithm provides a good blind equalization performance and convergence.

Key words: continuous phase modulation, hidden Markov model, soft-output Viterbi algorithm, expectation maximiza-tion algorithm, Turbo equalization

钟凯,彭华,葛临东. 基于隐马尔科夫模型的CPM信号盲Turbo均衡算法[J]. 通信学报, 2015, 36(3): 170-178.

Kai ZHONG,Hua PENG,Lin-dong GE. Blind Turbo equalization algorithm based on hidden Markov model for continuous phase modulation signals[J]. Journal on Communications, 2015, 36(3): 170-178.

图/表 9

图1

图2

图3

表1

本文算法与基于nSnOEE算法复杂度比较"

算法		复杂度		总复杂度	参数设置
算法	SOVA/MAP	自相关矩阵	最大化	总复杂度	参数设置
本文算法(SOVA)	QL_b	$Q L_{b} + Q D^{2}$	D³	$Q L_{b} + Q L_{b} D + Q D^{2} + D$	D=2,N=507,L_b=512,p=2,
本文算法(SOVA)	65 536	131 584	8	197 128	L=2,M =4,Q=pM^L+D_1=128
基于nSnOEE算法(MAP)	QML_b	$Q L_{b} + Q D^{2}$	D³	$Q M L_{b} + Q L_{b} D + Q D^{2} + D$	D=2,N=507,L_b=512,p=2,
基于nSnOEE算法(MAP)	262 144	131 584	8	393 736	L=2,M =4,Q=pM^L+D_1=128

表1

图4

图5

表2

图6

图7

参考文献 16

[1]	HOSSEINI E , PERRINS E . The Cramer-Rao bound for training se-quence design for burst-mode CPM[J]. IEEE Transactions on Communications, 2013,61(6):2396-2407.
[2]	NSENGA J , THILLO W V , HORLIN F . Comparison of OQPSK and CPM for communications at 60 GHz with a nonideal front end[J]. EURASIP Journal on Wireless Communications Networking, 2007,40(1):250-258.
[3]	CIRPAN H A , TSATSANIS M K . Blind receivers for nonlinearly modulated signals in multipath[J]. IEEE Transactions on Signal Processing, 1999,47(2):583-586.
[4]	HANSSON A , AULIN T . Generalized APP detection of continuous phase modulation over unknown ISI channels[J]. IEEE Transactions on Communications, 2005,53(10):1615-1619.
[5]	BIANCHI P , LOUBATON P . On the blind equalization of continuous phase modulated signals using the constant modulus criterion[J]. IEEE Transactions on Signal Processing, 2007,55(3):1047-1061.
[6]	JIN X D , ZHAO M J , SUN L . Equalization for continuous phase modulation over frequency-selective fading channels[J]. Journal of Zhejiang University SCIENCE A, 2007,8(12):1884-1888.
[7]	PARK C H , HEALTH R W , RAPPAPORT T S . Frequency domain channel estimation and equalization for continuous phase modulations with superimposed pilot sequences[J]. IEEE Transactions on Vehicular Technology, 2009,58(9):4903-4908.
[8]	李强，严庆，罗胜．基于符号频域均衡的CPM迭代检测算法[J]．电子与信息学报， 2011,33(7):1550-1555. LI Q , YAN Q , LUO S . Iterative detection algorithm based on symbol frequency domain equalization for continuous phase modulation[J]. Journal of Electronics ＆ Information Technology, 2011,33(7):1550-1555.
[9]	?ZGüL B , KOCA M , DELI? H . Pre-equalization for continuous phase modulation[J]. IEEE Transactions on Wireless Communications, 2011,10(7):2272-2282.
[10]	HASSAN E S , ZHU X , KHAMY S E . A chaotic interleaving scheme for the continuous phase modulation based single-carrier frequency-domain equalization system[J]. Wireless Personal Communications, 2012,62(5):183-199.
[11]	HASSAN E S , ZHU X , KHAMY S E . Chaotic interleaving scheme for single- and multi-carrier modulation techniques implementing continuous phase modulation[J]. Wireless Personal Communications, 2012,62(5):183-199.
[12]	LAOT C , BIDAN R L , LEROUX D . Low-complexity MMSE turbo equalization: a possible solution for EDGE[J]. IEEE Transactions on Wireless Communications, 2005,4(3):965-974.
[13]	许鹏，汪晋宽，祁峰．改进的高效MIMO-OFDM系统EM信道估计算法[J]．通信学报， 2011,32(1):87-93. LI Q , YAN Q , LUO S . Iterative detection algorithm based on symbol frequency domain equalization for continuous phase modulation[J]. Journal of Electronics ＆ Information Technology, 2011,32(1):87-93.
[14]	LIN S , COSTELLO D J . Error Control Coding[M]. USA: Prentice Hall, 2004.
[15]	BERTSEKAS D P . Nonlinear Programming[M]. USA: Athena Scientific, 1999.
[16]	MOON T K . The expectation-maximization algorithm[J]. IEEE Signal Processing Mag, 1996,13(11):47-60.

参数	真实信道	实验结果1	真实信道	实验结果2
抽头1	0.227 0	0.186 2	0.227 0	0.211 5
抽头2	0.460 0	0.520 5	0.460 0	0.493 2
抽头3	0.688 8	0.622 7	0.688 8	0.653 4
抽头4	0.460 0	0.415 8	0.460 0	0.432 7
抽头5	0.227 0	0.286 4	0.227 0	0.243 8
噪声方差(信噪比)	0.250 6 (3 dB)	0.236 2	0.125 6 (6dB)	0.128 4
信道NMSE	0.015 2		0.003 6
噪声NMSE	0.003 3		4.969 8×10^-4