基于EM算法的极大似然分布式量化估计融合新方法
doi: 10.3724/SP.J.1146.2010.00599
New Method for Distributed and Quantitative Estimation Fusion of Multi-sensor Based on EM Algorithm
-
摘要: 该文针对水下目标探测中的多传感器分布式量化估计融合问题,建立了分布式量化估计融合模型,在考虑信道噪声且其统计特性不完全已知条件下,充分利用EM算法在观测数据缺失时参数估计的优越性,提出了一种基于期望极大化(EM)算法的极大似然分布式量化估计融合新方法。该方法将未知的水声信道噪声参数以及局部量化器量化概率建模为EM算法中二元高斯混合模型参数,利用极大似然估计方法的估计不变性得到目标参数的估计融合结果。仿真实验表明:该方法在局部传感器观测样本数目大于5000和信噪比大于6 dB时与已有理想信道条件下的估计方法性能相当,该方法为水下目标探测中分布式量化估计融合系统的工程实现提供了理论依据。
-
关键词:
- 水下目标探测 /
- 期望极大化(EM)算法 /
- 估计融合 /
- 极大似然
Abstract: For multi-sensor distributed and quantitative estimation fusion problem of underwater target detection, a model of distributed and quantitative estimation fusion is established. The channel noise and its statistical property which is not fully known to fusion center is considered, The superiority of Expectation Maximization (EM) algorithm completely is used in parameter estimation problem when the observation data is missing. A new algorithm of distributed and quantitative estimation fusion is proposed based on EM algorithm. In this method, the unknown parameters of underwater acoustic channel noise and the quantization probability of local quantizer are modeled as the binary Gaussian mixture model parameters. Then, the invariance of the maximum likelihood estimation is used to get the result of the estimation fusion. Simulation results show that the estimation performance of the new algorithm is comparable to the methods which need ideal channel condition when the number of local sensors samples is larger than 5000 and the signal to noise ratio is higher than 6 dB. This new algorithm provides a theoretical basis for realizing the distributed and quantitative estimation fusion system of underwater target detection. -
王志胜, 姜斌, 甄子洋. 融合估计与融合控制[M]. 北京:科学出版社, 2009: 36-38.[3]Fang Jun and Li Hong-bin. Distributed estimation of Gauss- Markov random fields with one-bit quantized data[J].IEEE Signal Processing Letters.2010, 17(5):449-452[4]Chen Hao and Varshney P K. Performance limit for distributed estimation systems with identical one-bit quantizers[J].IEEE Transactions on Signal Processing.2010, 58(1):466-471[5]Ribeiro A and Giannakis G. Bandwidth-constrained distributed estimation for wireless sensor networkspart II: unknown probability density function [J].IEEE Transactions on Signal Processing.2006, 54(7):2784-2796[7]Ramanan S and Walsh J M. Distributed estimation of channel gains in wireless sensor networks[J].IEEE Transactions on Signal Processing.2010, 58(6):3097-3107[8]Senol H and Tepedelenlioglu C. Performance of distributed estimation over unknown parallel fading channels [J].IEEE Transactions on Signal Processing.2008, 56(12):6057-6068[9]Arindam k. das mehran mesbahiDistributed linear parameter estimation over wireless sensor networks[J].. IEEE Transactions on Aerospace and Electronic Systems.2009, 45(4):1293-1305[10]Cattivelli F S and Sayed A H. Diffusion LMS strategies for distributed estimation[J].IEEE Transactions on Signal Processing.2010, 58(3):1035-1048[11]Song En-bin, Zhu Yun-min, Zhou Jie, and You Zhi-sheng. Minimum variance in biased estimation with singular fisher information matrix[J].IEEE Transactions on Signal Processing.2009, 57(1):376-381[12]Ribeiro A and Giannakis G B. Bandwidth-constrained distributed estimation for wireless sensor networkspart I: Gaussian case [J].IEEE Transactions on Signal Processing.2006, 54(3):1131-1143[13]Fang Jun and Li Hong-bin. Distributed adaptive quantization for wireless sensor networks: from Delta modulation to maximum likelihood[J].IEEE Transactions on Signal Processing.2008, 56(10):5246-5257[14]Aysal T C and Barner K E. Constrained decentralized estimation over noisy channels for sensor networks [J].IEEE Transactions on Signal Processing.2008, 56(4):1398-1410
计量
- 文章访问数: 3426
- HTML全文浏览量: 85
- PDF下载量: 1002
- 被引次数: 0