No abstract

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 07/02/2015 Terms of Use: http://spiedl.org/terms Proc. of SPIE Vol. 4791 311 Downloaded From: http://proceedings.spiedigitallibrary.org/ on 07/02/2015 Terms of Use: http://spiedl.org/terms

Wireless sensor networks have been attracting increasing research interest given the recent advances in microelectronics, array processing, and wireless networking. Consisting of a large collection of small, wireless, low-cost, integrated sensing, computing, and communicating nodes capable of performing various demanding collaborative space-time processing tasks, wireless sensor network technology poses various unique design challenges, particularly for real-time operation. In this paper, we review the Approximate Maximum-Likelihood (AML) method for source localization and direction-of-arrival (DOA) estimations. Then, we consider the use of least-squares (LS) method applied to DOA bearing crossings to perform source localization. A novel virtual array model applicable to the AML-DOA estimation method is proposed for reverberant scenarios. Details on the wireless acoustical testbed are given. We consider the use of Compaq iPAQ 3760s, which are handheld, battery-powered device normally meant to be used as personal organizers (PDAs), as sensor nodes. The iPAQ provide a reasonable balance of cost, availability, and functionality. It has a build-in StrongARM processor, microphone, codec for acoustic acquisition and processing, and a PCMCIA bus for external IEEE 802.11b wireless cards for radio communication. The iPAQs form a distributed sensor network to perform real-time acoustical beamforming. Computational times and associated real-time processing tasks are described. Field measured results for linear, triangular, and square subarrays in free-space and reverberant scenarios are presented. These results show the effective and robust operation of the proposed algorithms and their implementations on a real-time acoustical wireless testbed.

Abstract. We propose to use the Approximate Maximum-Likelihood (AML) method to estimate the direction-of-arrival (DOA) of multiple targets from various spatially distributed sub-arrays, with each sub-array having multiple acoustical/seismic sensors. Localization of the targets can with possibly some ambiguity be obtained from the cross bearings of the sub-arrays. Spectra from the AML-DOA estimation of the target can be used for classification as well as possibly to resolve the ambiguity in the localization process. We use the Support Vector Machine (SVM) supervised learning method to perform the target classification based on the estimated target spectra. The SVM method extends in a robust manner to the nonseparable data case. In the learning phase, classifier hyperplanes are generated off-line via a primal-dual interior point method using the training data of each target spectra obtained from a single acoustical/seismic sensor. In the application phase, the classification process can be performed in real-time involving only a simple inner product of the classifier hyperplane with the AML-DOA estimated target spectra vector. Analysis based on Cramér-Rao bound (CRB) and simulated and measured data is used to illustrate the effectiveness of AML and SVM algorithms for wideband acoustical/seismic target DOA, localization, and classification.

No abstract