Collaborative Signal Processing

Wireless Sensor Networks consist of a powerful technology for monitoring the physical world. Particularly, innetwork data fusion techniques are very important to applications such as target classification and tracking to reduce the communication burden in these constrained networks. However, the efficiency of the solution can be affected by the data correlation among several sensor nodes. Thus, the application of value fusion (for clusters of nodes with correlated measurements) and decision fusion (combining the local decisions of the clusters) is a common strategy. In this work, we propose an algorithm for properly selecting the groups of nodes with correlated measurements. Experiments show that our algorithm is 30% better than a solution that considers only the spatial coherence regions.

We outline a framework for collaborative signal processing in distributed sensor networks. The ideas are presented in the context of tracking multiple moving objects in a sensor field. The key steps involved in the tracking procedure include event detection, target classification, and estimation/prediction of target location. Algorithms for various tasks are discussed and some recent results on experiments with real data are reported. Directions for ongoing and future research are discussed.

In many practical signal detection problems, the detectors have to designed from training data. Due to limited training data, which is usually the case, it is imperative to exploit some inherent signal structure for reliable detector design. The signals of interest in a variety of applications manifest such structure in the form of nuisance parameters. However, data-driven design of detectors by exploiting nuisance parameters is virtually impossible in general due to two major di culties: identifying the appropriate nuisance parameters, and estimating the corresponding detector statistics. We address this problem by using recent results that relate joint signal representations (JSRs), such as time-frequency and timescale representations, to quadratic detectors for a wide variety of nuisance parameters. We propose a general data-driven framework that: 1) identi es the appropriate nuisance parameters from an arbitrarily chosen nite set, and 2) estimates the second-order statistics that characterize the corresponding JSR-based detectors. Simulation results demonstrate that for limited training data, exploiting the structure of nuisance parameters via our framework can deliver substantial gains in performance over empirical detectors which ignore such structure.

Sensor networks provide virtual snapshots of the physical world via distributed wireless nodes that can sense in different modalities, such as acoustic and seismic. Classification of objects moving through the sensor field is an important application that requires collaborative signal processing (CSP) between nodes. Two main forms of CSP are possible. Data fusion-exchange of low dimensional feature vectors-is needed between correlated nodes, in general, for optimal performance. Decision fusion-exchange of likelihood values-is sufficient between independent nodes. Decision fusion is generally preferable due to its lower communication burden. We study CSP of multiple node measurements in the context of single target classification. Each measurement is modeled as a Gaussian signal vector (corresponding to the target class) corrupted by additive white Gaussian noise. The measurements are partitioned into groups. The signal components within each group are perfectly correlated whereas they vary independently between groups. Three classifiers are compared: the optimal maximum likelihood classifier, a data averaging classifier that treats all measurements as correlated, and a decision fusion classifier that treats them all as independent. Analytical and numerical results based on real data are provided to compare the performance of the three CSP classifiers. Our results indicate that the sub-optimal decision fusion classifier, that is most attractive in the context of sensor networks, is also a robust choice from a decision theoretic viewpoint.

Sensor networks provide virtual snapshots of the physical world via distributed wireless nodes that can sense in different modalities, such as acoustic and seismic. Classification of objects moving through the sensor field is an important application that requires collaborative signal processing (CSP) between nodes. Given the limited resources of nodes, a key constraint is to exchange the least amount of information between them to achieve desired performance. Two main forms of CSP are possible. Data fusion-exchange of low dimensional feature vectorsis needed between correlated nodes, in general, for optimal performance. Decision fusion-exchange of likelihood values-is sufficient between independent nodes. Decision fusion is generally preferable due to its lower communication and computational burden. We study CSP of multiple node measurements for classification, each measurement modeled as a Gaussian (target) signal vector corrupted by additive white Gaussian noise. The measurements are partitioned into groups. The signal components within each group are perfectly correlated whereas they vary independently between groups. Three classifiers are compared: the optimal maximum-likelihood classifier, a data-averaging classifier that treats all measurements as correlated, and a decision-fusion classifier that treats them all as independent. Analytical and numerical results based on real data are provided to compare the performance of the three CSP classifiers. Our results indicate that the sub-optimal decision fusion classifier, that is most attractive in the context of sensor networks, is also a robust choice from a decision-theoretic viewpoint.

The highly distributed infrastructure provided by sensor networks supports fundamentally new ways of designing surveillance systems. In this paper, we discuss sensor networks for target classification and tracking. Our formulation is anchored on location-aware data routing to conserve system resources, such as energy and bandwidth. Distributed classification algorithms exploit signals from multiple nodes in several modalities and rely on prior statistical information about target classes. Associating data to tracks becomes simpler in a distributed environment, at the cost of global consistency. It may be possible to filter clutter from the system by embedding higher level reasoning in the distributed system. Results and insights from a recent field test at 29 Palms Marine Training Center are provided to highlight challenges in sensor networks.

We outline a framework for collaborative signal processing in distributed sensor networks. The ideas are presented in the context of tracking multiple moving objects in a sensor field. The key steps involved in the tracking procedure include event detection, target classification, and estimation and prediction of target location. Algorithms for various tasks are discussed with an emphasis on classification. Results based on experiments with real data are reported which provide useful insights into the essential nature of the problems. Issues, challenges and directions for future research are identified.

Ground vehicle detection and classification with distributed sensor networks is of growing interest for border security. Different sensing modalities including electro-optical, seismic, and acoustic were evaluated individually and in combination to develop a more efficient system. Despite previous works that mostly studied frequency-domain features and acoustic sensors, in this work we analyzed the classification performance for both frequency and time-domain features and seismic and acoustic modalities. Despite their infrequent use, we show that when fused with frequency-domain features, time-domain features improve the classification performance and reduce the false positive rate, especially for seismic signals. We investigated the performance of seismic sensors and showed that the classification performance varies with the type of road due to the distinct spectral characteristics of the medium. Our proposed classifier fuses time and frequency-domain features and acoustic and seis...

Ground vehicle detection and classification with distributed sensor networks is of growing interest for border security. Different sensing modalities including electro-optical, seismic, and acoustic were evaluated individually and in combination to develop a more efficient system. Despite previous works that mostly studied frequency-domain features and acoustic sensors, in this work we analyzed the classification performance for both frequency and time-domain features and seismic and acoustic modalities. Despite their infrequent use, we show that when fused with frequency-domain features, time-domain features improve the classification performance and reduce the false positive rate, especially for seismic signals. We investigated the performance of seismic sensors and showed that the classification performance varies with the type of road due to the distinct spectral characteristics of the medium. Our proposed classifier fuses time and frequency-domain features and acoustic and seis...

Sensor networks provide virtual snapshots of the physical world via distributed wireless nodes that can sense in different modalities, such as acoustic and seismic. Classification of objects moving through the sensor field is an important application that requires collaborative signal processing (CSP) between nodes. Two main forms of CSP are possible. Data fusion-exchange of low dimensional feature vectors-is needed between correlated nodes, in general, for optimal performance. Decision fusion-exchange of likelihood values-is sufficient between independent nodes. Decision fusion is generally preferable due to its lower communication burden. We study CSP of multiple node measurements in the context of single target classification. Each measurement is modeled as a Gaussian signal vector (corresponding to the target class) corrupted by additive white Gaussian noise. The measurements are partitioned into groups. The signal components within each group are perfectly correlated whereas they vary independently between groups. Three classifiers are compared: the optimal maximum likelihood classifier, a data averaging classifier that treats all measurements as correlated, and a decision fusion classifier that treats them all as independent. Analytical and numerical results based on real data are provided to compare the performance of the three CSP classifiers. Our results indicate that the sub-optimal decision fusion classifier, that is most attractive in the context of sensor networks, is also a robust choice from a decision theoretic viewpoint.

Sensor networks provide virtual snapshots of the physical world via distributed wireless nodes that can sense in different modalities, such as acoustic and seismic. Classification of objects moving through the sensor field is an important application that requires collaborative signal processing (CSP) between nodes. Given the limited resources of nodes, a key constraint is to exchange the least amount of information between them to achieve desired performance. Two main forms of CSP are possible. Data fusion-exchange of low dimensional feature vectorsis needed between correlated nodes, in general, for optimal performance. Decision fusion-exchange of likelihood values-is sufficient between independent nodes. Decision fusion is generally preferable due to its lower communication and computational burden. We study CSP of multiple node measurements for classification, each measurement modeled as a Gaussian (target) signal vector corrupted by additive white Gaussian noise. The measurements are partitioned into groups. The signal components within each group are perfectly correlated whereas they vary independently between groups. Three classifiers are compared: the optimal maximum-likelihood classifier, a data-averaging classifier that treats all measurements as correlated, and a decision-fusion classifier that treats them all as independent. Analytical and numerical results based on real data are provided to compare the performance of the three CSP classifiers. Our results indicate that the sub-optimal decision fusion classifier, that is most attractive in the context of sensor networks, is also a robust choice from a decision-theoretic viewpoint.

In many practical signal detection problems, the detectors have to designed from training data. Due to limited training data, which is usually the case, it is imperative to exploit some inherent signal structure for reliable detector design. The signals of interest in a variety of applications manifest such structure in the form of nuisance parameters. However, data-driven design of detectors by exploiting nuisance parameters is virtually impossible in general due to two major di culties: identifying the appropriate nuisance parameters, and estimating the corresponding detector statistics. We address this problem by using recent results that relate joint signal representations (JSRs), such as time-frequency and timescale representations, to quadratic detectors for a wide variety of nuisance parameters. We propose a general data-driven framework that: 1) identi es the appropriate nuisance parameters from an arbitrarily chosen nite set, and 2) estimates the second-order statistics that characterize the corresponding JSR-based detectors. Simulation results demonstrate that for limited training data, exploiting the structure of nuisance parameters via our framework can deliver substantial gains in performance over empirical detectors which ignore such structure.

The highly distributed infrastructure provided by sensor networks supports fundamentally new ways of designing surveillance systems. In this paper, we discuss sensor networks for target classification and tracking. Our formulation is anchored on location-aware data routing to conserve system resources, such as energy and bandwidth. Distributed classification algorithms exploit signals from multiple nodes in several modalities and rely on prior statistical information about target classes. Associating data to tracks becomes simpler in a distributed environment, at the cost of global consistency. It may be possible to filter clutter from the system by embedding higher level reasoning in the distributed system. Results and insights from a recent field test at 29 Palms Marine Training Center are provided to highlight challenges in sensor networks.

We outline a framework for collaborative signal processing in distributed sensor networks. The ideas are presented in the context of tracking multiple moving objects in a sensor field. The key steps involved in the tracking procedure include event detection, target classification, and estimation/prediction of target location. Algorithms for various tasks are discussed and some recent results on experiments with real data are reported. Directions for ongoing and future research are discussed.

We outline a framework for collaborative signal processing in distributed sensor networks. The ideas are presented in the context of tracking multiple moving objects in a sensor field. The key steps involved in the tracking procedure include event detection, target classification, and estimation and prediction of target location. Algorithms for various tasks are discussed with an emphasis on classification. Results based on experiments with real data are reported which provide useful insights into the essential nature of the problems. Issues, challenges and directions for future research are identified.

Distributed sensing has been used for enhancing signal to noise ratios for space-time localization and tracking of remote objects using phased array antennas, sonar, and radio signals. The use of these technologies in identifying mobile targets in a field, emitting acoustic signals, using a network of low-cost narrow band acoustic micro-sensing devices randomly dispersed over the region of interest, presents unique challenges. The effects of wind, turbulence, and temperature gradients and other environmental effects can decrease the signal to noise ratio by introducing random errors that cannot be removed through calibration. This paper presents methods for dynamic distributed signal processing to detect, identify, and track targets in noisy environments with limited resources. Specifically, it evaluates the noise tolerance of adaptive beamforming and compares it to other distributed sensing approaches. Many source localization and direction-of-arrival (DOA) estimation methods based...

A new computing paradigm which utilizes mobile agents to carry out collaborative target classification in distributed sensor networks is presented in this paper. Instead of each sensor sending local classification results to a processing center where the fusion process is taken place, a mobile agent is dispatched from the processing center and the fusion process is executed at each sensor node. The advantage of using mobile agent is that it achieves progressive accuracy and is task-adaptive. To improve the accuracy of classification, we implement Behavior Knowledge Space method for multi-modality jhsion. W e also modified the classical k-nearest-neighbor method to be adaptive to collaborative classification in a distributed network of sensor nodes. Experimental results based on a field demo are presented at the end of the paper.

Target classification using distributed sensor arrays remains a challenging problem due to the non-stationarity of target signatures, large geographical area coverage of sensor arrays, and the requirements of time-critical and reliable information delivery. In this paper, we develop an algorithm to derive effective and stable features from both the frequency and the time-frequency domains of the acoustic signals. A modified data fusion algorithm for distributed sensor arrays is also developed in order to integrate the classification results from different sensors and provide fault-tolerance. By using data fusion, the accuracy of the classification can be increased by as many as 50%.

Distributed sensor networks are a significant technology nowadays. Inexpensive, smart devices with multiple sensors provide opportunities for instrumenting, monitoring and controlling targeting systems. Such sensor nodes have capability for acquiring and embeddedprocessing of variety of data forms. Collaborative signal processing and fusion algorithms are needed to aggregate the distributed data from among the nodes in the network, including possibly multiple modalities of data within a sensor node, to make decisions in a reliable and efficient manner. One of the important sensor network applications is target classification in battlefields. This paper presents improved moving vehicle target classification performance using data obtained from sensor networks with collaboration both across nodes and within a node in terms of multimodal fusion. Results show that a 50% relative improvement in classification error can be obtained using collaboration both in the case of single vehicle target and those involving multi-vehicle convoys. from among the nodes in the network, including possibly multiple modalities of data within a sensor node, to make decisions in a reliable and efficient manner. Even though problems in the fields of array signal processing and data fusion have been studied for a number of years, advances in sensor technologies, especially those aimed at military applications, have created new scenarios for applying signal processing ideas. One of the important sensor network applications is target classification in battlefields e.g. identifying types of moving vehicles in a field. Recently, this specific application has been discussed in the context of single sensor node processing [4] of single vehicle targets. This paper presents improved performance of classification algorithms applied over the sensor network with collaboration within a node, in terms of multimodal fusion, and across nodes in two different approaches i.e. data sharing and statistical confidence boost techniques. Proposed classification algorithms are also applied to the multiple targets scenarios i.e., target involving multiple vehicles in a convoy. The emphasis is on experiments using real seismic and acoustic. data collected from the SITEXOO experiments performed as a part of the DARPA SensIT program [9].

Heterogeneous sensor networks (HSNs) with multiple sensing modalities are gaining popularity in diverse fields. Tracking is an application that can benefit from multiple sensing modalities. If a moving target emits sound then both audio and video sensors can be utilized. These modalities can complement each other in the presence of high background noise that impairs the audio or visual clutter affecting the video. Audio-video tracking can also provide cues for the other modality for actuation. In this paper, we describe an approach for target tracking in urban environments utilizing an HSN of mote class devices equipped with acoustic sensor boards and embedded PCs equipped with web cameras. Our system employs a Markov Chain Monte Carlo Data Association algorithm for tracking vehicles emitting engine noise. Experimental results from a deployment in an urban environment are used to demonstrate our approach.

Distributed sensing has been used for enhancing signal to noise ratios for space-time localization and tracking of remote objects using phased array antennas, sonar, and radio signals. The use of these technologies in identifying mobile targets in a field, emitting acoustic signals, using a network of low-cost narrow band acoustic micro-sensing devices randomly dispersed over the region of interest, presents unique challenges. The effects of wind, turbulence, and temperature gradients and other environmental effects can decrease the signal to noise ratio by introducing random errors that cannot be removed through calibration. This paper presents methods for dynamic distributed signal processing to detect, identify, and track targets in noisy environments with limited resources. Specifically, it evaluates the noise tolerance of adaptive beamforming and compares it to other distributed sensing approaches. Many source localization and direction-of-arrival (DOA) estimation methods based...

A Virtual Global Bus Active Messaging Protocol for Sensor Webs David Andrews, Joe Evans University of Kansas dandrews@ ittc. ukans. edu, evans@ ittc. ukans. edu Mixed signal, micro-electro-mechanical, and wireless communication technologies have accelerated ...

In this work the author demonstrated a robust and efficient method for implementing Doppler classification through the use of Linear Discriminant Analysis (LDA). LDAs were used to reduce dramatically the data dimensionality and thereby eliminate redundancy and improve the efficiency of the classifier. The performance was assessed on a three-class problem of personnel, tracked and wheeled vehicles. Real radar data from a ground based system were used in the design and testing of the classifier. The classifier algorithm was optimised by choosing the best set of features that maximised the performance and the bootstrap method was used to measure the confidence interval. It was shown that only the first few LDA features were relevant. At the very least these were shown to contain information regarding the frequency extent of target Doppler sidebands. The classifier was shown to be robust to changes in target viewing geometry and speed. Overall, good classification was achieved for perso...

SCADDS data diffusion (at USC/ISI) and DRP (at MIT/LL) are both based on the core concept of subject-based routing. Although there are some fundamental differences between these approaches, we believe that both can be accommodated with the same Network Routing ...

In this research work, we present a novel application specific Ordered cell averaging-CFAR (OCA-CFAR) detector for anthropogenic seismic events. We investigate the use of geophones and real-time hardware for the detection of moving targets such as vehicle and human footsteps and activities such as hammering on the ground. We present here energy assisted variants of Constant False Alarm Rate (CFAR) detector implemented on a real-time embedded platform to obtain detection cues of the ground seismic source. The OCA-CFAR algorithm has been modified in terms of window length and position of Cell Under Test (CUT), suitably adapted to the nature of seismic events engendered from moving vehicles and humans. The proposed detector has been compared with the popularly known variants such as Cell Averaging-CFAR, Least Of-CFAR, Greatest Of-CFAR and Ordered Statistics-CFAR for probabilities of false alarm ranging from 1e-1 to 1e-5. The proposed method shows an increased detection rate with minimized false alarms on the field trials and a dataset containing the seismic signature of human activities and some of civilian vehicles.

We studied the possibility of using wireless sensor networks for vehicle identification in a large open field. This is exciting research in that it not only presents a challenge but has practicality. The challenge here is to develop algorithms and/or protocols for sensor nodes to execute a given task. Since each sensor node has limited computation and communication capabilities, these limitations prohibit the use of algorithms and/or protocols developed for conventional computers and networks. Each sensor is dispensable and easily deployable, it can do meaningful work when it is collaborated as a networked cluster; therefore it is very practical in application. Our goal is to identify vehicles in real time using acoustic signal sensors and wireless networks. Our contribution in this paper is three fold. First, we developed a simple vehicle sound identification algorithm enough to be implemented for capacity limited sensor nodes. Second, we proposed architecture and protocols of wireless sensor networks for vehicle identification using this developed sound classification algorithm. Third, we proposed a cooperation model among sensors to expedite the classification process. Our preliminary results show the proposed architecture and protocols are promising.

In this research work, we present a novel application specific Ordered cell averaging-CFAR (OCA-CFAR) detector for anthropogenic seismic events. We investigate the use of geophones and real-time hardware for the detection of moving targets such as vehicle and human footsteps and activities such as hammering on the ground. We present here energy assisted variants of Constant False Alarm Rate (CFAR) detector implemented on a real-time embedded platform to obtain detection cues of the ground seismic source. The OCA-CFAR algorithm has been modified in terms of window length and position of Cell Under Test (CUT), suitably adapted to the nature of seismic events engendered from moving vehicles and humans. The proposed detector has been compared with the popularly known variants such as Cell Averaging-CFAR, Least Of-CFAR, Greatest Of-CFAR and Ordered Statistics-CFAR for probabilities of false alarm ranging from 1e-1 to 1e-5. The proposed method shows an increased detection rate with minimized false alarms on the field trials and a dataset containing the seismic signature of human activities and some of civilian vehicles.

Distributed wireless sensor networks (WSNs) are being proposed for various applications including defense, security, and smart stages. The introduction of hardware wireless sensors in a signal processing education setting can serve as a paradigm for data acquisition, collaborative signal processing, or simply as a platform for obtaining, processing, and analyzing real-life real-time data. In this paper, a software interface that enables the java-digital signal processing (J-DSP) visual programming environment to communicate in a two-way manner with a wireless sensor network is presented. This interface was developed by writing nesC (an extension to the C programming language for sensors) code that enables J-DSP to issue commands to multiple wireless sensor motes, activate specific transducers, and analyze data using any of the existing J-DSP signal processing functions in real time. A series of exercises were developed and disseminated to provide hardware experiences to signals and systems and signal processing undergraduate students. The hardware with the J-DSP software has been used for two semesters in the senior level digital signal processing (DSP) course at Arizona State University. The interface, the exercises, and their assessment (instruments and results) are described in the paper.

MEMS technology has improved such that the capabilities of large sensor devices can now be encompassed in devices that are the size of a penny. These resource constraint devices are confronted with the challenges of ensuring accuracy of observations while conserving power resources. This paper analyzes the two modes of collaborative signal processing employed in sensor networks, namely, data fusion and decision fusion, to achieve greater accuracy. A breakdown of the computation and communication complexity is presented for each fundamental step of data and decision fusion processes, to help in identifying features for possible optimization. We show that a tradeoff exists between the two approaches depending on the underlying application domain.