Mahesh Vemula

Target Tracking by Particle Filtering in Binary Sensor Networks

We present particle filtering algorithms for tracking a single target using data from binary sensors. The sensors transmit signals that identify them to a central unit if the target is in their neighborhood; otherwise they do not transmit anything. The central unit uses a model for the target movement in the sensor field and estimates the targets trajectory, velocity, and power using the received data. We propose and implement the tracking by employing auxiliary particle filtering and cost-reference particle filtering. Unlike auxiliary particle filtering, cost-reference particle filtering does not rely on any probabilistic assumptions about the dynamic system. In the paper, we also extend the method to include estimation of constant parameters, and we derive the posterior Cramer-Rao bounds (PCRBs) for the states. We show the performances of the proposed methods by extensive computer simulations and compare them to the derived bounds.

Signal processing by particle filtering for binary sensor networks

In a wireless sensor network, limited power, communication, and computational resources are the major constraints that have to be overcome for their successful deployment and utilization. Binary sensor networks are a class of networks that get around these constraints. There, the sensors transmit only a binary digit on the occurrence of the event of interest, and therefore, the signals that reach the fusion center of these networks are highly compressed and pose challenging problems for recovering the sensed information. We consider the problem of tracking a vehicle, which moves along a 2-dimensional space, by using a binary sensor network that fuses information by particle filtering.

Sensor self-localization with beacon position uncertainty

We propose algorithms for distributed sensor self-localization using beacon nodes. These beacon nodes broadcast some information which describes their positions. The sensor nodes with unknown location information utilize these descriptions along with the characteristics of received signals to obtain estimates of their positions. Sensors with resolved positions, in the successive stages of the algorithm also broadcast their location information to other sensors so that they can resolve their own positions. Conditional upon the availability of probabilistic distributions of noise processes, we propose iterative and Monte Carlo sampling-based methods for obtaining sensor location descriptions. We also provide approximate hybrid Cramer-Rao bounds for distributed sensor self-localization and compare them with the proposed algorithms. We demonstrate the performance of the proposed algorithms through extensive computer simulations.

Tracking with particle filtering in tertiary wireless sensor networks

Recent advances of wireless sensor networks have presented some very interesting problems for signal processing. For practical reasons, many networks are composed of simple sensors that use very little power and do not consume much communication bandwidth. A class of sensors that satisfy these requirements are the tertiary sensors. They report an approaching event with one signal and a receding event with another signal. When the event is out of their range, they do not report anything. In this paper, we apply particle filtering for processing signals from tertiary sensor networks with the purpose of tracking events (targets) within the field of the sensor network. We present an algorithm for tracking and demonstrate its performance by computer simulations.

Target Tracking in a Two-Tiered Hierarchical Sensor Network

An important application of sensor networks is target tracking and localization. To deal with sensor nodes with limited energy supply and communication bandwidth we propose energy-efficient hierarchical architectures for solving the target tracking problem. In these networks, sensors form clusters and transmit minimal quantized information about a sensed event to a specialized node, known as a cluster head. Cluster heads are equipped with capability of communicating over large distances with a fusion center or a base station. We consider two different hierarchical architectures: (a) the target dynamics are probabilistically estimated at the cluster heads and their statistics combined at the fusion center, and (b) the cluster heads perform simple compression rules on the quantized sensor data and the fusion center estimates the target dynamics using these severely compressed data. Sequential Monte Carlo algorithms for estimation of the target dynamics are used. Through computer simulations the performances of these two architectures are studied

Performance Comparison of Gaussian-Based Filters Using Information Measures

In many situations, solutions to nonlinear discrete-time filtering problems are available through approximations. Many of these solutions are based on approximating the posterior distributions of the states with Gaussian distributions. In this letter, we compare the performance of Gaussian-based filters including the extended Kalman filter, the unscented Kalman fitter, and the Gaussian particle filter. To that end, we measure the distance between the posteriors obtained by these filters and the one estimated by a sequential Monte Carlo (particle filtering) method. As a distance metric, we apply the Kullback-Leibler and x2 information measures. Through computer simulations, we rank the performance of the three filters.

Particle Filtering-Based Target Tracking in Binary Sensor Networks Using Adaptive Thresholds

Target tracking in wireless sensor networks with constrained resources is a challenging problem. In this paper we consider scenarios where sensors sense an object of interest and process the received measurements using adaptive thresholds to obtain quantized data in the form of two levels. The data are quantized to address resource constraints in sensor networks. The processed data are then sent to a fusion center which resolves the tracking problem by means of a particle filter which can handle non-linearities in the state model. The performance of various strategies for threshold adaptation is studied by computer simulations and the results reveal that improvement is obtained over scenarios with fixed thresholds.

Target tracking by fusion of random measures

In this paper we propose fusion methods for tracking a single target in a sensor network. The sensors use sequential Monte Carlo (SMC) techniques to process the received measurements and obtain random measures of the unknown states. We apply standard particle filtering (SPF) and cost-reference particle filtering (CRPF) methods. For both types of filtering, the random measures contain particles drawn from the state space. Associated to the particles, the SPF has weights representing probability masses, while the CRPF has user-defined costs measuring the quality of the particles. Summaries of the random measures are sent to the fusion center which combines them into a global summary. Similarly, the fusion center may send a global summary to the individual sensors that use it for improved tracking. Through extensive simulations and comparisons with other methods, we study the performance of the proposed algorithms.

Cost-Based Monte Carlo Sampling Approaches for Sensor Self-Localization Under Beacon Position Uncertainty

Sensor localization methods based on Monte Carlo sampling approximate the sensor position distributions by a weighted set of samples. These approaches traditionally require complete knowledge of the probabilistic distributions of the uncertainties in the sensor system. In this paper, we propose alternative sampling-based methods which do not require complete knowledge of the probabilistic distributions. The sensor position distributions are represented by a set of samples and costs which are described by spatial parametric regions. Few parameters are needed to characterize these regions, and therefore the amount of information to be transmitted to the rest of the sensors to self-localize is simplified. Computer simulations show that the proposed methods are more robust and less computationally intensive than standard sampling approaches.

On Proposal Functions for Cost-Reference Particle Filtering

Standard particle filtering (SPF) schemes rely on the availability of probability distributions of the state and observation noises involved in the dynamic state space model. Cost reference particle filtering (CRPF) techniques have proven to be a viable and robust alternative in situations when the probability distributions of these noise processes are unknown. In this paper, we propose two new CRPF methods which use different proposal functions from the one of the original CRPF method. The proposed algorithms are applied to target tracking in a wireless sensor network. The performance of the proposed methods is demonstrated by computer simulations