Target location estimation in sensor networks using range information

Abstract

We consider the problem of target location estimation in the context of large scale, dense sensor networks. We model the probability of detection in each sensor, p d , as a function of the distance between the sensor and t h e target. Based on a binary (dctection vs. no detection) information from each sensor and the model of p d , we propose two different fusion rules for estimating the target location: a maximum likelihood estimate and an empirical risk minimization method. Moreover, we also consider the case where only sensors with a positive detection transmit their reading. T h i s can be helpful to economize the power of sensor units. By employing gaussian like pd models, we develop versions of both methods based on simple initialization procedures and ; I gradient search. We compare and discuss both algorithms in terms of complexity and accuracy.

Algorithms for the selection of the active sensors in distributed tracking: comparison between Frisbee and GNS methods

“…Let di E {0,1} be the result of the i-th node in the active set (1 if detection occurs, 0 otherwise), and D represent the agglomeration of the Na nodes, i.e., D = (d1,...dN). The likelihood that the target is located at (x,y) is given by (see also [5] Ri(X,y)= _(-x) +( yi)2…”

Performance Characterization of Random Proximity Sensor Networks

“…For all simulations, P F A = 0.001, SN R 1 = 67.6db, λ = 2, μ = 1, and the deployment area is 1 km 2 . The purpose of the experiments is to validate the bound for connectivity given by (10) and the ability of the formula for the mean-squared error given by (20) to predict the performance of centroid estimator and the maximum likelihood estimator (MLE) algorithm proposed in [9]. We ran one-thousand Monte-Carlo simulations fixing = 0.03 and empirically calculated that the smallest constant c in (10) that guarantees the network is connected with probability larger than 97% is c ≈ 1.5166.…”

Connectivity for the Frisbee Architecture

“…As in [10], simple sensors have only the capability of sensing their coverage area, compute binary information and transmit data to complex sensors. Binary information is encoded by a 1 if sensor detects something crossing its coverage area and by a 0 otherwise.…”

“…Binary information is encoded by a 1 if sensor detects something crossing its coverage area and by a 0 otherwise. Complex sensors, instead, have computation capabilities; they are able to locate the target by applying the maximum likelihood estimation algorithm described in [ 10].…”

“…* Simple sensors monitor their coverage area, and when they detect the presence of a target, awake close complex sensors, transmit their binary information to them and fall asleep * Complex sensors collect data from simple sensors, process their information as in [10] and compute the estimated position of the target. Finally they choose which complex sensors will be used to detect the target at the next step and activate a mechanism to awake them.…”