{[ promptMessage ]}

Bookmark it

{[ promptMessage ]}

Target Localization using Ensemble Support Vector

Target Localization using Ensemble Support Vector - This...

Info iconThis preview shows pages 1–2. Sign up to view the full content.

View Full Document Right Arrow Icon
Target Localization using Ensemble Support Vector Regression in Wireless Sensor Networks Woojin Kim, Jaemann Park, and H. Jin Kim School of Aerospace and Mechanical Engineering Seoul National University, Seoul, Korea 151-744 Email: [email protected] Abstract —This paper considers a target localization problem whose goal is to estimate the location of an unknown object. It is one of the key issues in applications of wireless sensor net- works (WSNs). With recent advances in fabrication technology, deployment of a large WSNs has become economically feasible. On the other hand, this has caused the curse of dimensionality in applying learning algorithms such as support vector regression (SVR). To handle this, we use an ensemble implementation of SVRs for target localization and validate it experimentally. This paper draws a comparison between the conventional SVR method and the proposed method in terms of the accuracy and robustness. Experimental results show that the prediction performance of the proposed method is more accurate and robust to the measurement noise than conventional SVR predictor. I. I NTRODUCTION Dramatic advances in communications and MEMS technol- ogy have enabled the use of distributed sensor nodes [1]. Wire- less sensor networks consist of a large number of scattered sensor nodes, which can measure miscellaneous modalities of their local environment and communicate with their neighbors. The wireless sensor networks can be viewed as a distributed system, in which each node makes measurements, processes the raw data and transmits the acquired information. WSNs have been considered for various monitoring and control applications such as target detection, target recognition, target localization [2], biomedical applications [3], structure health monitoring applications [4], and home and office ap- plications [5]. These applications have a common issue of efficiently processing the distributed information gathered. In this paper, we discuss one of the key issues in ap- plications of WSNs, target localization. The goal of target localization is to estimate the location of an unknown ob- ject. Many practical implementations of sensor networks use acoustic feature, and we consider a sensor network consisting of acoustic sensor nodes, whose signal source comes from a target. In order to estimate precise target location, many chal- lenges for target localization have been addressed with various approaches. Various target localization techniques in sensor networks are proposed, such as time difference of arrival (TDOA) [6], received signal strength indication (RSSI) [7], binary-detecting model [8], and learning theory [9]. The methods based on TDOA require heavy traffic load of com- munications and consume much energy to maintain time synchronization. The performance of the RSSI-based methods depends on the sensitivity of sensors, which may not be applicable for low-cost sensor networks. For low-cost and low-capability sensor networks, the methods based on binary-
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
Image of page 2
This is the end of the preview. Sign up to access the rest of the document.

{[ snackBarMessage ]}