Scholarlink

<p>Aftermath of disaster, efficient and accurate localization of victims and resources is crucial for effective rescue operations. Traditional localization algorithms face challenges posed by environmental noise, multipath reflection, and various uncertainties, all of which can compromise range accuracy and therefore, affect the localization precision. This paper addresses the localization problem by leveraging machine learning techniques for a disaster area network. We first generate a set of Received Signal Strength Indicator (RSSI) values collected by Anchor Nodes (ANs) for signals received from Target Nodes (TNs). Using the RSSI data collected from victims, we train machine learning models to accurately estimate the locations of individuals with maximum coverage. In this paper, we investigate the influence of placing the anchor nodes in a grid pattern as well as randomly distributing these nodes throughout the deployment area. We examine the effect on various network parameters with varying number of anchor nodes and target nodes. Through simulation, we investigate the performance of machine learning algorithms including Support Vector Regression (SVR), Random Forest (RF), k-Nearest Neighbor (k-NN), and Gradient Boosting (GB) with optimization-based techniques such as Genetic Algorithm (GA) and Particle Swarm Optimization (PSO) using Received Signal Strength Indicator (RSSI) in terms of Root Mean Square Error (RMSE), Mean Localization Error (MLE) and Node Localization Efficiency (NLE).</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Localization of Victims in Disaster Area Using RSSI with Machine Learning Techniques

  • 期刊论文

  • 发表日期：

    2025年4月15日

查看全文
  • Bidyarani Langpoklakpam,
  • Lithungo K. Murry

摘要

Aftermath of disaster, efficient and accurate localization of victims and resources is crucial for effective rescue operations. Traditional localization algorithms face challenges posed by environmental noise, multipath reflection, and various uncertainties, all of which can compromise range accuracy and therefore, affect the localization precision. This paper addresses the localization problem by leveraging machine learning techniques for a disaster area network. We first generate a set of Received Signal Strength Indicator (RSSI) values collected by Anchor Nodes (ANs) for signals received from Target Nodes (TNs). Using the RSSI data collected from victims, we train machine learning models to accurately estimate the locations of individuals with maximum coverage. In this paper, we investigate the influence of placing the anchor nodes in a grid pattern as well as randomly distributing these nodes throughout the deployment area. We examine the effect on various network parameters with varying number of anchor nodes and target nodes. Through simulation, we investigate the performance of machine learning algorithms including Support Vector Regression (SVR), Random Forest (RF), k-Nearest Neighbor (k-NN), and Gradient Boosting (GB) with optimization-based techniques such as Genetic Algorithm (GA) and Particle Swarm Optimization (PSO) using Received Signal Strength Indicator (RSSI) in terms of Root Mean Square Error (RMSE), Mean Localization Error (MLE) and Node Localization Efficiency (NLE).