Locating sensors with fuzzy logic algorithms

Abstract

In a system formed by hundreds of sensors deployed in a huge area it is important to know the position where every sensor is.This information can be obtained using several methods. However, if the number of sensors is high and the deployment is based on ad-hoc manner, some auto-locating techniques must be implemented.In this paper we describe a novel algorithm based on fuzzy logic with the objective of estimating the location of sensors according to the knowledge of the position of some reference nodes.This algorithm, called LIS (Localization based on Intelligent Sensors) is executed distributively along a wireless sensor network formed by hundreds of nodes, covering a huge area.The evaluation of LIS is led by simulation tests. The result obtained shows that LIS is a promising method that can easily solve the problem of knowing where the sensors are located.

Gaussian Process Regression Plus Method for Localization Reliability Improvement

“…The optimal estimation for $x_{*}$ is the mean of this distribution $\stackrel{true¯}{x_{*}}=K_{*}{K}^{-1}x$ [20,21], and the uncertainty in the estimation is the variance $\text{var}\left(x_{*}\right)=K_{**}-K_{*}{K}^{-1}{K}_{*}^T$. The aforementioned description indicates that the GPR is suitable for predicting unknown data with limited known data.…”

Location Determination of Dynamically Moving Mobile Nodes in Real Time

“…Infrastructure bound wired solutions do have an advantage in terms of security over wireless ones, due the inevitable physical connection to the target network [1][2][3][4][5]. The possibility to determine the physical location of any connected client would result in a significant gain of the achievable security level.…”

A new fuzzy location indicator for Interval Type-2 indoor Fuzzy Localization System

“…Depends on the used localization methods [3]: based on Time of Arrival (ToA), Angle of Arrival (AoA), Radio Signal Strenght (RSS) or Time Difference of Arrival (TDoA), localization techniques merged from classical and probabilistic approaches [7] to Evolutionary computation (EC) using Neural networks (NN) [8], Support Vector Machines (SVM) [9], Fuzzy logic (FL) [10,11,12] and hybrid approaches.…”