Geolocation of Narrow Band Radio Frequency Emitter Using RSS based Localization Techniques: The Case of Addis Ababa
No Thumbnail Available
Date
2021-09
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Addis Ababa University
Abstract
Currently, determining the coordinate of fixed or moving objects using propagation features of radio waves such as signal strength, propagation delay, signal phase, and angle of arrival have been a subject of interest in radiodetermination. Radiodetermination can be categorized into radiolocation, ranging, and radio navigation. Ranging is to estimate the distance between the receiver and the transmitter. Radiolocation (geolocation for the object on the earth) predicts the position of an object. Radio navigation encompasses both ranging and radiolocation. Triangulateration and trilateration (range-based) techniques use for radiolocation purposes. Triangulateration technique relies on line of bearing from receiver to emitter incorporating with antenna arrays. Whereas, Trilateration considers received signal strength (radio-wave propagation path loss models) and propagation delay. Following radio wave properties and applied technologies, measurements have been taken by deploying radio wave measurement systems on the ground, or in aircraft.
In this thesis, passive radio frequency (RF) emitter geolocation was analyzed by considering received signal strength (RSS) techniques to compare and analyze the performance of RSS-based geolocation techniques and algorithms. For this purpose, data were gathered using a mobile radio spectrum monitoring unit (receiver), and it was analyzed using software tools such as excel and Python. Radio wave propagation is statistically modeled using the lognormal shadowing model and regressive methods to determine the distance. The results indicate that the error of estimated distance with regressive and clustering approach on average by 96% over the non-clustered approach. With this approach, weighted least square (WLS) estimates location with the error of 170 meters, whereas linear least-squares (LLS) estimate with the error of 206 meters. The lognormal model estimates distance with a minimum error than the regressive method when the standard deviation of data is less than 2dB. At the end, there is a positive correlation between estimated distance and location. It means the error of estimated location increases, as the error of estimated distance increases.
Description
Keywords
Geolocation algorithms, RSS-based geolocation techniques, radio wave measurement, range-based geolocation