Open AccessCompressive sensing for localisation in wireless sensor networks: an approach for energy and error control
Author(s) -
Alwan Nuha A.S.,
Hussain Zahir M.
Publication year - 2018
Publication title -
iet wireless sensor systems
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.433
H-Index - 27
eISSN - 2043-6394
pISSN - 2043-6386
DOI - 10.1049/iet-wss.2016.0112
Subject(s) - wireless sensor network , computer science , real time computing , node (physics) , energy consumption , context (archaeology) , efficient energy use , interpolation (computer graphics) , energy (signal processing) , sampling (signal processing) , sensor node , sample (material) , sensor fusion , wireless , key distribution in wireless sensor networks , wireless network , computer network , telecommunications , engineering , electrical engineering , artificial intelligence , mathematics , statistics , detector , geography , archaeology , chemistry , structural engineering , frame (networking) , chromatography
Energy efficiency is an important requirement in wireless sensor networks in order to achieve cost‐effectiveness and practical implementation. The present work deals with the problem of minimising node power consumption in the context of moving‐node localisation and tracking. Time‐of‐arrival measurements are sent from anchor nodes to a powerful, usually sophisticated, central node, called the fusion centre, where all computations are performed. Low data rates are desirable to economise on node energy but result in sub‐optimal localisation accuracy. It makes sense, therefore, to sample measurements at a low data rate while interpolating the data stream at the fusion centre to improve localisation. The localisation error is remarkably reduced and energy efficiency increased by using this conventional sample rate conversion technique. A further improvement in terms of localisation error is achieved using compressive sensing (via random sampling and interpolation), whereby the localisation error function is shown to decrease with higher‐average random sampling periods.