Signal Processing for Sensor Networks With Compressive Sampling
Loading...
Date
item.page.authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
ABSTRACT
newlineSensor Networks (SNs) are becoming an inevitable part of the processes in almost all the
newlinefields. Networking a large number of varied sensors to monitor, collect, disseminate and
newlineintelligently combine information about a specific task is the emerging paradigm of the era.
newlineThe flexibility, fault tolerance, high sensing fidelity, low cost and rapid deployment characteristics
newlineof SNs create many new and exciting application areas for them. As the usage and
newlineapplication areas are becoming wider, the design and implementation of SNs face variety
newlineof challenges. The conventional signal processing methods in SNs are based on the Nyquist
newlineSampling theorem, which requires sampling rate at least twice the message signal bandwidth
newlinein order to achieve exact recovery. This gigantic amount of data received from these sensors
newlinemust be processed to estimate, detect and predict various phenomena, which needs huge
newlinebandwidth, processing power, storage space etc. The huge data generated by these systems
newlineare creating overload for the systems of processing, communicating, storage, power and control.
newlineThis area needs a lot of contributions to cover up the issues to have properly oriented
newlineand well functioning networks.
newlineCompressive Sensing (CS) is of interest in the scenario where sampling by Nyquist
newlinerate is not efficient. CS can find sparse solutions to underdetermined linear systems and can
newlinereconstruct the signals from far fewer samples than suggested by the Nyquist rate. The research
newlinefocuses on the field ofWireless Sensor Networks (WSNs). After a detailed study and
newlineimplementation of a typical WSN, it investigates the signal processing activities and challenges
newlineinvolved in WSN. The research tries to elucidate the drawbacks of traditional signal
newlineprocessing methods. The data deluge problem generated by the SNs has been portrayed to
newlineunderstand the critical situation pertaining in the field.
newlineThis research is an attempt to redefine the WSNs with the new concept of CS, in which
newlinethe sparsity of the signals can be exploited to avoid the wastage of resources. The traditional
newlinemethods are modified and the regeneration of sensor data with lesser samples than Nyquist
newlinerate has been accomplished with the CS methods. Along with this, the potential of CS to
newlinesolve the issues mentioned for the WSNs that aroused as a result of data deluge due to high
newlinesampling rate, is examined. The capability of CS has been proved with different performance
newlinemeasures and concluded that CS can effectively solve the existing issues in SNs.
newlineiii
newline