Study of Signal Transmission and Equalization in Wireless Underground Sensor Communication Networks

Abstract

In this thesis, modeling of the channel is segregated into three stages. First stage involves newlinethe study of electrical and magnetic properties, i.e. permittivity and permeability of soil newlinemedium. Modeling these properties is crucial as it governs the propagation of EM waves newlinethrough a medium. Second stage involves study of various paths traced by the transmitted newlinesignal such as direct wave and reflected wave. Final stage consists of parametric analysis newlineof path loss and BER with variables as inter-node distance, depth of sensor nodes and soil newlineparameters. However, the primary concern in wireless communication is Inter Symbol newlineInterference (ISI) which causes a detrimental effect to the reliability of the newlinecommunication link. To mitigate the effects of ISI, application of different equalization newlinealgorithms such as Least Mean Square (LMS), Recursive Least Square (RLS) and newlineNormalized Least Mean Square (NLMS) can be done at the receiver end. newline newlineIn this thesis, underground-underground (UG-UG) channel has been characterized using newlinethe Mineralogy Based Spectroscopic Dielectric Model (MBSDM) which takes into newlineconsideration the varying soil parameters such as volumetric water content, burial depth, newlinesink antenna height, sensor density, clay and sand percentage and operating frequency of newlinethe transmitted signal. The channel coefficients were then obtained using the two ray newlinemodel. Although the underground channel is relatively stable with respect to time, a newlinelocation dependent Rayleigh factor is considered to accommodate for the varying newlinereceived signal levels along each of the multipath. The data signal was modulated using newlineBinary Phase Shift Keying (BPSK), Quadrature Phase Shift Keying (QPSK) and newlineM-Quadrature Amplitude Modulation (QAM). The simulation of these stages is done in newlineMATLAB for UG-UG, UG-AG and AG-UG scenarios. The received signal corrupted newlinedue to noise and interference was equalized using LMS, RLS or NLMS algorithms to newlinecancel the channel effects.