Performance Analysis of Underwater Visible Light Communication Systems

Abstract

Underwater wireless communication is gaining considerable research interest due to increased exploitation of natural underwater resources, especially in the ocean, which has led to a number of technological advances in the domains of oil and gas exploration, environmental monitoring, and the military, among others. Recently, under water visible light communication (UWVLC) has been preferred, which works well in the blue-green wavelength range and requires positioning along a line of sight. Due to support of large bandwidth, UWVLC is suitable for real-time image and video transmission. However, UWVLC suffers from high path-loss due to absorption and scattering, and turbulence-induced fading. The underwater optical turbulence is caused by fluctuations in the refractive index of water, which is influenced by variations in temperature and salinity of water. Furthermore, the behaviour of channel in horizontal and vertical communications is different as concentration of water is homogeneous in horizontal direction, but it is non homogeneous in vertical direction. Thus, these impairments severely degrade the performance of UWVLC system and limit its communication range to a short distance. Furthermore, the performance degrades severely with the depth due to the variation of sea parameters such as temperature, eddy particles and pressure. In this thesis, we have modelled the UWVLC channel using the statistical model with a log normal distribution, where the statistical parameters of the log nor mal channel depend upon the oceanic parameters such as water temperature, eddy particles, pressure, scintillation index of water etc. The statistical parameters of horizontal channel will be uniform, but they vary in the vertical direction. There fore, the vertical channel has been modelled by cascading log normal distributions with different statistical channel parameters. To combat the adverse effect of the underlying channel and improve the performance and range of communication, we have used spatial diversity with SIMO systems and cooperative diversity with multi hop relays with Decode and Forward (DF) principle. Further, to observe the effect of real time scenario of not having perfect information of channel, we have assumed imperfect channel state information at the receiver for detection. To support the high data rate, we have used square quadrature amplitude modulation (SQAM), rectangular QAM (RQAM) and Cross QAM (XQAM) constellations. Our major contribution is derivations of the expressions of probability density functions (PDF) of end to end (e2e) instantaneous SNR and then using Gauss Hermite approach, we have derived closed form expressions of the performance parameters such as outage probability, ergodic capacity, average symbol error probability (ASEP) and aver age diversity order (ADO). We have presented the performance parameters versus average SNR using Monte Carlo simulations and analytical expressions. A close matching between them validates the derived analytical expressions. We conclude that the performance degrades with imperfection in channel information available at the receiver. Further, the performance of vertical communication is very poor compared to horizontal communication, in terms of channel capacity, ASEP and distance. We have also observed that the performance degrades with increase in temperature.