Underwater wireless communications refer to transmitting data in an unguided water environment by wireless carriers including acoustic, radio frequency (RF), and optical waves. Relative to acoustic and RF, the optical wave is more promising to offer higher bandwidth at a lower energy consumption rate. However, an optical wave has its challenges such as attenuation due to absorption, scattering and turbulence effects. Therefore, this work attempts to investigate the performance of lightwave propagation for underwater optical wireless communication (UOWC) using simulation and experimental approaches. First, the performance of optical waves was analyzed using MATLAB by simulating the light attenuation model which based on depth-dependent chlorophyll concentration. A depth profile that related to the surface chlorophyll levels for the range 0-4 mg/m3 was used to represent the open ocean. The simulation showed that the attenuation of light less affected for operating wavelength range of 450 – 550 nm. Further, an experimental set-up was developed which consists of a transmitter, receiver, and aquarium to emulate the UOWC channel. Three types of water including clear, sea and cloudy were tested to analyze their interaction with the light emitted by a light-emitting diode (LED) and a laser diode. The emitted light detected by the light sensor and the strength of an audio signal transmitted through the UOWC were measured using a light meter and sound meter respectively. The measured power was plotted against distance and the attenuation constant c was deduced through curve fitting method. The analysis showed irrespective of the light sources, UOWC in cloudy water suffered the highest attenuation relative to still clear and seawater. The received power emitted by laser was at least 41% higher than the LED. This study contributes to identify the potential and limitations of different operating schemes to optimize UOWC performance.
One of the biggest power consuming devices in wireless communications system is the Power Amplifier
(PA) which amplifies signals non-linearly when operating in real-world systems. The negative effects of
PA non-linearity are energy inefficiency, amplitude and phase distortion. The increases in transmission
speed in present day communication technology introduces Memory Effects, where signal spreading
happens at the PA output, thus causing overhead in signal processing at the receiver side. PA Linearization
is therefore required to counter the non-linearity and Memory Effects. Digital Pre-distortion (DPD) is
one of the outstanding PA Linearization methods in terms of its strengths in implementation simplicity,
bandwidth, efficiency, flexibility and cost. DPD pre-distorts the input signal, using an inversed model
function of the PA. Modelling of the PA is therefore vital in DPD, where the Memory Polynomial
Method (MP) is used to model the PA with memory effects. In this paper, the MP method is improved
in Memory Polynomial using Binomial Reduction method (MPB-imag-2k). The method is simulated
using a modelled ZVE-8G Power Amplifier and sampled 4G (LTE) signals. It was found MPB-imag-2k
is capable of achieving comparable anti-scattering/anti-distortion in MP for non-linearity order of 3,
memory depth of 3 and pre-amplifier gain of 2.