Procreation of Energy Efficient Topologies for Data Transmission in Underwater Wireless Sensor Network

Abstract

Water, being an essential component for survival of flora and fauna and covers almost 71.4% area on the Earth. It undergoes various chemical, thermal and physical changes due to man made and natural hazard which deteriorate the water quality. In order to monitor the quality of water with respect to its physical parameters such as Electrical Conductivity, pH, Tem perature, Chloride and Dissolved Oxygen, a consistent monitoring methodology is required. For the purpose, the different sampling techniques were suggested earlier that focused on evaluating underwater quality through sample collection manually from various locations. These samples are further sent for laboratory testing which is time consuming process. The techniques face various challenges like manpower, continuous monitoring, high cost and complex installation. To address the challenges, Underwater Sensor Network (UWSN) came into existence that targets to gather data through the sensors as a wired or wireless network. These network exhibits some limitations such as acoustic communication, large number of sensors, low bandwidth, high latency, high bit error rate, limited energy and data storage. Due to these limitations, they suffer inefficient means for communication between sensors and surface station. As an improvement, the Underwater Acoustic Sensor network (UWASN), a networking methodology of UWSN, is used for communication. UWASN consists of variable number of sensors and autonomous vehicles, to perform collaborative monitoring tasks over targeted area. These network captures physical parameters and images using different kind of sensors and forward them to surface station through acoustical signals. The acoustic signals have less propagation speed in underwater due to water constraints like path loss, ambient noise, propagation delay, high bit error rates, fading and Doppler effect, therefore, affects the data transmission and produces impaired sensed data. Moreover, the battery power of sensors is limited and are prone to failure. In summary, to deal with water constraints and to perform efficient data monitoring in underwater, there is a need to design energy efficient sensor and topologies. Therefore, in this work, a Multi-Parametric Sensor (MPS), two topologies, namely, Multi-Parametric Sensor Topology (MPST), Triangular Pyramidal Topology (TPT) and a technique for image transmission, are proposed. The two image refinement techniques i.e.,Mosaicing and Ridgelet based Fusion, are applied. The MPS is an integration of different parametric sensors and a camera fabricated on Arduino board. It is used in MPST to mea sure physical parameters in river water. It minimizes energy consumption due to single hop communication. Its main features are energy efficient, in-expensive infrastructure, less man power, longer sustainability and can cope with varying currents of water. MPST provides optimal coverage by minimizing the overlapping sensing areas and has assumed the area under anchored sensor as negligible. In MPST few of the sensors get displaced from its position due to currents of river that leads to increase in errors of localization and commu nication. Therefore, to improve its efficiency, TPT is proposed. As an extension to MPST, TPT provides full coverage of the targeted region considering the communication range un der anchored sensor. It helps in detecting physical parameters’ data in cost-effective manner and uses Dynamic Source Routing (DSR) protocol for data transmission. In order to transmit captured data and images to surface station without negligible er rors, a Hybrid-OFDM based image transmission technique is proposed. The technique uses various modulation schemes such as Phase Shift Keying for communication. The gathered images are blended and refined to produce big canvas using mosaicing technique. It helps to visualize wider area of targeted region. In order to improve pixel’s intensities along edges and visual quality of panoramic images, Ridgelet based Fusion technique is proposed. The technique helps to overcome few issues like poor visibility, distortion, light scattering, wave length attenuation due to water currents, varying colors with increase in depth, and refines images using fusion techniques. It provides better visual quality panoramic images having high pixel intensities, high content information and minimal erred pixels.