Machine Learning Techniques for Underwater Wireless Sensor Networks: A Comprehensive Study

Machine Learning Techniques for Underwater Wireless Sensor Networks: A Comprehensive Study

Deepti Rani, Anju Sangwan, Anupma Sangwan, Tajinder Singh
Copyright: © 2021 |Pages: 18
DOI: 10.4018/978-1-7998-3640-7.ch013
OnDemand:
(Individual Chapters)
Available
$37.50
Current Special Offers
No Current Special Offers
TOTAL SAVINGS: $37.50

Abstract

With the enormous growth of sensor networks, information seeking from such networks has become an invaluable source of knowledge for various organizations to enhance the comprehension of people interests. Not only wireless sensor networks (WSNs) but its various classes also remain the hot topics of research. In this chapter, the primary focus is to understand the concept of sensor network in underwater scenario. Various mechanisms are used to recognize the activities underwater using sensor which examines the real-time events. With these features, a few challenges are also associated with sensor networks, which are addressed here. Machine learning (ML) techniques are the perfect key of success to resolve such issues due to their feasibility and adaption in complex problem environment. Therefore, various ML techniques have been explained to enhance the operational performance of WSNs, especially in underwater WSNs (UWSNs). The main objective of this chapter is to understand the concepts of UWSNs and role of ML to address the performance issues of UWSNs.
Chapter Preview
Top

Introduction

Advances in the area of hardware and wireless network technologies have taken us to the verge of a new age where little wireless devices give access and participate actively in smart environments anytime, anywhere. WSN has achieved great popularity universally due to its potential applications in variety of fields. Wireless sensor network refers to a network of geographical distributed devices, known as nodes, having dedicated sensors. These can sense the habitat and communicate the environment's physical conditions collected from the field monitored via wireless links to base stations (Yick et al., 2008). The major objective of such networks is information collection by environment sensing and placing the collected data to relevant location. The number of sensor nodes depends on various applications and can vary significantly. Consequently, the management of so many nodes needs an algorithm that is scalable and efficient. Furthermore, the WSNs can change dynamically as desired by system designers or due to other external factors. Such situations can be handled by reprogramming and network reconfiguration (Kumar et al., 2019).

Underwater Wireless Sensor Networks (UWSNs)

Majority of the earth surface, approximately 70 percent, is covered with water. Out of the total water covered area, more than 95 percent area is oceanic. Rest of the water subsists in seas, rivers, streams, ponds, and lakes. Hence, a lot of activities take place under water which could neither be supervised by human beings nor by devices that are used for above earth surveillance. Underwater sensor network (UWSN) is a key technology for the surveillance of various under-water activities. Technologies used underwater are very much different from technologies used above water. UWSN is a technique that facilitates users and devices to send and receive messages in underwater scenario. Wireless sensor-based technology is globally adopted by various users as per need and requirement in different domains of WSNs. Architecture of Underwater WSNs is vulnerable to various challenges. UWSN provide solution for many area related problems.

Figure 1.

A typical underwater sensor network (Tuna et al., 2017)

978-1-7998-3640-7.ch013.f01

The UWSNs consist of several kinds of sensors like acoustic sensors used to track underwater physical information such as noise, temperature, pressure, disaster symptoms, attacks, pollution. Underwater Acoustic Sensor Networks (UASNs) are the network of sensor nodes that are used for monitoring the aquatic activities. UWSN is very frequently used for underwater object exploration, navigation of underwater vehicles, monitoring the behaviour of aquatic living and non-living bodies. Data from underwater sensors is gathered through some autonomous vehicles specially designed for collaborative monitoring of sea bottom (Awan et al., 2019). Sensors are the backbone of every kind of smart network which extract the data associated with physical and environmental conditions. The information based on collected data is diffused and shared among connected nodes through wireless links. Sharing of information is based on network topologies (bus, star, ring, tree, or mesh), which can be decided as per need and requirement as WSNs are deployed in different kinds of environments. Characteristics of UWSNs are also different from other traditional networks. Before the emergence of UWSNs, there were some flaws in the traditional approaches used for data monitoring and other operations used for underwater environments. It was very difficult to monitor the area covered with water in depth such as oceans, Marianas Trench, and sea. Acoustic Sensors are efficient in such circumstances (Singh et al., 2013).

Complete Chapter List

Search this Book:
Reset