A Review on Machine Learning Techniques for QoS in WSN

Abstract

WSN is circulated, self-directed and distributed in nature. WSN is a kind of network consist of multiple nodes which are wireless sensors connected to the base station. In recent days WSN is widely used for sensing vital information and communicate to the destination by means of the base station. In this transmission, it is essential to use the best efficient path and proper utilization of the available resources. Generally, nodes in the WSN are energy constraints, on the absence of efficient path it leads to lengthening of network lifetime and results in severe causes. The wide growth of WSN and its importance in used application increase its attention in the researched area. There are several traditional approaches were designed for WSN in the motto of limited energy usages. Most of the existing methods are one -size-fits-all approaches which are reactive and centrally-managed. But these are not properly fit for satisfying and serving the future complex networks on the aspect of cost-effective as well as optimization. On this way, Hierarchical routing protocols result effective on the concern of energy efficiency. These hierarchical protocols utilize clustering approach in collecting and disseminating the data. The need for huge data to be processed on WSN during sending and receiving makes this approach still in development stage. The important factor to be considered during the WSN process is bandwidth, sensor energy consumption, and time consumptions. To overcome these issues an ML (machine learning) based effective algorithm need to develop in order to improvise the WSN characteristic in all manner. The intention of this survey is to establish Machine Learning as an applied methodology in overcoming the WSN problems especially in the term of energy efficient routing.