Data Handling Approach for Machine Learning in Wireless Communication: A Survey

Recently, wireless communication network has evolved with different types of communication architectures and protocols. Operation and management of such heterogeneous networks with huge demand are manually difficult for network engineers. In the recent past, Machine Learning (ML) has proven its capability by significantly improving the performance in various fields such as natural language processing and medical diagnostic. Using ML in wireless communication is also not a simple task, as we have to track the user’s Quality of Experience (QoE) on the one hand and network resource management on the other, with continuously changing wireless scenarios. Identifying channel variability with proper decision-making is the crucial task of ML in Wireless Communication Network. In this paper, based on a systematic review of the current use of machine learning techniques in WCN, a set of crucial design limitations are identified, and a novel computationally efficient data exchange approach is proposed.