Abstract—The most intriguing research areas in mobile communication networks recently is call drop prediction. Several artificial intelligence algorithms are created for this goal in earlier research projects; however, they struggle with the problems of large system complexity, poor efficiency, and lack of accuracy. Therefore, the goal of this research work is to create a cutting-edge framework, named as, DAI-BOTS for predicting mobile call drops using clever deep learning algorithms. Here, the special Deep Auto-Encoder based Data Imputation (DAE-DI) technique is used to generate the imputed data with normalized characteristics after data gathering. Then, a hybrid Skill Search based Levy Flight Optimization (S2LFO) method is created to select the most demanding features from the imputed data to lessen the classifier’s training and testing complexity. In addition, for precise call dropout prediction, the Bayes Optimized Transfer Learning Network (BOT-LN) classification algorithm is used. In this work, there are four distinct and emerging datasets such as Call Detail Record (CDR), Synthetic Minority Over-sampling Technique (SMOTE), Cell to Cell and IBC Telco used to assess the call drop prediction results of the proposed DAI-BOTS mechanism. In addition, this study also uses a few other publicly available open-source datasets for system validation, including SMOTE, Cell to Cell, and IBM Telco. Furthermore, the results of the DAI-BOTS system are compared based on a number of evaluation factors, taking into consideration the most recent state-of-the-art model techniques.
 
Keywords—mobile networks, call drop prediction, data imputation, feature optimization, machine learning, classification


Copyright © 2024 by the authors. This is an open access article distributed under the Creative Commons Attribution License (CC BY-NC-ND 4.0), which permits use, distribution and reproduction in any medium, provided that the article is properly cited, the use is non-commercial and no modifications or adaptations are made.