Achieving QoS in GSM Network by Efficient Anomaly Mitigation and Data Prediction Model

The cellular communication forms the backbone of the present society. The GSM network for its deployment and maintenance costs a major revenue. Efficient deployment and utilization form the key factor of the invested infrastructure. In this paper, the Call Detail Record (CDR) of the real-time data set is analyzed to find the traffic intense region over a spanning area. The k-means clustering algorithm is used and it is enhanced with the help of the elbow method. The anomalies in the CDR dataset are the abnormal behavior of the users in the coverage area. The dataset is subjected to regressive optimization and the anomalies are removed to find the usage characteristics and behavior in the coverage area. The dataset is fed to the Bayesian Generalized linear model to predict the usage of the coverage area, which is proposed as a novelty in this paper, and in future, this data can be very crucial for the network service provider to reconFigure the bandwidth of the network in the signal traffic intense areas. Efficient bandwidth allocation results in the organized load balancing in the network which on a prolonged time frame will improve the Quality Of Service (QoS) in the network.