A Smart Data Analytics System Generating for 5G N/W System Via ML Based Algorithms for the Better Communications

In comparison to previous cellular networks, 5G cellular networks provide a host of new features, one of which is the Network Data Analysis Function (NWDAF). With the help of NWDAF, network operators may easily incorporate solutions from outside vendors into their networks or use their own unique mathematical modelling (ML) centred data analytics methodology. The first section of this article explains the NWDAF's organisation and procedures as they are outlined in the standard papers of the Third Generation Partnership Project (3GPP). Next, using a cell-based methodology, a synthetic database for 5G networks is created that complies with the fields defined by 3GPP standards. Anomalies, including sudden surges in traffic inside a particular cell, are added to the data set to replicate real-world situations. The next step is to classify these abnormalities within any cell, subscriber type, and piece of user equipment. Three different machine learning (ML) models are used to estimate behavioural information, especially anomalies of network traffic, and anticipate network load: recursive neural networks, long-short term retention, and linear regression. Three different models are used to evaluate network load prediction with the goal of minimising the absolute error of the mean. Extreme gradient boosting and logistic regression are used for anomaly classification in order to maximise the volume under the receiver's operational characteristics curve. Simulation findings show that neural network techniques beat linear regression in the prediction of network load, while the tree-based gradient-boost approach performs better in anomaly identification than logistic regression. It is believed that the results of these approximations will improve 5G network performance by efficiently using NWDAF.