Abstract:Mobile traffic, which has increased significantly with the emergence of Fourth generation longterm evolution (4G-LTE) communications and advances in video streaming services, is still currently increasing at an incredible pace. Fifth-generation (5G) mobile communication systems, which were developed to deal with such a drastic increase in mobile traffic, aim to achieve ultra-high-speed data transmission, low latency, and the accommodation of many more connected devices compared to 4G-LTE systems. 5G communication uses high-frequency bandwidth to implement these features, which leads to an inevitable drawback of a high path loss. In order to overcome this disadvantage, small cell technology was developed, and is defined as small, low-power base stations that can extend the network coverage and solve the shadow area problem. Although small cell technology has these advantages, different problems, such as the effects of interference due to the deployment of a large number of small cells and the differences in devices accessing the network, need to be solved. To do so, it is necessary to develop an algorithm for a service method. However, general algorithms have difficulties in responding to the diverse environment of mobile communication systems, such as sudden increase in traffic in certain areas or sudden changes in the mobile population, and machine learning technology has been applied to solve this problem. This study employs a machine learning algorithm to determine small cell connections. In addition, a 5G macro system, the application of small cells, and the application of machine learning algorithms are compared to determine the performance improvement in the machine learning algorithm. Moreover, Support Vector Machine (SVM), Logistic Regression and Decision Tree algorithm are employed to show a training method that uses basic training data and a small cell on-off method, and the performance enhancement is verified based on this method.

A Machine Learning Based Approach to Mobile Network Analysis

Traffic Analysis Zone Based Urban Activity Study with Aggregate Mobile Network Data

Machine Learning for 5G and Beyond: from Model-Based to Data-Driven Mobile Wireless Networks

Mobile big data analysis with machine learning

When Machine Learning Meets Big Data: A Wireless Communication Perspective

Anomaly Detection for Cellular Networks Using Big Data Analytics.

Big Data Analytics in Mobile Cellular Networks

Deep Neural Mobile Networking

Big Data Analysis Based Network Behavior Insight of Cellular Networks for Industry 4.0 Applications

A survey of 5G network systems: challenges and machine learning approaches

Call Detail Records Driven Anomaly Detection and Traffic Prediction in Mobile Cellular Networks

Mobile Big Data Based Network Intelligence

Big Data-driven Automated Anomaly Detection and Performance Forecasting in Mobile Networks

Community Based User Behavior Analysis on Daily Mobile Internet Usage.

Machine Learning Assisted Security Analysis of 5G-Network-Connected Systems

Empirical performance analysis and ML-based modeling of 5G non-standalone networks

Machine Learning Enabled Wireless Communication Network System.

Improving the Performance of Heterogeneous Network Systems in Machine Learning-based 5G Mobile Communication System

A Review of Network Traffic Analysis Targeting Mobile Devices

The Dark Side(-Channel) of Mobile Devices: A Survey on Network Traffic Analysis

A Human-in-the-Loop Anomaly Detection Architecture for Big Traffic Data of Cellular Network