Prediction of Dynamic Crowd Formation in Cellular Networks for Activating Small Cells

In today´s world, mobile communication has become a ubiquitously used technology. In real world scenarios such as open air festivals, shopping malls, stadiums or public events, several mobile users gather to form a crowd. This poses high load situation to the respective serving base station at the site of crowd formation, thereby leading to congestion. As a consequence, mobile users are subjected to poor Quality of Experience (QoE) due to high dropping and blocking rates. This paper exploits mobility patterns of the users in neighboring cells to predict formation of crowd at the cell of interest. The work presented here combines concepts of next cell prediction, cluster detection and trajectory prediction to indicate the severity of crowd formation and to predict probable location of crowd formation. This information could be used to design smart and efficient resource management schemes. In the presented work, this context information about formation of crowd is used to proactively trigger load balancing at the predicted site and free up resources. Alternatively, a small cell is activated at the predicted site of crowd formation. The simulation results show substantial reductions in blocking and dropping of users, thereby improving the QoE of users in the crowd.