Network Coverage Improvement during Natural Disaster using Self-Organizing Maps

This paper presents an emergency response management system to tackle  the problem of the absence of network connectivity during the time of a  natural disaster. Network connectivity is often enabled by the base  stations on the ground. However, during the time of the disaster, the  connectivity is disrupted due to the base station being damaged. During  such scenarios, the Unmanned Aerial Vehicles (UAV) based stations could  help in partially providing the network connectivity and help in the  rescue operations. But, the UAVs need to be quickly deployed and placed  at a suitable location based on the population coverage and base  stations being impacted due to the disruptions. In this paper, we  propose the Self Organizing Map (SOM) based optimal UAV deployment to  enhance the network coverage, and increase the percentage of people  having network access. In contrast to other Artificial  Intelligence-based approaches, like Deep Neural Networks, our method  does not require to be heavily trained using train and the test dataset.