AROM: An Adaptive Resource Optimization Model for Mobile IoT Edge

Crowdsourced approach using mobile phones towards rescue and relief operations in post-disaster situations is becoming popular. The mobile phones play a critical role in estimation, evacuation and relief operations as they can be used for sensing and transmitting information such as location, elevation, acceleration, images and other user generated messages, but the mobile IoT edge devices are often resource constrained by battery and network. In this paper we present AROM, an Adaptive Resource Optimization Model for optimizing mobile sensing and communication systems, by considering network and battery constraints. AROM works by adaptively computing a resource criticality score and controlling sensing mechanisms and throttling and prioritizing network communications based on the score. We also present the details of performance tests done to validate our model and the analysis of results. We have implemented this architecture in our Amrita-Kripa disaster relief application which was deployed during 2018 and 2019 Kerala floods for rescues and relief activities.