LAUSR

The recent remarkable advancement of smart devices is enabling a higher-level flexibility of mobile sensing. Along with the rapid development of mobile devices and applications, a challenging issue is becoming more serious than ever before. A large number of mobility-based services have brought heavy workloads to mobile devices. Resource outsourcing via resource allocations is a type of method to mitigate local workloads. However, most current solutions are restricted by two issues, namely, the variety of inputs and the contradiction between optimal outputs and latency. In this article, we utilize the mechanism of Reinforcement Learning (RL) and propose a novel approach, named Smart Reinforcement Learning-based Resource Allocation (SRL-RA), to achieve optimal allocation through a self-learning process.