Improving Mobile Device Efficiency Through Dew Computing and Multi-Language Communication

Low processing power and limited battery life are typical constraints for most mobile devices. To address these challenges, many researchers advocate for computational offloading, a technique where devices with limited computational capacity transfer tasks to more robust machines, thereby saving time and computational resources. However, previous research has identified network latency and inefficient programming languages as significant obstacles to the effectiveness of computational offloading. In this context, this paper proposes the adoption of multilanguage communication in conjunction with the Dew Computing paradigm as a strategy to replicate offloading server processes written in programming languages other than Java or Kotlin on Android smartphones. The main goal is to mitigate network effects while enhancing mobile applications' computational performance. We conducted experiments using real devices to evaluate the impact of this approach on task computation. The results indicate that the Dew approach is up to 7.2 times faster than conventional local computing and up to 4.4 times faster than remote processing in environments with high network overhead. Additionally, it reduces data transmission by up to 2.5 times, leading to significant bandwidth savings.