A Taxonomy of Offloading in Mobile Cloud Computing

Mobile computing provides benefits of mobility, convenience and convergence. However, they have limited resources for computation, memory, networking, and battery. An effective approach to remedying the constraints is to offload resource-intensive components of a mobile app to a resourceabundant server and run it; called offloading and more generally Mobile Cloud Computing. There are a number of different approaches to offloading, and they result in different gains on benefit. In this paper, we propose a comprehensive taxonomy of offloading schemes and an assessment on the schemes. We first propose five criteria for classifying offloading schemes and define a taxonomy using the criteria. And, we choose a subset of practically configurable offloading schemes out of 32 schemes in the table. Then, we characterize the resulting offloading schemes by using key quality gains with offloading; Time Efficiency, Resource Consumption, and Adaptability to Load Evolution. Using the proposed taxonomy and guidelines, we present a case study of designing an offloadable mobile app and assess the applicability of the proposed taxonomy and guidelines.