Mobile Cloud Computing Techniques for Extending Computation and Resources in Mobile Devices

This paper addresses whether Mobile Cloud Computing models can be used to extend the capabilities of resource-constrained mobile devices to provide the illusion of infinite, elastic resources on demand. We identify five key limited resources as being CPU, memory, battery, data usage and time. Existing solutions for these limitations are explored, and we identify offloading computation and storage from the device as a possible solution. Offloading techniques mainly support two types of offloading; offloading to a remote cloud and to a peer device. Four different modes of execution are analysed; local SQL database, local NoSQL database, cloud NoSQL database and NoSQL peer-to-peer database. We develop a prototype which showcases these execution scenarios. Healthcare is identified as an ideal use case for our prototype due to the varied purposes it provides (e.g. in a hospital, in a clinic, on site etc.). An open dataset of anonymised healthcare data is used as test data on which a set of experiments are run. Data generated via monitoring and logging of the prototype is collected and evaluated in terms of different execution scenarios. Using existing research we determine that Mobile Cloud Computing is a useful model for extending computation and resources in mobile devices. We identify that the discrepancy between our test results and existing research is due to the research approach, which evaluates results on a limited set of criteria and does not expand to other factors. Industries that would benefit from Mobile Cloud Computing include e-commerce, healthcare and e-learning, as well as education, finance, point-of-sale and transportation.