NearBy-Offload: An Android based Application for Computation Offloading

Abstract

Computation offloading in fog computing not only addresses the computation-intensive and delay-sensitive needs of IoT applications but also improves the network lifetime of the IoT-Fog paradigm. As fog node suits as a place for availing offloading services, it introduces another challenge, i.e., selection of most suitable fog node among available heterogeneous fog nodes. Since fog node selection involves multiple criteria such as CPU power, energy, etc., the selection of fog node can be modeled as a multi-criteria decision making (MCDM) problem. In this work, we have designed an android based application, named Nearby-Offload, to provide a realtime implementation of computation offloading. Nearby-Offload also integrates two well-known MCDM methods, i.e., MAAU (Multi-Attribute Additive Utility) and TOPSIS (Technique for Order Preference by Similarity to Ideal Solution), for selecting the most suitable fog node based on the preference given by the user for criteria. This application is designed in such a way that new MCDM methods can be integrated by the user if required.