ORIGINAL RESEARCH Efficient resource allocation scheme for on-the-fly computing based mobile grids Amit Sadanand Savyanavar 1,2 • Vijay Ram Ghorpade 3 Received: 29 April 2018 / Accepted: 7 December 2018 Ó Bharati Vidyapeeth’s Institute of Computer Applications and Management 2018 Abstract Mobile grid (MG) is emergi ng as a new com- puting paradigm due to the ubiquitous availability of mobile devices. With the advancement in the capability of these devices, computationally intensive tasks can be executed using a peer-to-peer grid of such devices. MG can provide an edifice to execute parallel computationally intensive tasks. Key challenges that crop up while com- puting on a MG are resource constrained environment, inefficient resource allocation, high failure probability, etc. As a result, selection of appropriate nodes for task execu- tion becomes critical for successful execution of the application. In this paper, we propose an efficient resource allocation model (ERAM) which provides resource allo- cation with failure handling. We created a MG comprising of Wi-Fi Direct connected Android smartphones. Different scenarios are considered for the purpose of experimenta- tion. Our approach performs well with respect to applica- tion completion time, % battery consumption and recovery time from failure in comparison with existing techniques. Keywords Checkpointing Á Mobile grid computing Á Peer- to-peer computing Á Resource allocation Á Rough set theory Abbreviations MG Mobile grid ERAM Efficient Resource Allocation Model DRA Distance based resource allocation scheme NLRA Next-location based resource allocation scheme AMT Application Metadata Template 1 Introduction Peer-to-peer computing is an architecture for distributed application execution. The application comprises of tasks which can be allocated to peers. Peers are potentially capable as service and resource providers that can con- tribute for application execution. The computing and storage resources of the peers or nodes are used by the members of the network without being coordinated by a centralized server. The nodes act as both, resource supplier and consumer depending on the application requirement. This is in stark contrast with a client–server model. Based on such peer-to-peer computing model, powerful collabo- rative systems came into existence empowering them to accomplish tasks beyond the capacity of individual nodes. One such contemporary architecture is known as Mobile Grid (MG), where a network of mobile devices are deployed to execute a parallel, computationally intensive application. A true MG operates without any pre-existing network infrastructure. It can identify, connect and collaborate with the ubiquitously available Android devices in the vicinity. It can be deployed fast and hence is suitable for on-the-fly computing in real-life situations like disaster management (shown in Fig. 1). Being an anytime-anywhere computing paradigm, it provides a potentially viable solution for military [1], & Amit Sadanand Savyanavar amitsavyanavar@gmail.com Vijay Ram Ghorpade vijayghorpade@hotmail.com 1 Department of Computer Science and Engineering, Shivaji University, Kolhapur, India 2 Department of Computer Engineering, MIT College of Engineering, Pune, India 3 Department of Computer Science and Engineering, Bharati Vidyapeeth’s College of Engineering, Kolhapur, India 123 Int. j. inf. tecnol. https://doi.org/10.1007/s41870-018-0269-y