An efficient and scalable framework for mobile cloud computing environment

Abstract

In this research, various existing frameworks of different computing s explored and a comprehensive computing framework for MCC and MEC is proposed. This proposed SKYR framework is scalable as we can add functionality module as per requirement and it also manages the computing resources such as primary and secondary cloudlet and can scale up and down these resources in the computing environment to satisfy the utility of users and resource providers. We have surveyed various pricing theories and incentive mechanism to propose a pricing mechanism which increases the profit for resource provider as well maximize the utility satisfaction to users. In this dissertation, we have proposed three pricing model, two for user-based task which are SPOIMR and SUPMAR pricing mechanisms and one for sensor-based task which is named as SRRSIM. The major contribution of this research work is error free and energy efficient SEOR algorithm for sensed data in cloudlet-based computing for MCC and MEC. This algorithm reduces the complexity of computation as compared to the existing competitors. We introduce a STAR algorithm which caters scheduling of both independent and workflow-based tasks at both local level at primary cloudlet and remote level at cloud and federated cloud. These four tiers include the edge and dew layer at the bottom, then next upward layer is fog or primary cloudlet layer and, third and fourth layer comprise of cloud and federated cloud respectively. The purpose of introducing this four-tier computing is to improve the ease of computation which indeed optimizes the satisfaction for user and profit to the resource provider. During this complete research work, the initially proposed SKYR framework upgraded gradually with the introduction of different functionality modules. SKYR framework upgraded to SE-SKYR then to ASME-SKYR and finally, it became part of four-tier EVACON-Rainsnow computing. newline