Towards the Advanced Data Processing for Medical Applications Using Task Offloading Strategy

Abstract

Broad adoption of resource-constrained devices for medical use has additional limitations in terms of execution of delay-sensitive medical applications. As one of the solutions, new ways of computational offloading could be developed and integrated. The recently emerged Mobile Edge Computing (MEC) and Mobile Cloud Computing (MCC) paradigms attempt to address this problem by offloading tasks to a the resource-rich server. In the context of the availability of eHealth services for all patients, independently of the location, the implementation of MEC and MCC could help ensure a high availability of medical services. Remote medical examination, robotic surgery, and cardiac telemetry require efficient computing solutions. This work discusses three alternative computing models: local computing, MEC, and MCC. We have designed a Matlab-based tool to calculate and compare the response time and energy efficiency. We show that local computing demands 48 times more power than MEC/MCC with increasing packet workload. On the other hand, the throughput of MEC/MCC highly depends on the parameters of the communication channel. Finding an optimal trade-off between the response time and energy consumption is an important research question that could not be solved without investigating the system's bottlenecks.