Development of a framework for adaptive productivity management for edge computing based IoT applications

Data collected from IoT devices can be encrypted and analysed by computer resources at the network edge in your nearby cloud to alleviate congestion load in the network infrastructure, whilst different IoT applications can indeed be run in clouds to reduce the time among both IoT users such as mobile network user devices) and clouds. Taking into account the geographical and temporal characteristics of the cloud working loads in each application, the allocation of the load across the clouds for each IoT application influences the application’s response time. While it can decrease network latency by allocating IoT users’ requests to the local cloud, computer delay can be unpleasant if an application’s associated virtual machine is overextended in a cloudlet. In order to overcome this problem, we are developing an application-known IoT workload assignment scheme to minimise the response times of IoT application requests by deciding the destination cloud for the different types of requests for each IoT user and the amount of computing resources assigned in each cloud for each application. This approach takes into consideration both network delays and compute delay, requests of IoT users more likely be assigned to clouds nearby and less loaded. In the meantime, the system adjusts computer resources from the various applications on a cloud basis on the basis of its workload and therefore reduces computational delays for all cloud requests. Comprehensive simulations have confirmed the efficiency of the control strategy.