IEEE Project Abstract

In the emerging paradigm of Edge Computing forInternet of Things (IoT), data processing is pushed to the edgeof the IoT network (e.g. gateways and embedded IoT devices).IoT devices must support multiple operation modes in order toadapt to varying runtime situations, like preserving energy atlow battery, while still maintaining some crucial functionalityetc. Adapting the optimal operation mode is a challenge for edgedevices given the limited resources at the edge of the network(both bandwidth and processing power of the shared gateway),various constraints (e.g. battery lifetime), etc.This article proposes a fast and low-overhead scheme todetermine and adapt the operation mode of edge devices atruntime and orchestrate devices in a way that the efficiencyof IoT devices is optimized with respect to the gateway’s resourceconstraints. The proposed scheme breaks the optimizationproblem into several smaller ones (i.e. sub-problems) whosesolutions are aggregated to find the final solution. We presenta novel memoization technique that determines the solution to arange of sub-problems based on sub-problems that are alreadysolved. In addition, we present a novel pruning technique thatreduces the search space and consequently reduces both memoryand execution time overhead. The experimental results show upto 50% reduction in memory overhead and 14 reduction inexecution time overhead compared to the state-of-the-art solutionwhich is a major step towards efficient edge computing for IoT.