IoT connections emerge exponential growth trend, which poses a serious challenge to cloud computing due to the limit of network bandwidth. Edge computing provides a promising solution. However, the computing power of edge side isn’t enough to complete large scale tasks. Hence, to solve computing power shortage, a collaborative edge network (CEN) model is proposed, including system model and network communication model. Further, to estimate the performance of CEN, we design and implement ant colony algorithm to schedule tasks for IoT scenarios. The simulation results show, as the mobile devices increase to 1500, the CEN reduces significantly average task complete time by 90% and task failure rate (TFR) by 10% compared with OnlyEdge. And the CEN reduces sharply TFR by 70% compared with OnlyCloud. As a consequence, the comprehensive performance of CEN is leading, which has more powerful access ability to serve enormous IoT devices.
Collaborative Edge Network Research and Design Based on Ant Colony Algorithm for IoT Haijun Fu, Gang Xiong, Yue Liu, Bin Hu, Sheng Liu, Shichao Chen