Skip to the main content

International Journal for Engineering Modelling, Vol. 33 No. 3-4 Regular Issue, 2020.

Original scientific paper

https://doi.org/10.31534/engmod.2020.3-4.ri.04d

Offloading Decision Algorithm Based on Distance Weighted K-Nearest Neighbor in Power Internet of Things

Jun Jia ; State Grid Jiangsu Electric Power Co. LTD. Research Institute, Nanjing 211103, CHINA
Jiangtao Xu ; State Grid Jiangsu Electric Power Co. LTD. Research Institute, Nanjing 211103, CHINA
Wenqing Cui ; School of Information Science and Engineering, Southeast University, Nanjing 210096, CHINA
Fengbo Tao ; State Grid Jiangsu Electric Power Co. LTD. Research Institute, Nanjing 211103, CHINA
Guojiang Zhang ; State Grid Jiangsu Electric Power Co. LTD., Nanjing 210000, CHINA
Chengbo Hu ; State Grid Jiangsu Electric Power Co. LTD. Research Institute, Nanjing 211103, CHINA
Zhaohui Zhang ; State Grid Jiangsu Electric Power Co. LTD. Research Institute, Nanjing 211103, CHINA
Ziquan Liu ; State Grid Jiangsu Electric Power Co. LTD. Research Institute, Nanjing 211103, CHINA

Full text: english pdf 808 Kb

page 63-73

downloads: 396

cite

Abstract

With the widespread popularity of power Internet of Things (PIoT), the data collected from smart meters are growing explosively, which makes the calculation task of power data more and more complex. In order to improve computing power and maximize resource utilization, an offloading decision algorithm based on weighted K-nearest neighbor (WKNN) is proposed. It first collects the training set required by the WKNN-based algorithm, including the Received Signal Strength (RSS) required for offloading, the transmission rate, and the load balance of the Access Point (AP), and then the Euclidean distance between the training set and the sample is weighted by Gaussian function. Finally, the result with the largest K similarities in the training set is the offloading result. The simulation results show that the proposed algorithm reduces the offloading delay of the computing tasks and improves the resource utilization rate effectively when the number of meters increases in the network, which ensures that the resources of the mobile edge computing (MEC) servers in the system can be effectively and evenly utilized.

Keywords

Power Internet of Things (PIoT); offloading decision; weighted K-nearest neighbor (WKNN)

Hrčak ID:

247775

URI

https://hrcak.srce.hr/247775

Publication date:

14.12.2020.

Visits: 1.215 *