Smart network maintenance in edge cloud computing environment: An allocation mechanism based on comprehensive reputation and regional prediction model

Abstract

With the rapid development of communication networks, there are more stringent requirements for its maintenance management. This paper uses edge cloud computing technology and wearable technology to improve the inconvenient information flow of on-site maintenance of communication networks. We propose an Emergent Task Allocation Mechanism based on Comprehensive Reputation and Regional Prediction Model (ETARR) in the edge cloud computing environment to solve emergency task allocation in smart network maintenance. Firstly, based on the basic reputation value, we add the work enthusiasm and work activity as the indicators to measure the work efficiency of the maintenance personnel. By using Long Short-Term Memory (LSTM) model to predict the historical reputation value sequence, the work enthusiasm of the maintenance personnel is quantified. Furthermore, based on the historical movement route of the maintenance personnel, the location of maintenance personnel is predicted by using the modified Workers’ Movement Patterns (WMP) model. Finally, combined with the reputation value requirements of the emergency task and the geographical location of the maintenance personnel, the maintenance emergency task is assigned. Simulation results show that the ETARR mechanism proposed in this paper reduces the cost of task allocation, improves the efficiency of the completion of the emergency task, and can be better applied to the emergency task scenario in the smart network maintenance environment.