Resource Scheduling Of Loosely Coupled Multi Cognitive Networks Based On Edge Computing

With the rapid development of communication services and big data computing services,the society are seeing higher demand for wireless spectrum resources.Increasing data traffic also puts forward higher requirements for the capacity of the core network.Cognitive wireless networks and edge computing research have been emerging because of uneven distribution of spectrum and computing resources.Aiming at the problems of interference and frequency band limitation in the network,the cognitive network can improve the spectrum utilization rate and system load capacity of the network by accessing spectrum opportunistically.Also,the decision of placing computing resources on the edge side of the network can alleviate the pressure of core network and the performance in terms of delay and fluctuation of edge computing tasks.Firstly,this paper lists the research status of cognitive networks and edge computing,and expounds some key technologies of these two researches as the theoretical basis of this research.Next,for multiple loosely coupled cognitive network models,two resource allocation and offloading decision models for edge computing tasks are proposed.The first model is aimed at reducing channel interference between multiple distributed cognitive networks.Considering the short-term stability of network equipment,a network model that records network channel occupancy and a control strategy based on interference perception and active interference by base stations is proposed.This strategy allocates channels semi-centrally based on historical interference information records and simulated annealing algorithm.With the exchange of channel usage information,it assists each network to gradually stabilize and find the optimal channel allocation scheme,reducing channel contention among multiple networks.The proposed algorithm effectively increases the number of tasks successfully unloaded in each communication round.For dynamically changing edge computing task states,the second model proposes a task offloading schema that jointly optimizes latency and energy performance by combining the channel conditions in the cognitive network and the end user’s preference coefficients.This decision-making and spectrum resource allocation algorithm optimizes energy consumption performance and latency performance.By recording and predicting the performance of each channel in the network,the algorithm calculates and optimizes the plan with the highest expected benefit according to different task offloading strategies and channel allocation plans.The Kuhn-Munkers algorithm reduces the implementation complexity of this strategy and the strategy improves the latency and energy consumption performance of computing tasks under various device layouts.To sum up,the cognitive network resource scheduling based on edge computing studied in this paper effectively improves the performance of edge computing services in wireless networks,and is of great significance to the future research on mobile terminal computing equipment.