Research On Task Offloading In Cell-Free Heterogeneous Edge Network

With the rapid growth of big data traffic,many applications that require higher network transmission rates and lower transmission delays have emerged,such as virtual reality and intelligent driving.There are some problems in cellular networks,such as limited data transmission rate,poor fairness to cell edge users,and long propagation distance of data signals from mobile user terminals to remote cloud servers,which will lead to high transmission delay of applications and cannot meet the low latency requirements of computationally intensive real-time applications.In view of the above problems,this paper proposed a communication network based on cell-free massive multiple input multiple output(MIMO)system and heterogeneous edge computing(CF-HetMEC).A task offloading scheme was discussed in the network to reduce system delay.In terms of reducing transmission delay,the system capacity can be improved by designing a high-speed wireless network connection.We modeled the new communication system and transmission signals and obtained the uplink and downlink transmission rates of signals.Taking spectral efficiency as a performance index,the variation of spectral efficiency with the number of users and the selection radius of users is analyzed.Then we compare this network with the traditional cellular network.In terms of reducing computation delay,a more balanced and low-delay network transmission can be achieved by designing a reasonable offloading scheme in the network.In this paper,the problems of joint communication networks and computing task offloading are formulated,and we propose a directed acyclic graph(DAG)based task offloading algorithm in CFHetMEC network.Firstly,the user’s tasks are broken down into dependent subtasks,and multiple wireless access points with heterogeneous processor capabilities are introduced to provide services for large-scale users at the same time.Then,the performance of system delay and user energy consumption under the condition of different number of users and access points is verified.Simulation results show that compared with traditional cellular networks,the CF-HetMEC network has better spectral efficiency and can effectively reduce signal transmission delay.During task offloading,the proposed algorithm can significantly reduce the system delay under the constraint of user energy consumption.