Research On Resource Allocation Of NOMA-MEC System

Multi-access edge computing(MEC)can efficiently process data generated by edge devices and provide computing resources for devices at the edge of the wireless network.Applying non-orthogonal multiple access(NOMA)technology to MEC network can effectively increase the number of device connections and improve the offloading efficiency of users.Although NOMA-MEC system has many advantages,it still has some disadvantages,such as the risk of information leakage brought by the use of relay-assisted transmission and the freshness of information reception in Industrial Internet control application scenarios.Therefore,this paper mainly studies the resource allocation of NOMA-MEC system from two aspects of physical layer security and the freshness of information.First,distance is an important factor that affects the offloading performance of NOMA-MEC system.Relay-assisted transmission can effectively improve the coverage and offloading efficiency of the NOMA-MEC system.However,the relay will be untrusted relay when it is attacked by malicious users,resulting in information leakage.To solve this problem,the MEC server is used to actively interfere with untrusted relays to improve the information security performance of the system.This article jointly optimizes the user’s transmission power,CPU frequency and offloading time to maximize the physical layer security energy efficiency of the user under the condition of limited time delay and energy consumption.The single-objective sub-problem is transformed into a convex problem using the Taylor approximation method,then be solved by the Lagrange dual method.The alternate iterative algorithm is used to jointly optimize the single objective optimal solution of user’s transmission power,CPU frequency and offloading time.The simulation results show that the proposed offloading method can effectively improve the security offloading performance and physical layer security energy efficiency.Secondly,Industrial Internet control services have strict requirements on timeliness of information,but the current research on the timeliness of the NOMA-MEC system is rarely involved.Therefore,the age of information(AoI)is studied,which represents the timeliness of offloading of the NOMA-MEC system.Due to the random burst characteristics of network traffic,compared with traditional theoretical analysis methods such as queuing theory,the stochastic network calculus can fully simulate the arrival characteristics of network traffic and allow certain violation probability to exceed the network performance boundary.Therefore,this study uses the stochastic network calculus theory and the MEC offloading process to derive the moment generating function of the arrival process and the service process,combines with the moment generating function and the AoI evolution of service process to obtain the expression of average AoI,finally derives the violation probability of AoI.The theoretical analysis shows that the power allocation of NOMA-MEC system is an important factor which affects the AoI of the system.With the increase of the power allocation coefficient of MEC1 server,the average AoI of the near-distance server link decreases gradually,and the average AoI of the long-distance server link increases gradually.The minimum average AoI value of the system is obtained when the power distribution coefficient is0.1.Finally,the simulation results verify the correctness of theoretical analysis.