Research On Optimization Scheme Of Task Allocation For Edge Computing In 5G Heterogeneous Network

With the advent of the internet of everything era under the 5th generation mobile networks(5G),mobile edge computing(MEC)has become a key technology in 5G,by transferring various computing and service functions to the edge of the mobile network,it provides real-time information transmission and efficient computing services for mobile terminals with limited energy and computing constraints.Its advantages of high bandwidth,low latency and low cost are widely used in network scenarios that require real-time data processing and low power consumption.Current researches on MEC mostly consider reducing the time delay,energy consumption and cost of the user task offloading under ideal conditions,but the consumption generated by the queuing process of tasks to be offloaded is very important for the whole offloading decision system.Moreover,in the multi-server multi-user communication scenario,the service capacity restriction of heterogeneous servers and the transmission quality between users and servers will have an important impact on the user offloading experience,existing research ignores the quantitative research and analysis of task offloading process and influencing factors in real scenes.At present,the application of MEC technology in the actual communication scenarios of 5G network faces great challenges,that is,how to provide users with highly reliable and low consumption of network services.In order to solve this problem,the thesis takes MEC as the core technical solution to carry out user task offloading and scheduling in the network edge scenario of multi-user and multi-server.The main work and research content are summarized as follows:1.Investigate the key issues involved in the MEC technology,take the current research deficiencies as the starting point,and put forward the research significance and goals of the thesis.Secondly,the techniques and theoretical methods used in the thesis including computational offloading technology,queuing theory,constrained optimization algorithm and chaotic optimization theory are introduced and summarized.2.In view of the congestion problems in actual offloading scenarios,the thesis uses the multi-access characteristics of 5G heterogeneous networks and queuing theory,comprehensively considers the heterogeneity of base stations,by establishing the waiting and transmission consumption model of user task offloading,the goal of optimizing user delay and energy consumption is proposed.An optimization scheme based on the probability of task allocation is designed,and a task allocation algorithm based on the quasi-Newton interior point method is adopted,which speeds up the convergence speed and reduces the algorithm complexity,and is closer to the optimal solution of the objective function.The simulation results prove the effectiveness of the proposed algorithm,and the scheme can effectively reduce the total consumption of user offloading tasks and alleviate system congestion.3.Aiming at the problem of invalid appropriation of edge service resources and high energy consumption of user offloading caused by inefficient communication of multiple users in edge scenarios,the goal of optimizing the energy consumption of expected successful task offloading is proposed.On the premise of ensuring the minimum communication quality,consider the impact of transmission quality and congestion on task offloading performance,and the opportunity function for the successful establishment of communication between the user and the base station is constructed.Comprehensive consideration of the service capacity limitations of heterogeneous base stations,the queuing theory is used to model the queuing mechanism of offloading tasks,and efficient task assignment strategies are set up for users to realize more reliable transmission and accept idler service resources.The chaotic search task allocation algorithm based on the mixed penalty function is adopted,and the pseudo-random and ergodic characteristics of chaos are used to search and obtain the optimal solution of decision.The simulation results show that the proposed scheme can accelerate convergence,effectively reduce user offloading energy consumption,and improve the communication quality.