Research On Dynamic Resources Optimal Allocation Algorithm Of 5g Network Slices Under Multiple Scenarios

The development of 5G technology has promoted drastic changes in the world's informatization landscape,and has also given birth to many emerging application fields.While bringing convenience to human production and life,5G networks also face many challenges.For example,the deployment scale and construction speed of network equipment such as 5G base stations cannot meet the rapidly increasing data traffic and user service requirements.Therefore,it is of great significance to optimize the allocation of network resources within a relatively limited range.At present,traditional network resource allocation algorithms are restricted by the network structure,and must rely on basic network facilities for resource allocation,and have poor adaptability when facing complex application scenarios of 5G networks.In order to enable network resources to be scheduled at the application layer,researchers consider applying network slicing technology to resource allocation algorithms.This article mainly studies the dynamic resource optimal allocation algorithm of 5G network slicing in multiple scenarios.The main tasks completed are:(1)In order to meet the differentiated business needs of 5G networks,the key network performance requirements of the three major application scenarios of 5G networks are analyzed,and network resource optimal allocation algorithms are designed for them respectively.(2)A Dynamic Resource Optimal Allocation(DROA)is proposed to meet the massive demand for network bandwidth of e MBB slicing services.By introducing an admission control strategy,the algorithm always keeps the network operating state within a reasonable range,and can comprehensively consider the increase in throughput and the increase in resource scheduling costs.Finally,the simulation experiments are compared with the classic PFRA algorithm to verify the superiority of the DROA algorithm in improving network bandwidth and resource utilization.(3)An Improved Non-Linear Programming(INLP)is proposed to solve the network resource allocation problem of u RLLC slices.The introduction of the delay factor enables the resource allocation model of u RLLC slicing to not only meet the service delay time standard,but also to maintain the balance between the slicing revenue and the resource scheduling cost.Finally,simulation experiments verify the significant improvement of the INLP algorithm in terms of convergence performance and its effectiveness in delay control.(4)Proposed the Automatic Frame Control(AFC)algorithm.This algorithm introduces the idea of dynamic clustering.Based on the classic LEACH protocol,the cluster head is selected according to the remaining energy of the node itself,and the overall energy consumption of the network is balanced by controlling the number of data transmission frames in the stable operation stage.The comparison results of simulation experiments show that the clustering structure of the AFC algorithm is more reasonable and can effectively extend the life of the network,thereby having greater promotion and application value in the m MTC slicing business scenario.