Research On Resource Scheduling Based On Multi Access Edge Computing

After years of development of mobile Internet technology,emerging applications such as augmented reality,virtual reality and the Internet of things have put forward higher requirements for 5g wireless networks,including large bandwidth,high stability and low latency.Referring to Ericsson’s report,the total global mobile data traffic will exceed 237 EB/month in 2026.As the key technology and special architecture feature of 5g network,multi access edge computing reduces the delay of user interactive operation and strengthens the stability of related services by virtue of its proximity to end users.At the same time,wireless network operators have the opportunity to provide more wireless side information to software engineers,so that application providers can provide better services for users.Optimizing the access performance of multi access edge computing based on the characteristics of wireless network can not only effectively improve the operation capacity of existing edge computing platform applications,but also rely on edge computing to optimize the transmission capacity of existing services and improve user experience.Firstly,this paper uses mp-quic and edgegallery to build a set of multi access edge computing platform,and then constructs the edge information service system and data transmission display platform.Then,based on the state of wireless link,relevant solutions are proposed for congestion control and path selection algorithm.The main work of this paper includes:1.The construction of multi access edge computing platform and the design and implementation of wireless network information service system and transmission status display system.Based on the multi access edge computing standard of 3 GPP,the edge computing platform is constructed by using mp-quic and edgegallery.Then,redis,springboot and Tomcat are used to build the data transmission display platform,and the wireless network information service platform is designed and implemented in combination with redis and websocket protocols.Finally,the platform is tested.The test results show the good performance of the whole multi access edge computing platform.2.The problem of congestion control in multi access edge computing system is solved based on wireless link state.Firstly,the working principles and implementation methods of some original multipath congestion control algorithms are described;Then it introduces the principle and implementation of an algorithm based on wireless channel quality.By comparing the advantages and disadvantages of the above algorithms,a solution strategy based on olia and observing the link quality is designed.By detecting the quality of the link,dynamically adjust the transmission state of the link and adjust its transmission rate,so that the transmission can adapt to the change of the quality of the link itself more quickly.Experiments are designed and compared with the existing congestion control schemes.The mechanism can ensure that the transmission can flexibly use the changeable wireless network bandwidth.The results show that this algorithm has obvious advantages over other algorithms in this experimental case.3.Research on improving the path selection algorithm of multi access edge computing system based on wireless network link asymmetry.The model of the original algorithm is analyzed in detail,and the advantages and disadvantages of LRF and STTF algorithms are compared.Combined with the asymmetry of link states in multiple wireless networks,a solution of mp-quic path selection algorithm based on one-way link delay is proposed.Referring to the one-way delay,adjust the transmission sequence and path of control frame and data frame to complete the file data transmission process earlier.Considering a variety of scenarios and designing related experimental processes,it is verified that the newly designed algorithm can fully use the bandwidth of multiple paths.The results show that the transmission delay of the link has been significantly reduced and the transmission rate of the link has been greatly increased.