Research And Implementation Of Cognitive Service Network Function Towards 6G Core Network

The US Commerce Department added Huawei to the BIS Entity List in May,2019,making 5G an important part of the competitive field and national strategy for major countries worldwide.It is exactly the great strategic significance of the mobile communication network that whole world gradually pays more attention to 6G research.The competition for 6G has quietly begun.This is not only a hot spot in the world’s technological frontier,but also a competitive field for major powers.As the core part of the mobile communication network,the core network plays a decisive role.However,current research mainly focuses on the wireless air interface,the research on the 6G core network is rarely involved.This paper first analyzes the problems of complex service invocation and insufficient cognitive ability in the 5G core network.Then unified cognitive flow is abstracted,and cognitive message processing rules are designed through deep research on cognitive service network element.Further,cognition network function is designed and implemented based on cognitive service based architecture.Finally,a path selection algorithm for core network user plane is proposed based on cognitive information.The main work of this paper is as follows:(1)The cognitive network funciton for 6G core network is designed and implemented.CNF can manage the cognitive flow generated by cognitive objects,together with the rule sets of all cognitive flows in the cognitive domain.Then cognitive flow messages can be uniformly processed based on the rules to generate cognitive information,providing general cognitive ability for the core network.This provides an important reference value for the 6G core network standards in the future.(2)A core network user plane path selection algorithm is proposed.The algorithm first builds a knowledge graph based on cognitive information,and then uses knowledge graph embedding to train and learn the feature relations of users,services,links and topology.Finally,the experimental results based on real historical data show that the algorithm effectively improves the service hit rate of the session path.