Research On Conflict Avoidance Technology In Intent-driven Radio Access Network

In the face of the increasingly complex wireless network and its diverse performance and business requirements in the future,the radio access network needs to be adaptive reconfiguration according to different performance objectives,network environment and business requirements.Intent-driven radio access network has become an important component of the introduction of artificial intelligence to wireless network.When multiple intents are implemented in intent-driven radio access networks,resource competition and conflict of different configurations will be caused,which will bring challenges to the deployment and implementation of intentions and the stability of network performance.Therefore,the conflict avoidance technology in intent-driven radio access networks is studied in this paper.First of all,in order to deal with the problems of high cost and low efficiency brought by traditional management of radio access network,and considering that the research of intent conflict avoidance technology needs effective network management methods as support,an intent-driven radio access network management method is proposed in this paper,which introduces the concept of intent to manage radio access network.With the designed network function modules and protocol flow,the input intents can be transformed into the deployment and implementation of radio access network configuration strategies,and the active network performance maintenance and optimization can be realized.Based on this management method,it can effectively reduce the complexity of network configuration and the cost of manual operation and maintenance management,making the management of radio access network more intelligent,flexible and efficient.Secondly,considering the limited edge network resources,a multi performance intention conflict avoidance method is proposed,to solve the problem of different intent conflict in single service network in this paper.By considering two types of performance intents,the weighted fairness index is introduced,and the multi performance intent conflict avoidance problem is transformed into an optimization problem to maximize the weighted satisfaction of each intention,and the resource allocation strategy is obtained through the deep reinforcement learning method.Simulation results show that compared with the existing schemes,the proposed method can effectively avoid multi performance intent conflict and improve network performance.Finally,considering that the future radio access network needs to face the different requirements of different types of services,under the limited bandwidth resources,this paper models the data flow of each service in the form of packets,and takes the packet arrival rate and delay of each service as the indicators to calculate the satisfaction.By considering the best effort and guaranteed service intention,a differentiated intention conflict avoidance method is proposes based on service priority,and a resource allocation strategy based on the deep reinforcement learning algorithm is introduced.Simulation results show that the proposed algorithm can alleviate the conflict between multiple intents and ensure the quality of service of high priority services.This paper lays a theoretical foundation for the research of intent-driven radio access network,and also provides technical support for its standardization and commercial deployment.