Research On Resource Allocation Mechanism For D2D Communication Considering Delay Requirement And Reliability

D2D(Device-to-Device)communication technology has been widely studied due to its advantages in terms of enhancing spectrum efficiency,reducing transmission delay,decreasing terminal energy consumption and so on,especially for networks where D2 D users and cellular users coexist,how to improve various aspects of network system performance by optimizing resource allocation schemes is the main research direction.However,in the traditional D2 D communication,the resources used for the D2 D devices are shared with the cellular users,and there is no dedicated D2 D communication resource pool.In addition,the traditional D2 D communication has little consideration for transmission service delay.In future communication networks,it is very important to satisfy the time delay of D2 D services especially in ultra reliable and low latency communication scenarios,and the concept of sidelink is introduced in such D2 D communication.This paper studies the resource configuration of the sidelink-D2 D communication.There are two types of existing resource configuration modes for sidelink-D2 D communication.In Mode 1,the transmission resources are totally scheduled by the base station,while in Mode 2 resources can be selected competitively by D2 D users.The former mode can effectively guarantee the transmission reliability of D2 D communication,but the delay caused by signaling interaction is relatively high and cannot meet the delay requirement of the low-latency D2 D services.The latter mode can save signaling overhead and reduce scheduling delay,however,the spontaneous competition of selecting resources may lead to the reduction of transmission reliability.In view of this situation,the paper proposes an adaptive D2 D resource configuration mode switching scheme,which allows the D2 D device to autonomously switch between the two resource configuration modes according to the service delay requirement and the current congestion level of the resource pool,so as to ensure the delay demand and transmission reliability.In order to support the proposed scheme,there also defines a new uplink control information(UCI)format for D2 D devices to send the resource configuration mode switching request.The respective D2 D communication service delays are theoretically analyzed for Mode 1,Mode 2 and the scheme proposed in this paper.The simulation results show the effectiveness of proposed scheme in meeting the delay requirements of D2 D services.Based on the Q-learning algorithm in reinforcement learning,the proposed scheme can be further optimized.The sidelink-D2 D network is considered as the learning environment,the mode-switching decisions are taken as action set,and the two resource configurationmodes as state set.Considering the delay requirements and priorities of D2 D services,the corresponding rewards will be given to D2 D users that meet the service delay demands.Through continuous learning and exploration,the performance of the proposed scheme in resource configuration for sidelink-D2 D communication can be optimized.The simulation results show that the proposed scheme combined with the Q-learning algorithm can remain stable performance in delay demand satisfaction rates when D2 D users increase rapidly.Accordingly,the proposed sidelink-D2 D resource configuration scheme has good applicability to the future D2 D communication scenarios with ultra reliability,low latency and high equipment density.