面向D2D组播内容推荐的语义通信资源管理  

Resource Management in D2D Multicast Content Recommendation for Semantic Communications

作　　者：姬智 管新荣 吴丹 邓诚 申歆歆 JI Zhi;GUAN Xinrong;WU Dan;DENG Cheng;SHEN Xinxin(College of Communications Engineering,Army Engineering University of PLA,Nanjing 210007,China;Unit 32579 of PLA,Guilin 541000,China;School of Electronic Science and Engineering,Nanjing University,Nanjing 210046,China)

机构地区：[1]陆军工程大学通信工程学院,江苏南京210007 [2]32579部队,广西桂林541000 [3]南京大学电子科学与工程学院,江苏南京210046

出　　处：《陆军工程大学学报》2024年第6期17-24,共8页Journal of Army Engineering University of PLA

基　　金：国家自然科学基金面上项目(62171461);国家优秀青年科学基金(62122094)。

摘　　要：未来语义通信将广泛运用于蜂窝无线网络。针对设备到设备(device-to-device,D2D)内容推荐场景,研究了蜂窝网络中按需求语义通信的问题。提出了一种以目标用户接收体验(quality of experience,Qo E)为指标的语义传输模式,使簇头目标用户可根据时延和图像保真度的需求来调节语义压缩比例与选择接入链路,进而从语义压缩比例和信道分配的角度设计了资源优化问题。为解决所提问题,针对图像语义通信中的语义编码解码过程,设计了比例压缩因子,提出了一种小批量模型训练方法,通过随机选择批次来训练子语义模型,兼顾模型性能与训练复杂度。引入博弈论的方法解决信道分配与语义比例压缩因子选择问题,该问题被建模为势能博弈问题,设计了效用函数和势能函数,并证明纳什均衡的存在。提出了参数修正的空间自适应算法来达到纳什均衡。仿真结果证明了所提方法在目标用户整体Qo E上的有效性。In the future,the semantic communications will be widely used in cellular wireless networks.This paper focuses on the issue of the on-demand semantic communications in cellular networks in the context of device-to-device(D2D)content recommendation scenarios.A semantic transmission mode is proposed in this paper,which takes the target user's quality of experience(QoE)as the indicator.The mode allows the cluster head's target user to adjust the semantic compression ratio and select the access link based on the requirements of delay and image fidelity,and further designs a resource optimization problem from the perspective of semantic compression ratio and channel allocation.To solve this problem,a ratio compression factor is designed for the semantic encoding and decoding process in image semantic communications,and a small batch model training method is proposed to train sub-semantic models by randomly selecting batches,taking into account the model performance and training complexity.In addition,the game theory approach is introduced to solve the problems of channel allocation and semantic compression factor selection.The problem is modeled as a potential game problem,and utility functions and potential functions are designed,and the existence of Nash equilibrium is proved.Finally,a parameter correction space adaptive algorithm is proposed to achieve Nash equilibrium.The simulation results demonstrate the effectiveness of the proposed method on the overall QoE of the target user.

关 键 词：蜂窝网络 设备到设备通信 语义通信 用户接收体验 博弈论 

分 类 号：TN926[电子电信—通信与信息系统]