摘要
在超宽带通信技术中,获取高质量的信道冲激响应数据对系统设计和性能优化至关重要。引入最小二乘生成对抗网络和改进的损失函数,能显著提升信道数据的捕捉和复现能力。结合特征匹配技术和条件生成对抗网络,可以增强生成数据的细节准确性和多样性,还能使模型根据不同通信环境和信号场景进行数据生成。在模型训练阶段,采用能够代表全局特征的重构信道数据,而在测试阶段使用了经历无线衰落的实际信道数据。实验结果显示,模型在小样本数据集和复杂衰落信道环境下的表现优于带有梯度惩罚的Wasserstein生成对抗网络(WGAN-GP),识别准确率提高4.8%,模式崩溃问题减少5%。
In ultra-wideband communication technology,high-quality channel impulse response data is crucial for sys‐tem design and performance optimization.A least squares generative adversarial network(LSGAN)and an improved loss function were introduced,which significantly enhanced the ability to capture and reproduce channel data.By combining feature matching techniques with conditional generative adversarial networks(CGAN),it was able to im‐prove the detail accuracy and diversity of the generated data.The model was allowed to generate data according to dif‐ferent communication environments and signal scenarios.During the model training phase,reconstructed channel data representing global features were used,while actual channel data experiencing wireless fading were employed during the testing phase.Experimental results demonstrate that the model outperforms the WGAN-GP in small sample datas‐ets and complex fading channel environments,with a 4.8%increase in recognition accuracy and a 5%reduction in mode collapse issues.
作者
诸葛斌
王正贤
汪盈
蔡晓丹
董黎刚
张子天
蒋献
李华
徐越倩
ZHUGE Bin;WANG Zhengxian;WANG Ying;CAI Xiaodan;DONG Ligang;ZHANG Zitian;JIANG Xian;LI Hua;XU Yueqian(College of Information and Electronic Engineering,Zhejiang Gongshang University,Hangzhou 310018,China;UTStarcom,Hangzhou 310059,China;YingXian School of Philanthropy,Zhejiang Gongshang University,Hangzhou 310018,China)
出处
《电信科学》
北大核心
2024年第11期27-39,共13页
Telecommunications Science
基金
国家自然科学基金资助项目(No.62301488)
浙江省新型网络标准与应用技术重点实验室项目(No.2013E10012)
浙江省自然科学基金项目(No.LZ23F010003)
2022年度浙江省高等教育学会高等教育研究重点立项课题项目(No.KT2022017)
浙江工商大学省属高校基本科研业务费项目(No.QRK23010)
浙江工商大学“数字+”学科建设管理项目重大项目(No.SZJ2022A003)。
关键词
数字孪生
信道建模
生成对抗网络
智能通信网络
digital twin
channel modeling
generative adversarial network
intelligent communication network
