基于DDPM的盲解调实验系统设计  

Design of a Blind Demodulation Experimental System Based on DDPM

作　　者：臧淼[1] 邢志强[1] 李响 ZANG Miao;XING Zhiqiang;LI Xiang(School of Information Science and Technology,North China University of Technology,Beijing 100144,China)

机构地区：[1]北方工业大学信息学院,北京100144

出　　处：《实验室研究与探索》2025年第1期91-95,170,共6页Research and Exploration In Laboratory

基　　金：国家自然科学基金项目(61771169);2022年北京市高等教育本科教学改革创新项目(202210009002);2024年北方工业大学教改项目(108051360024XN142-4)。

摘　　要：盲解调技术因其在非协作通信领域的重要应用而广受关注,因其理论的抽象、复杂性成为课程教学中的一个难点。基于Matlab设计一套非协作通信环境下的盲解调系统。系统涵盖载波频率估计、符号速率估计、调制方式识别、载波同步、符号同步以及盲均衡等多个模块。通过采用去噪扩散概率模型(DDPM)对调制信号星座图样点去噪,基于最小均方误差自动识别多进制相移键控的调制方式。实验结果表明,该盲解调系统能有效估计信号参数、提取信号星座图和识别调制方式,成功解调还原信号,并保持较低的误码率。该系统具有较强的综合性、直观性和可扩展性,有利于学生理解盲解调系统的工作原理和DDPM的应用场景,培养学生的工程实践能力和创新思维。ion and complexity pose challenges in related course teaching.To address this issue,a blind demodulation system for non-cooperative communication was designed by using Matlab.This system encompasses several modules,including carrier frequency estimation,symbol rate estimation,modulation mode recognition,carrier synchronization,symbol synchronization,and blind equalization.Specifically,the system employs the DDPM(denoising diffusion probabilistic model)to denoise the constellation points of the modulated signal and automatically recognizes the MPSK(multiple phase shift keying)modulation modes based on LMSE(least mean square error).The experimental results demonstrate that the designed blind demodulation system can effectively estimate signal parameter information,extract signal constellation diagrams,recognize modulation modes,successfully demodulate and restore signals,and error rate is low.The system is characterized by its comprehensiveness,intuitiveness,and scalability,which facilitates students’understanding of the working principles of blind demodulation systems and the application scenarios of DDPM.It also fosters students’practical engineering skills and innovative thinking.

关 键 词：去噪扩散概率模型 盲解调 星座图 调制方式识别 

分 类 号：TN911.72[电子电信—通信与信息系统]