基于深度学习的MIMO系统联合优化
|
摘要
自动编码器神经网络可将通信系统重新构建为端到端的任务,从而实现整个系统的联合优化。针对基于深度学习的2用户与4用户多输入多输出(multiple-input multiple-output,MIMO)系统联合优化问题,提出将自动编码器运用到系统中,将整个通信系统的发射端和接收端视为自动编码器的编码和译码部分,利用交叉熵损失加权和函数进行训练学习,从而获得优化的系统模型,并进一步分析得出每个用户的误比特率及所有用户的平均误比特率。实验结果表明,基于自动编码器所构建的MIMO通信系统相比于传统的通信系统具有更优的系统性能。
Reconstructing the communication system as an end-to-end task,the autoencoder neural network enables the joint optimization of the whole system. In order to deal with the joint optimization problem of MIMO( multiple-input multiple-output) systems based on deep learning with two users and four users respectively,the autoencoder was applied to the communication system. The encoding and decoding parts of autoencoder were considered as the transmitter and receiver of the overall communication system respectively. The cross-entropy loss weighted sum functions were applied to train and learn,so that the optimized system model was obtained and the bit error ratio of each user and average bit error ratio of all users were further analyzed and achieved. Compared with the traditional communication system,the experiment results show that MIMO communication system based on the autoencoder demonstrates better system performance.
Reconstructing the communication system as an end-to-end task,the autoencoder neural network enables the joint optimization of the whole system. In order to deal with the joint optimization problem of MIMO( multiple-input multiple-output) systems based on deep learning with two users and four users respectively,the autoencoder was applied to the communication system. The encoding and decoding parts of autoencoder were considered as the transmitter and receiver of the overall communication system respectively. The cross-entropy loss weighted sum functions were applied to train and learn,so that the optimized system model was obtained and the bit error ratio of each user and average bit error ratio of all users were further analyzed and achieved. Compared with the traditional communication system,the experiment results show that MIMO communication system based on the autoencoder demonstrates better system performance.
引文
[1]WANG Tianqi,WEN Chaokai,WANG Hangqing,et al.Deep learning for wireless physical layer:Opportunities and challenges[J].Wireless Communication Over Zigbee for Automotive Inclination Measurement China Communications,2017,14(11):92-111.
[2]张静,金石,温朝凯,等.基于人工智能的无线传输技术最新研究进展[J].电信科学,2018,34(08):46-55.ZHANG Jing,JIN Shi,WEN Chaokai,et al.Latest research progress of wireless transmission technology based on artificial intelligence[J].Telecommunications Science,2018,34(08):46-55.
[3]ERPEK T,O'SHEA T J,CLANCY T C.Learning a Physical Layer Scheme for the MIMO Interference Channel[C]//2018 IEEE International Conference on Communications(ICC).Kansas City,Missouri,USA:IEEEPress,2018:1-5.
[4]DANIELS R,HEATH R W.Online Adaptive Modulation and Coding with Support Vector Machines[C]//2010European Wireless Conference(EW).Lucca,Italy:IEEE Press 2010:718-724.
[5]HE Hengtao,WEN Chaokai,JIN Shi,et al.Deep Learning-Based Channel Estimation for Beamspace mm Wave Massive MIMO Systems[J].IEEE Wireless Communication Letters,2018,7(5):852-855.
[6]YE H,LI G Y,JUANG B H.Power of deep learning for channel estimation and signal detection in OFDM systems[J].IEEE Wireless Communications Letters,2018,7(1):114-117.
[7]FARSAD N,GOLDSMITH A.Neural network detection of data sequences in communication systems[J].IEEETransactions on Signal Processing,2018,66(21):5663-5678.
[8]YE H,LI G Y.Initial results on deep learning for joint channel equalization and decoding[C]//2017 IEEE 86t Vehicular Technology Conference(VTC).Toronto,Ontario,Canada:IEEE Press,2017:1-5.
[9]JEON Y S,HONG S N,LEE N.Blind detection for MI-MO systems with low resolution ADCs using supervised le-arning[C]//2017 IEEE International Conference on Communications(ICC).Paris,France:IEEE Press,2017:1-6.
[10]XU Zongben,SUN Jian.Model-driven deep-learning[J].National Science Review,2017,5(1):22-24.
[11]O’SHEA T,HOYDIS J.An introduction to deep learning for the physical layer[J].IEEE Transactions on Cognitive Communications and Networking,2017,3(4):563-575.
[12]WANG C,AU E K S,MURCH R D,et al.On the performance of the MIMO zero-forcing receiver in the presence of channel estimation error[J].IEEE Transactions on Wireless Communications,2007,6(3):805-810.
[13]ERASLAN E,DANESHRAD B,LOU C Y.Performance indicator for MIMO MMSE receivers in the presence of channel estimation error[J].IEEE Wireless Communications Letters,2013,2(2):211-214.
[14]GUO Jingcong,BURATTI L.On Adam Trained Models and a Parallel Method to Improve the Generalization Performance[C]//2018 IEEE/ACM Machine Learning in HPC Environments(MLHPC).Dallas,Texas,USA:IEEE Press,2018:85-94.
[2]张静,金石,温朝凯,等.基于人工智能的无线传输技术最新研究进展[J].电信科学,2018,34(08):46-55.ZHANG Jing,JIN Shi,WEN Chaokai,et al.Latest research progress of wireless transmission technology based on artificial intelligence[J].Telecommunications Science,2018,34(08):46-55.
[3]ERPEK T,O'SHEA T J,CLANCY T C.Learning a Physical Layer Scheme for the MIMO Interference Channel[C]//2018 IEEE International Conference on Communications(ICC).Kansas City,Missouri,USA:IEEEPress,2018:1-5.
[4]DANIELS R,HEATH R W.Online Adaptive Modulation and Coding with Support Vector Machines[C]//2010European Wireless Conference(EW).Lucca,Italy:IEEE Press 2010:718-724.
[5]HE Hengtao,WEN Chaokai,JIN Shi,et al.Deep Learning-Based Channel Estimation for Beamspace mm Wave Massive MIMO Systems[J].IEEE Wireless Communication Letters,2018,7(5):852-855.
[6]YE H,LI G Y,JUANG B H.Power of deep learning for channel estimation and signal detection in OFDM systems[J].IEEE Wireless Communications Letters,2018,7(1):114-117.
[7]FARSAD N,GOLDSMITH A.Neural network detection of data sequences in communication systems[J].IEEETransactions on Signal Processing,2018,66(21):5663-5678.
[8]YE H,LI G Y.Initial results on deep learning for joint channel equalization and decoding[C]//2017 IEEE 86t Vehicular Technology Conference(VTC).Toronto,Ontario,Canada:IEEE Press,2017:1-5.
[9]JEON Y S,HONG S N,LEE N.Blind detection for MI-MO systems with low resolution ADCs using supervised le-arning[C]//2017 IEEE International Conference on Communications(ICC).Paris,France:IEEE Press,2017:1-6.
[10]XU Zongben,SUN Jian.Model-driven deep-learning[J].National Science Review,2017,5(1):22-24.
[11]O’SHEA T,HOYDIS J.An introduction to deep learning for the physical layer[J].IEEE Transactions on Cognitive Communications and Networking,2017,3(4):563-575.
[12]WANG C,AU E K S,MURCH R D,et al.On the performance of the MIMO zero-forcing receiver in the presence of channel estimation error[J].IEEE Transactions on Wireless Communications,2007,6(3):805-810.
[13]ERASLAN E,DANESHRAD B,LOU C Y.Performance indicator for MIMO MMSE receivers in the presence of channel estimation error[J].IEEE Wireless Communications Letters,2013,2(2):211-214.
[14]GUO Jingcong,BURATTI L.On Adam Trained Models and a Parallel Method to Improve the Generalization Performance[C]//2018 IEEE/ACM Machine Learning in HPC Environments(MLHPC).Dallas,Texas,USA:IEEE Press,2018:85-94.