Doppler Channel Series Prediction Using Recurrent Neural Networks 


Vol. 43,  No. 4, pp. 629-636, Apr.  2018
10.7840/kics.2018.43.4.629


PDF Full-Text
  Abstract

In fast time-varying channel environments, pilot resources should be used frequently to track the channel variation, which leads to lower time-efficiency. In this paper, we propose a recurrent-neural-network (RNN)-based channel prediction method to improve both time-efficiency and system performance compared to the conventional pilot-based channel estimator. During the channel training period, the channels are estimated based on the pilot resources, and the estimated channels are used to train the RNN. In the actual channel prediction period, only the RNN (not the pilots) is used to predict the channel series. Our proposed method can be easily applied to the conventional pilot-based channel estimator and improves time-efficiency by achieving lower bit-error rate (BER) performance within a certain prediction period.

  Statistics
Cumulative Counts from November, 2022
Multiple requests among the same browser session are counted as one view. If you mouse over a chart, the values of data points will be shown.


  Cite this article

[IEEE Style]

S. Jo, J. Sohn, D. Han, J. Moon, H. Ahn, "Doppler Channel Series Prediction Using Recurrent Neural Networks," The Journal of Korean Institute of Communications and Information Sciences, vol. 43, no. 4, pp. 629-636, 2018. DOI: 10.7840/kics.2018.43.4.629.

[ACM Style]

Sunyoung Jo, Jy-yong Sohn, Dong-Jun Han, Jaekyun Moon, and Hyunjun Ahn. 2018. Doppler Channel Series Prediction Using Recurrent Neural Networks. The Journal of Korean Institute of Communications and Information Sciences, 43, 4, (2018), 629-636. DOI: 10.7840/kics.2018.43.4.629.

[KICS Style]

Sunyoung Jo, Jy-yong Sohn, Dong-Jun Han, Jaekyun Moon, Hyunjun Ahn, "Doppler Channel Series Prediction Using Recurrent Neural Networks," The Journal of Korean Institute of Communications and Information Sciences, vol. 43, no. 4, pp. 629-636, 4. 2018. (https://doi.org/10.7840/kics.2018.43.4.629)

Search
(IN TITLE, AUTHOR, ABSTRACT,KEYWORDS)

Advanced Search ▼

POPULAR KEYWORDS
(TOP 10 KEYWORDS)

Recent Publications
(LAST 3 YEARS)






pISSN : 1226 - 4717
eISSN : 2287 - 3880