A Real-Time Missing Data Recovery Method Using Recurrent Neural Network for Multiple Transmissions

  • Bor-Shing LinEmail author
  • Yu-Syuan Lin
  • I-Jung Lee
  • Bor-Shyh Lin
Conference paper
Part of the Smart Innovation, Systems and Technologies book series (SIST, volume 109)


Data loss and recovery is a critical issue in data transmission. Traditional data recovery methods are impractical for use in real-time systems that require multiple transmissions. To solve this problem, this study proposed a recovery method based on a recurrent neural network, which is then used to build a pre-diction model. When a data gap occurs, the missing data can be recovered immediately using the predicted value. This method distributes the calculation and can immediately recover the data gap. Through a series of experiments, this study optimized different parameters in the neural network, thus optimizing the prediction model.


Missing data recovery Wearable technology Recurrent neural network 


  1. 1.
    Janssen, A., Veldhuis, R.: Adaptive interpolation of discrete-time signals that can be modeled as autoregressive processes. IEEE Trans. Acoust. Speech Signal Process. 34(2), 317–330 (1986)CrossRefGoogle Scholar
  2. 2.
    Esquef, P.A.A., Biscainho, L.W.P.: An efficient model-based multirate method for reconstruction of audio signals across long gaps. IEEE Trans. Audio Speech Lang. Process. 14(4), 1391–1400 (2006)CrossRefGoogle Scholar
  3. 3.
    Gunduzhan, E., Momtahan, K.: Linear prediction based packet loss concealment algorithm for PCM coded speech. IEEE Trans. Speech Audio Process. 9(8), 778–785 (2001)CrossRefGoogle Scholar
  4. 4.
    Milner, B., Semnani, S.: Robust speech recognition over IP networks. In: 2000 IEEE International Conference on Acoustics, Speech, and Signal Processing on Proceedings, vol. 3, pp. 1791–1794. IEEE, Istanbul (2000)Google Scholar
  5. 5.
    Milner, B.: Robust speech recognition in burst-like packet loss. In: 2001 IEEE International Conference on Acoustics, Speech, and Signal Processing on Proceedings, vol. 1, pp. 261–264. IEEE, Salt Lake City (2001)Google Scholar
  6. 6.
    Segarra, S., Marques, A.G., Leus, G., Ribeiro, A.: Sampling of graph signals: successive local aggregations at a single node. In: 2015 49th Asilomar Conference on Signals, Systems and Computers, pp. 1819–1823. IEEE, Pacific Grove (2015)Google Scholar
  7. 7.
    Kim, K.J., Agrawal, V., Gaunaurd, I., Gailey, R.S., Bennett, C.L.: Missing sample recovery for wireless inertial sensor-based human movement acquisition. IEEE Trans. Neural Syst. Rehabil. Eng. 24(11), 1191–1198 (2016)CrossRefGoogle Scholar
  8. 8.
    Hsiao, P.-C., Yang, S.-Y., Lin, B.-S., Lee, I-J., Chou, W.: Data glove embedded with 9-Axis IMU and force sensing sensors for evaluation of hand function. In: 37th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC 2015), pp. 4631–4634. IEEE, Milan (2015)Google Scholar
  9. 9.
    Lin, B.-S., Lee, I.-J., Yang, S.-Y., Lo, Y.-C., Lee, J., Chen, J.-L.: Design of inertial-sensor-based data glove for hand function evaluation. Sensors 18(5), 1–18 (2018). Article 1545CrossRefGoogle Scholar
  10. 10.
    Inertial measurement unit: In Wikipedia, the Free Encyclopedia. Accessed 14 July 2018
  11. 11.
    Madgwick, S.O.H., Harrison, A.J.L., Vaidyanathan, R.: Estimation of IMU and MARG orientation using a gradient descent algorithm. In: 2011 IEEE International Conference on Rehabilitation Robotics, pp. 1–7. IEEE, Zurich (2011)Google Scholar
  12. 12.
    Samarawickrama, A.J.P., Fernando, T.G.I.: A recurrent neural network approach in predicting daily stock prices an application to the Sri Lankan stock market. In: 2017 IEEE International Conference on Industrial and Information Systems (ICIIS), pp. 1–6. IEEE, Peradeniya (2017)Google Scholar
  13. 13.
    Xia, Y., Hu, R.: Fuzzy neural network based energy efficiencies control in the heating energy supply system responding to the changes of user demands. J. Netw. Intell. 2(2), 186–194 (2017)MathSciNetGoogle Scholar
  14. 14.
    Cho, K. et al.: Learning phrase representations using RNN encoder–decoder for statistical machine translation. In: Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing (EMNLP) on Proceedings, pp. 1724–1734. Association for Computational Linguistics, Doha (2014)Google Scholar
  15. 15.
    Kingma, D., Ba, J.: Adam: a method for stochastic optimization. In: 2015 International Conference on Learning Representations, pp. 1–13. arXiv, San Diego (2015)Google Scholar

Copyright information

© Springer Nature Switzerland AG 2019

Authors and Affiliations

  1. 1.Department of Computer Science and Information EngineeringNational Taipei UniversityNew Taipei CityTaiwan
  2. 2.College of Electrical Engineering and Computer ScienceNational Taipei UniversityNew Taipei CityTaiwan
  3. 3.Institute of Imaging and Biomedical PhotonicsNational Chiao Tung UniversityTainanTaiwan

Personalised recommendations