Skip to main content

Advertisement

SpringerLink
Log in

Hardware-Optimized Reservoir Computing System for Edge Intelligence Applications

  • Published:
Cognitive Computation Aims and scope Submit manuscript

Abstract

Edge artificial intelligence or edge intelligence is an ever-growing research area due to the current popularization of the Internet of Things. Unfortunately, incorporation of artificial intelligence (AI) in smart devices operating at the edge is a challenging task due to the power-hungry characteristics of deep learning implementations, such as convolutional neural networks (CNNs). As a feasible alternative, reservoir computing (RC) has attracted a lot of attention in the field of machine learning due to its promising performance in a wide range of applications. In this work, we propose a simple hardware-optimized circuit design of RC systems presenting high energy-efficiency capacities that fulfill the low power requirements of edge intelligence applications. As a proof of concept, we used the proposed design for the implementation of a low-power audio event detection (AED) application in FPGA. The measurements and simulation results obtained show that the proposed approach may provide significant accuracy with the advantage of presenting ultra-low-power characteristics (the energy efficiency estimated is below the microjoule per inference). These results make the proposed system optimal for edge intelligence applications in which energy efficiency and accuracy are the key issues.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6

Similar content being viewed by others

Notes

  1. Top-N accuracy is computed by interpreting as correct those predictions for which the ground truth is one of the N most likely categories.

References

  1. Chen D, Cong J, Gurumani S, Wm Hwu, Rupnow K, Zhang Z. Platform choices and design demands for IoT platforms: cost, power, and performance tradeoffs. IET Cyber-Physical Systems: Theory & Applications 2016;1(1):70–77.

    Article  Google Scholar 

  2. O’Leary DE. Artificial intelligence and big data. IEEE Intelligent Systems 2013;28(2):96–99.

    Article  Google Scholar 

  3. Silva DA, Stuchi JA, Violato RPV, Cuozzo LGD. Exploring convolutional neural networks for voice activity detection. Cognitive technologies. Springer; 2017. p. 37–47.

  4. Zhang Y, Suda N, Lai L, Chandra V. 2017. Hello edge: keyword spotting on microcontrollers. arXiv:171107128.

  5. Liu B, Wang Z, Zhu W, Sun Y, Shen Z, Huang L, Li Y, Gong Y, Ge W. An ultra-low power always-on keyword spotting accelerator using quantized convolutional neural network and voltage-domain analog switching network-based approximate computing. IEEE Access 2019;7:186456–186469.

    Article  Google Scholar 

  6. Maass W, Natschläger T, Markram H. Real-time computing without stable states: a new framework for neural computation based on perturbations. Neural Computation 2002;14(11):2531–2560.

    Article  MATH  Google Scholar 

  7. Jaeger H, Haas H. Harnessing nonlinearity: predicting chaotic systems and saving energy in wireless communication. Science 2004;304(5667):78–80.

    Article  Google Scholar 

  8. Lukoševičius M, Jaeger H. Reservoir computing approaches to recurrent neural network training. Computer Science Review 2009;3(3):127–149.

    Article  MATH  Google Scholar 

  9. Maass W. Liquid state machines: motivation, theory, and applications. Computability in context: computation and logic in the real world. World Scientific; 2011. p. 275–296.

  10. Appeltant L, Soriano MC, Van der Sande G, Danckaert J, Massar S, Dambre J, Schrauwen B, Mirasso CR, Fischer I. Information processing using a single dynamical node as complex system. Nature Communications 2011;2(1):1–6.

    Article  Google Scholar 

  11. Scardapane S, Butcher J, Bianchi F, Malik Z. Advances in biologically inspired reservoir computing. Cognitive Computation 2017;9(3):295–296.

    Article  Google Scholar 

  12. Katumba A, Freiberger M, Bienstman P, Dambre J. A multiple-input strategy to efficient integrated photonic reservoir computing. Cognitive Computation 2017;9(3):307–314.

    Article  Google Scholar 

  13. Gallicchio C, Micheli A. Echo state property of deep reservoir computing networks. Cognitive Computation 2017;9(3):337–350.

    Article  Google Scholar 

  14. Alomar M, Skibinsky-Gitlin ES, Frasser CF, Canals V, Isern E, Roca M, Rosselló JL. 2017. Efficient parallel implementation of reservoir computing systems. Neural Comput & Applic: 32 1–15.

  15. Rodan A, Tino P. Minimum complexity echo state network. IEEE Trans Neural Netw 2010; 22(1):131–144.

    Article  Google Scholar 

  16. Alomar ML, Soriano MC, Escalona-Morán M, Canals V, Fischer I, Mirasso CR, Rosselló JL. Digital implementation of a single dynamical node reservoir computer. IEEE Transactions on Circuits and Systems II: Express Briefs 2015;62(10):977–981.

    Google Scholar 

  17. Penkovsky B, Larger L, Brunner D. Efficient design of hardware-enabled reservoir computing in fpgas. Journal of Applied Physics 2018;124(16):162101.

    Article  Google Scholar 

  18. Larger L, Soriano MC, Brunner D, Appeltant L, Gutiérrez JM, Pesquera L, Mirasso CR, Fischer I. Photonic information processing beyond turing: an optoelectronic implementation of reservoir computing. Optics Express 2012;20(3):3241–3249.

    Article  Google Scholar 

  19. Skibinsky-Gitlin ES, Alomar ML, Isern E, Roca M, Canals V, Rossello JL. Reservoir computing hardware for time series forecasting. 2018 28th international symposium on power and timing modeling, optimization and simulation (PATMOS). IEEE; 2018. p. 133–139.

  20. Skibinsky-Gitlin ES, Alomar ML, Canals V, Frasser CF, Isern E, Galán-Prado F, Morán A, Roca M, Rosselló JL. Fpga-based echo-state networks. International conference on time series and forecasting. Springer; 2018. p. 135–146.

  21. Salamon J, Jacoby C, Bello JP. A dataset and taxonomy for urban sound research. Proceedings of the 22nd ACM international conference on multimedia; 2014. p. 1041–1044.

  22. Costoya AM, Frasser CF, Roca M, Rossello JL. 2019. Energy-efficient pattern recognition hardware with elementary cellular automata. IEEE Transactions on Computers.

  23. Jaeger H. 2001. The echo state approach to analysing and training recurrent neural networks. GMD Report.

  24. Maass W, Natschläger T, Markram H. Real-time computing without stable states: a new framework for neural computation based on perturbations. Neural Comput 2002;14(11):2531– 2560.

    Article  MATH  Google Scholar 

  25. Matsubara T, Torikai H, Hishiki T. A generalized rotate-and-fire digital spiking neuron model and its on-fpga learning. IEEE Transactions on Circuits and Systems II: Express Briefs 2011;58(10):677–681.

    Google Scholar 

  26. Nouri M, Jalilian M, Hayati M, Abbott D. A digital neuromorphic realization of pair-based and triplet-based spike-timing-dependent synaptic plasticity. IEEE Transactions on Circuits and Systems II: Express Briefs 2018;65(6):804–808.

    Google Scholar 

  27. Yilmaz O. Symbolic computation using cellular automata-based hyperdimensional computing. Neural Comput 2015;27(12):2661–2692.

    Article  MathSciNet  MATH  Google Scholar 

  28. Corporation A. 2008. Nios ii processor reference handbook.

  29. Xilinx I. 2006. Microblaze processor reference guide. reference manual 23.

  30. Han W, Chan CF, Choy CS, Pun KP. An efficient mfcc extraction method in speech recognition. 2006 IEEE international symposium on circuits and systems. IEEE; 2006. p. 4–pp.

  31. Meyer M, Cavigelli L, Thiele L. 2017. Efficient convolutional neural network for audio event detection. arXiv:170909888.

  32. Salamon J, Bello JP. Deep convolutional neural networks and data augmentation for environmental sound classification. IEEE Signal Processing Letters 2017;24(3):279–283.

    Article  Google Scholar 

  33. Kücüktopcu O, Masazade E, Ünsalan C, Varshney PK. A real-time bird sound recognition system using a low-cost microcontroller. Appl Acoust 2019;148:194–201.

    Article  Google Scholar 

  34. Schaetti N, Salomon M, Couturier R. Echo state networks-based reservoir computing for mnist handwritten digits recognition. 2016 IEEE Intl conference on computational science and engineering (CSE) and IEEE intl conference on embedded and ubiquitous computing (EUC) and 15th intl symposium on distributed computing and applications for business engineering (DCABES). IEEE; 2016. p. 484–491.

  35. Saleh R, Wilton S, Mirabbasi S, Hu A, Greenstreet M, Lemieux G, Pande PP, Grecu C, Ivanov A. System-on-chip: reuse and integration. Proc IEEE 2006;94(6):1050–1069.

    Article  Google Scholar 

  36. Escalona-Morán MA, Soriano MC, Fischer I, Mirasso CR. Electrocardiogram classification using reservoir computing with logistic regression. IEEE Journal of Biomedical and health Informatics 2014;19 (3):892–898.

    Article  Google Scholar 

  37. Patra JC, Pal RN. A functional link artificial neural network for adaptive channel equalization. Signal Process 1995;43(2):181–195.

    Article  MATH  Google Scholar 

  38. Inoue T, Vinayavekhin P, Wang S, Wood D, Greco N, Tachibana R. 2018. Domestic activities classification based on cnn using shuffling and mixing data augmentation. DCASE 2018 Challenge.

Download references

Funding

This work has been funded by a research grant from Endura Technologies and by the Ministerio de Ciencia e Innovación (MICINN/FEDER, UE), Spain, under project TEC2017-84877-R.

Author information

Authors and Affiliations

  1. Universitat de les Illes Balears, Palma de Mallorca, 07122, Spain

    Alejandro Morán, Vincent Canals, Fabio Galan-Prado, Christian F. Frasser & Josep L. Rosselló

  2. Balearic Islands Health Research Institute, Palma de Mallorca, 07010, Spain

    Vincent Canals & Josep L. Rosselló

  3. Endura Technologies, Greater San Diego Area, CA, USA

    Dhinakar Radhakrishnan & Saeid Safavi

Authors
  1. Alejandro Morán
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Vincent Canals
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Fabio Galan-Prado
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Christian F. Frasser
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Dhinakar Radhakrishnan
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Saeid Safavi
    View author publications

    You can also search for this author in PubMed Google Scholar

  7. Josep L. Rosselló
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Josep L. Rosselló.

Ethics declarations

Conflict of Interest

The authors at the Balearic Islands University received a research grant from Endura Technologies (San Diego, CA, USA).

Ethical Approval

This article does not contain any studies with human participants or animals performed by any of the authors.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

This article belongs to the Topical Collection: Trends in Reservoir Computing

Guest Editors: Claudio Gallicchio, Alessio Micheli, Simone Scardapane, Miguel C. Soriano

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Morán, A., Canals, V., Galan-Prado, F. et al. Hardware-Optimized Reservoir Computing System for Edge Intelligence Applications. Cogn Comput 15, 1461–1469 (2023). https://doi.org/10.1007/s12559-020-09798-2

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s12559-020-09798-2

Keywords

Search

Navigation