Sinyal Kodlama ve Yeniden Yapılandırma İçin Izhikevich Nöron Modelinin İncelenmesi

Year 2025, Volume: 1 Issue: 1, 7 - 11, 30.07.2025

Obeid Abdalla , Jean Luck Randrıanantenaına , Ahmet Yasin Baran , Recai Kılıç

Abstract

Nöromorfik sistemlerde geleneksel sinyal kodlama yöntemleri genellikle analog girişleri ateşleme dizilerine dönüştürmek için Sızıntılı Toplayıcı-Ateşleyici (Leaky Integrate-and-Fire – LIF) gibi basitleştirilmiş nöron modellerine dayanır. Ancak bu modeller, zamanla değişen sinyaller için genellikle yüksek ateşleme oranları gerektirir ve bu da enerji tüketimini artırır. Bu çalışmada, sinyal kodlaması için biyolojik anlamlılığı yüksek ve düşük güç tüketimine sahip Izhikevich (IZ) nöron modeli incelenmiştir. IZ nöronunun, giriş genliğini ateşleme zamanlamasına kodlayarak 1-bit Sigma-Delta (ΣΔ) modülatörüne benzer davranış sergilediği gösterilmiştir. Modelin performansını değerlendirmek için farklı sinyal-gürültü oranlarında (SNR) Gauss beyaz gürültüsü (GWN) eklenmiş sinüzoidal giriş sinyalleri IZ nöronuna uygulanmıştır. Oluşan ateşleme dizileri, ateşlemeler arası zaman aralıkları (interspike intervals – ISIs) analiz edilerek ve anlık ateşleme oranı tahmin edilerek çözülmüş; böylece orijinal sinyal yeniden yapılandırılmıştır. Kodlama yapısının dayanıklılığı, farklı SNR seviyelerinde yeniden yapılandırılan sinyalin ortalama kare hata (MSE) değeri ölçülerek değerlendirilmiştir. Sonuçlar, IZ modelinin gürültülü koşullarda dahi yüksek yeniden yapılandırma doğruluğunu koruduğunu göstermekte ve bu modeli nöromorfik sistemlerde dayanıklı, olay
temelli sinyal kodlaması için uygun bir seçenek haline getirmektedir.

Keywords

nöron model , Sinyal kodlama , Gauss beyaz gürültüsü (GWN). , Delta (ΣΔ) modülasyonu , Izhikevich (IZ) nöron modeli

Exploring The Izhikevich Neuron Model for Robust Signal Encoding and Reconstruction

Abstract

Traditional signal encoding in neuromorphic systems often relies on simplified neuron models such as the Leaky Integrateand-Fire (LIF) to convert analog inputs into spike trains. However, these models typically demand high firing rates for timevarying signals, resulting in increased energy consumption. In this work, we explore the Izhikevich (IZ) neuron model as a low-power, biologically inspired alternative for signal encoding. We show that the IZ neuron exhibits behavior analogous to a 1-bit Sigma-Delta (ΣΔ) modulator by encoding the input’s amplitude into spike timing. To assess its performance, sinusoidal input signals with added Gaussian white noise (GWN) at various signal-to-noise ratios (SNRs) have been applied to IZ neuron. The spike trains are decoded by analyzing interspike intervals (ISIs) and estimating the instantaneous firing rate to reconstruct the original signal. The robustness of the encoding scheme has been evaluated by measuring the mean squared error (MSE) of the reconstructed signal across SNR levels. Results indicate that the IZ model maintains high reconstruction fidelity under noisy conditions, demonstrating its suitability for robust, event-driven signal encoding in neuromorphic systems.

Keywords

neuron model , Izhıkevich , Signal encoding , Sigma-delta modulatiın , Gaussian white noise

References

• Kang, M., Lee, Y., Park, M.: Energy Efficiency of Machine Learning in Embedded Systems Using Neuromorphic Hardware. Electronics (Basel). 9, 1069 (2020). https://doi.org/10.3390/electronics9071069
• Yang, Y., Bartolozzi, C., Zhang, H.H., Nawrocki, R.A.: Neuromorphic electronics for robotic perception, navigation and control: A survey. Eng Appl Artif Intell. 126, 106838 (2023). https://doi.org/10.1016/j.engappai.2023.106838
• Rajendran, B., Sebastian, A., Schmuker, M., Srinivasa, N., Eleftheriou, E.: Low-Power Neuromorphic Hardware for Signal Processing Applications. (2019). https://doi.org/10.1109/MSP.2019.2933719
• Tavanaei, A., Ghodrati, M., Kheradpisheh, S.R., Masquelier, T., Maida, A.: Deep learning in spiking neural networks. Neural Networks. 111, 47–63 (2019). https://doi.org/10.1016/j.neunet.2018.12.002
• Auge, D., Hille, J., Mueller, E., Knoll, A.: A Survey of Encoding Techniques for Signal Processing in Spiking Neural Networks. Neural Process Lett. 53, 4693–4710 (2021). https://doi.org/10.1007/s11063-021-10562-2
• Brette, R.: Dynamics of one-dimensional spiking neuron models. J Math Biol. 48, 38–56 (2004). https://doi.org/10.1007/s00285-003-0223-9
• Nair, M. V, Indiveri, G.: An Ultra-Low Power Sigma-Delta Neuron Circuit. In: 2019 IEEE International Symposium on Circuits and Systems (ISCAS). pp. 1–5. IEEE (2019). https://doi.org/10.1109/ISCAS.2019.8702500
• Yoon, Y.C.: LIF and Simplified SRM Neurons Encode Signals Into Spikes via a Form of Asynchronous Pulse Sigma–Delta Modulation. IEEE Trans Neural Netw Learn Syst. 28, 1192–1205 (2017). https://doi.org/10.1109/TNNLS.2016.2526029
• Hovin, M., Wisland, D., Berg, Y., Marienborg, J.T., Lande, T.S.: Delta-sigma modulation in single neurons. In: 2002 IEEE International Symposium on Circuits and Systems. Proceedings (Cat. No.02CH37353). pp. V-617-V–620. IEEE. https://doi.org/10.1109/ISCAS.2002.1010779
• Thao, N.T., Rzepka, D., Miśkowicz, M.: Bandlimited Signal Reconstruction From Leaky Integrate-and-Fire Encoding Using POCS. IEEE Transactions on Signal Processing. 71, 1464–1479 (2023). https://doi.org/10.1109/TSP.2023.3256269
• Dazhi Wei, Harris, J.G.: Signal reconstruction from spiking neuron models. In: 2004 IEEE International Symposium on Circuits and Systems (IEEE Cat. No.04CH37512). pp. V-353-V–356. IEEE. https://doi.org/10.1109/ISCAS.2004.1329535
• Schrauwen, B., Van Campenhout, I.: BSA, a fast and accurate spike train encoding scheme. In: Proceedings of the International Joint Conference on Neural Networks, 2003. pp. 2825–2830. IEEE. https://doi.org/10.1109/IJCNN.2003.1224019
• Izhikevich, E.M.: Simple model of spiking neurons. IEEE Trans Neural Netw. 14, 1569–1572 (2003). https://doi.org/10.1109/TNN.2003.820440
• Izhikevich, E.M.: Which Model to Use for Cortical Spiking Neurons? IEEE Trans Neural Netw. 15, 1063–1070 (2004). https://doi.org/10.1109/TNN.2004.832719
• Janssen, E., van Roermund, A.: Basics of Sigma-Delta Modulation. Presented at the (2011). https://doi.org/10.1007/978-94-007-1387-1_2
• Ko, D., Wilson, C.J., Lobb, C.J., Paladini, C.A.: Detection of bursts and pauses in spike trains. J Neurosci Methods. 211, 145–158 (2012). https://doi.org/10.1016/j.jneumeth.2012.08.013
• Lánský, P., Rodriguez, R., Sacerdote, L.: Mean Instantaneous Firing Frequency Is Always Higher Than the Firing Rate. Neural Comput. 16, 477–489 (2004). https://doi.org/10.1162/089976604772744875