Revealing the Secrets of Spiking Neural Networks: The Case of Izhikevich Neuron

Garaffa, Luiza C.; Aljuffri, Abdullah; Reinbrecht, Cezar; Hamdioui, Said; Taouil, Mottaqiallah; Sepúlveda, Johanna

doi:10.1109/dsd53832.2021.00083

Cited by 8 publications

(3 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The Izhikevich Spiking Neural Network used, due to its dynamic configuration [43,44], can reproduce different spikes and different triggering behaviors of neurons. Specifically, the dynamics of the model were governed by two key variables [45,46]. The following equation describes the membrane potential [38,47]:…”

Section: The Proposed Artificial Immune Systemmentioning

confidence: 99%

An Ensemble Transfer Learning Spiking Immune System for Adaptive Smart Grid Protection

et al. 2022

View full text Add to dashboard Cite

The rate of technical innovation, system interconnection, and advanced communications undoubtedly boost distributed energy networks’ efficiency. However, when an additional attack surface is made available, the possibility of an increase in attacks is an unavoidable result. The energy ecosystem’s significant variety draws attackers with various goals, making any critical infrastructure a threat, regardless of scale. Outdated technology and other antiquated countermeasures that worked years ago cannot address the complexity of current threats. As a result, robust artificial intelligence cyber-defense solutions are more important than ever. Based on the above challenge, this paper proposes an ensemble transfer learning spiking immune system for adaptive smart grid protection. It is an innovative Artificial Immune System (AIS) that uses a swarm of Evolving Izhikevich Neural Networks (EINN) in an Ensemble architecture, which optimally integrates Transfer Learning methodologies. The effectiveness of the proposed innovative system is demonstrated experimentally in multiple complex scenarios that optimally simulate the modern energy environment. The most significant findings of this work are that the transfer learning architecture’s shared learning rate significantly adds to the speed of generalization and convergence approach. In addition, the ensemble combination improves the accuracy of the model because the overall behavior of the numerous models is less noisy than a comparable individual single model. Finally, the Izhikevich Spiking Neural Network used here, due to its dynamic configuration, can reproduce different spikes and triggering behaviors of neurons, which models precisely the problem of digital security of energy infrastructures, as proved experimentally.

show abstract

Section: The Proposed Artificial Immune Systemmentioning

confidence: 99%

An Ensemble Transfer Learning Spiking Immune System for Adaptive Smart Grid Protection

et al. 2022

View full text Add to dashboard Cite

show abstract

“…1 shows a typical setup of an EM side-channel attack targeting model (or input) recovery from an EdgeML device. The objective of the adversary under such settings can be listed as follows: [16] Architecture (activation only) MLP & CNN Batina et al [17] Input MLP Dong et al [18] Input MLP & CNN Jap et al [19] Architecture & Weights Decision trees FPGA Dubey et al [20] Weights BNN Yli-Mäyry et al [21] Architecture & Weights BNN Yu et al [22] Architecture BNN Yoshida et al [23] Weights MLP & CNN Garaffa et al [24] Weights a SNN Wei et al [25] Input…”

Section: Survey Of Em Side-channel Attacks On Edgemlmentioning

confidence: 99%

“…Different ML types can be implemented on these TPUs, including MLP and CNN. Garaffa et al [24] investigate the vulnerability of Spiking Neural Networks (SNN) against sidechannel attacks. They exploit the power leakage or the electromagnetic radiation from the target device to learn information on the spiking activity of the network, which is then combined with timing information to recover the weights of the network.…”

Section: B Field Programmable Gate Array (Fpga)mentioning

confidence: 99%

On (in)Security of Edge-based Machine Learning Against Electromagnetic Side-channels

Bhasin

Jap

Picek

2022

2022 IEEE International Symposium on Electromagnetic Compatibility &Amp; Signal/Power Integrity (EMCSI)

View full text Add to dashboard Cite

Machine (deep) learning represents mainstream research and development direction. This success can be linked to the ever-increasing computational capabilities and vast amounts of available data, resulting in ever more sophisticated machine learning models. The design and training of such machine learning models are challenging and expensive tasks, which incentivize companies to protect and keep it secret.Additionally, the wide applicability of machine learning results in diverse deployment scenarios. Many machine learning architectures are deployed on edge devices, such as sensors or actuators, making them susceptible to side-channel attacks. This work surveys the research works considering electromagnetic side-channel and edge-based machine learning models. After discussing state-of-the-art attacks and countermeasures, we propose several open problems to be investigated in future research.

show abstract

Placement of Chains of Real-Time Tasks on Heterogeneous Platforms under EDF Scheduling

Casini

Biondi

2022

2022 25th Euromicro Conference on Digital System Design (DSD)

View full text Add to dashboard Cite

Revealing the Secrets of Spiking Neural Networks: The Case of Izhikevich Neuron

Abstract: Spiking Neural Networks (SNNs) are a strong can-

Cited by 8 publications

References 13 publications

An Ensemble Transfer Learning Spiking Immune System for Adaptive Smart Grid Protection

An Ensemble Transfer Learning Spiking Immune System for Adaptive Smart Grid Protection

On (in)Security of Edge-based Machine Learning Against Electromagnetic Side-channels

Placement of Chains of Real-Time Tasks on Heterogeneous Platforms under EDF Scheduling

Contact Info

Product

Resources

About