Current-Mode FPAA with CMRR Elimination and Low Sensitivity to Mismatch

Foundations of Computing and Decision Sciences

Rato

2019

Self Cite

This paper addresses the problem of effective processing using third generation neural networks. The article features two new models of spiking neurons based on the cusp catastrophe theory. The effectiveness of the models is demonstrated with an example of a network composed of three neurons solving the problem of linear inseparability of the XOR function. The proposed solutions are dedicated to hardware implementation using the Edge computing strategy. The paper presents simulation results and outlines further research direction in the field of practical applications and implementations using nanometer CMOS technologies and the current processing mode.

Section: Discussion and Summarymentioning

confidence: 99%

Spiking Neural Network Based on Cusp Catastrophe Theory

Huderek

Foundations of Computing and Decision Sciences

Rato

2019

Self Cite

“…It is worth mentioning, when discussing the current mode of network operation in the presynaptic area, that spiking neurons in hardware implementations are mainly found in the form of silicon structures designed using CMOS technologies [ 6 , 34 , 35 , 36 ]. The miniaturization of nanometer CMOS technologies forces the designers of such circuits to replace the voltage processing mode with the current mode [ 7 , 37 ]. An SNN application for amperometric techniques in the IoT is an excellent answer to the problem of the miniaturization of the technology for implementing modern neuroprocessors.…”

Section: Pattern Classificationmentioning

confidence: 99%

Spiking Neural Network with Linear Computational Complexity for Waveform Analysis in Amperometry

Huderek

Przyborowski

2021

Sensors

Self Cite

The paper describes the architecture of a Spiking Neural Network (SNN) for time waveform analyses using edge computing. The network model was based on the principles of preprocessing signals in the diencephalon and using tonic spiking and inhibition-induced spiking models typical for the thalamus area. The research focused on a significant reduction of the complexity of the SNN algorithm by eliminating most synaptic connections and ensuring zero dispersion of weight values concerning connections between neuron layers. The paper describes a network mapping and learning algorithm, in which the number of variables in the learning process is linearly dependent on the size of the patterns. The works included testing the stability of the accuracy parameter for various network sizes. The described approach used the ability of spiking neurons to process currents of less than 100 pA, typical of amperometric techniques. An example of a practical application is an analysis of vesicle fusion signals using an amperometric system based on Carbon NanoTube (CNT) sensors. The paper concludes with a discussion of the costs of implementing the network as a semiconductor structure.

“…Such voltage mode elements are often energy-expensive, [4], although they are used as components for example in SAR ADCs [5][6][7]. Hence, current mode circuits can be useful in analog to digital converters [8][9][10].…”

Section: Introductionmentioning

confidence: 99%

ADC based on a fully differential current mode integrator

Handkiewicz¹,

Kropidłowski

Analog Integr Circ Sig Process

et al. 2019

Self Cite

The paper presents current mode Analog to Digital Converters (ADC) for a standard digital CMOS technology in a nm scale. It is shown that such converters composed of a fully differential current mode integrator make it possible to achieve a several-bit resolution necessary for pipeline processing. The proposed fully differential structure provides a 1-bit higher resolution in comparison to a non-balanced structure. The goal of the paper is obtained without voltage increasing above the standard 1.2 V supply. The pipelined ADC, composed of two 5-bits converters, consumes 1.5 mW of power and its Walden figure-of-merit equals 104.34 fJ/conv.