Use this URL to cite or link to this record in EThOS: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.668216
Title: Synthesis and analysis of nonlinear, analog, ultra low power, Bernoulli cell based CytoMimetic circuits for biocomputation
Author: Papadimitriou, Konstantinos
ISNI:       0000 0004 5365 8906
Awarding Body: Imperial College London
Current Institution: Imperial College London
Date of Award: 2014
Availability of Full Text:
Access from EThOS:
Full text unavailable from EThOS. Please try the link below.
Access from Institution:
Abstract:
A novel class of analog BioElectronics is introduced for the systematic implementation of ultra-low power microelectronic circuits, able to compute nonlinear biological dynamics. This class of circuits is termed ``CytoMimetic Circuits'', in an attempt to highlight their actual function, which is mimicking biological responses, as observed experimentally. Inspired by the ingenious Bernoulli Cell Formalism (BCF), which was originally formulated for the modular synthesis and analysis of linear, time-invariant, high-dynamic range, logarithmic filters, a new, modified mathematical framework has been conceived, termed Nonlinear Bernoulli Cell Formalism (NBCF), which forms the core mathematical framework, characterising the operation of CytoMimetic circuits. The proposed nonlinear, transistor-level mathematical formulation exploits the striking similarities existing between the NBCF and coupled ordinary differential equations, typically appearing in models of naturally encountered biochemical systems. The resulting continuous-time, continuous-value, low-power CytoMimetic electronic circuits succeed in simulating with good accuracy cellular and molecular dynamics and found to be in very good agreement with their biological counterparts. They usually occupy an area of a fraction of a square millimetre, while consuming between hundreds of nanowatts and few tenths of microwatts of power. The systematic nature of the NBCF led to the transformation of a wide variety of biochemical reactions into nonlinear Log-domain circuits, which span a large area of different biological model types. Moreover, a detailed analysis of the robustness and performance of the proposed circuit class is also included in this thesis. The robustness examination has been conducted via post-layout simulations of an indicative CytoMimetic circuit and also by providing fabrication-related variability simulations, obtained by means of analog Monte Carlo statistical analysis for each one of the proposed circuit topologies. Furthermore, a detailed mathematical analysis that is carefully addressing the effect of process-parameters and MOSFET geometric properties upon subthreshold translinear circuits has been conducted for the fundamental translinear blocks, CytoMimetic topologies are comprised of. Finally, an interesting sub-category of Neuromorphic circuits, the ``Log-Domain Silicon Synapses'' is presented and representative circuits are thoroughly analysed by a novel, generalised BC operator framework. This leads to the conclusion that the BC operator consists the heart of such Log-domain circuits, therefore, allows the establishment of a general class of BC-based silicon synaptic circuits, which includes most of the synaptic circuits, implemented so far in Log-domain.
Supervisor: Drakakis, Emmanuel Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.668216  DOI: Not available
Share: