SPAIC: Modelling of an Event-Driven Spike Encoding System

Using our measurement framework, data needs to be collected from the chip using which a realistic model needs to be developed. This model should quantify the non-idealities in the signal conditioning/processing pipeline preceding the spike encoders and their effects on the encoding.

• The primary objective of this project is to do a thorough literature review of the ”event-driven/level-crossing/asynchronous” type of ADCs and develop a realistic model as close as possible to the real circuit implementation.
• Study the effect of spread in voltage thresholds used for A/D conversion and offset of the digitizing comparators.
• Study the effect of noise accumulated in the signal conditioning chain.
• Study the effect of comparator delay/conversion delay and include this in the model
• Quantify accuracy of encoding and validate the model against existing silicon implementation

Requirements

We are looking for an enthusiastic MSc student who has some background and interest in analog circuit design and strong programming skills with python/Matlab.
Familiar with bench measurement of circuits/chips
Familiar with the programming of microcontroller boards (Arduino, Teensy etc)
This work could potentially lead to architectural/circuit level changes for the next generation of event-driven ADC implementation and someone interested in taking this forward with involvement in the circuit design/simulation would be the ideal candidate.

Contact

Shyam Narayanan (PhD Student) : shyam (at) ini.uzh.ch
Prof. Dr. Giacomo Indiveri : giacomo (at) ini.uzh.ch

© 2022 Institut für Neuroinformatik