Adaptive Compressed Integrate-and-Fire Time Encoding Machine

TL;DR

The paper introduces ACIF-TEM, a novel adaptive compressed time encoding system that combines adaptive and compressed sampling techniques, resulting in more efficient signal encoding with lower error and fewer bits.

Contribution

It presents a new integrated sampler architecture that combines adaptive and compressed time encoding, along with an efficient clockless TDC design for improved performance.

Findings

Achieves at least 3-bit compression over AIF-TEM.

Provides 60% compression over IF-TEM for fixed MSE.

Demonstrates lower MSE with fewer bits on real audio signals.

Abstract

Integrate-and-Fire Time Encoding Machine (IF-TEM) is a power-efficient asynchronous sampler that converts analog signals into non-uniform time-domain samples. Adaptive IF-TEM (AIF-TEM) improves this machine by adapting its process to the characteristics of the input signal, thereby reducing the sampling rate. Compressed IF-TEM (CIF-TEM) reduces bit usage by performing analog compression before quantization. In this paper, we introduce a combined Adaptive Compressed IF-TEM (ACIF-TEM) -- a new sampler that leverages the two machines, AIF-TEM and CIF-TEM, where each reinforces the effectiveness of the other. We propose an efficient adaptive clockless time-to-digital converter (TDC) architecture for the novel sampler that integrates the compression stage within the TDC, facilitating the realization of the intended integrated system. \ifconf \else We analyze the total bit usage, and contrast its performance with that of IF-TEM, AIF-TEM, and CIF-TEM.\fi Via an evaluation study, we demonstrate that the proposed ACIF-TEM sampler achieves lower Mean Square Error (MSE) with fewer bits, offering compression gains of at least 3-bit out of 9-bits over AIF-TEM and 60\% compression over IF-TEM, for fixed recovery MSE with real audio signals.