Adaptive Learning-Based Compressive Sampling for Low-power Wireless Implants

Abstract

Implantable systems are nowadays being used to interface the human brain with external devices, in order to understand and potentially treat neurological disorders. The most predominant design constraints are the system's area and power. In this paper, we implement and combine advanced compressive sampling algorithms to reduce the power requirements of wireless telemetry. Moreover, we apply variable compression, to dynamically modify the device performance, based on the actual signal need. This paper presents an area-efficient adaptive system for wireless implantable devices, which dynamically reduces the power requirements yielding compression rates from 8× to 64×, with a high reconstruction performance, as qualitatively demonstrated on a human data set. Two different versions of the encoder have been designed and tested, one with and the second without the adaptive compression, requiring an area of 230×235 μm and 200 × 190 μm, respectively, while consuming only 0.47 μW at 0.8 V. The system is powered by a 4-coil inductive link with measured power transmission efficiency of 36%, while the distance between the external and internal coils is 10 mm. Wireless data communication is established by an OOK modulated narrowband and an IR-UWB transmitter, while consuming 124.2 pJ/bit and 45.2 pJ/pulse, respectively.Index Terms-Implantable integrated circuit, area-efficient, low-power, compressive sensing, neural signals, learning-based digital signal processing, signal recovery.

An AC-Coupled Wideband Neural Recording Front-End With Sub-1 mm2×fJ/conv-step Efficiency and 0.97 NEF

“…The power consumption is the lowest [7] JSSC'15 [13] JSSC'17 [9] JSSC'18 [4] SSCL'18 [14] JSSC'18 [8] TBCAS'19 [15] TBCAS'20 [10] This work 2 On-chip decimation filter. 3 Off-chip decimation filter. 4 ADC only.…”

“…Recording and decoding high-frequency neural features through intracortical brain-computer interfaces has allowed accurate control of complex actuators [1]. Moving from laboratory demonstrations to widespread use of such systems requires combining signal acquisition and processing in an implantable system on chip [2], [3], under stringent energy and size constraints. Therefore, accommodating the maximum number of electrodes requires the lowest energy-area cost per recording channel.…”

Challenges and Perspectives on Impulse Radio-Ultra-Wideband Transceivers for Neural Recording Applications

“…Among different DPAs, Class D power amplifiers are more prevalent due to their superior power efficiency [60], [64], [72]. Class-C PAs are also adopted in some designs [45], [73]- [75] with a large inductor in their drain and a tuned LC tank to reduce the distortion of inherently nonlinear amplifiers. A power oscillation is introduced in [44] to maximize energy efficiency, which is connected directly to an on-chip dipole antenna.…”

A 16-Channel Neural Recording System-on-Chip With CHT Feature Extraction Processor in 65-nm CMOS

“…Therefore, the correct subset must be learned via a machine learning algorithm that maximizes the classification performance. The selection problem can be formally described similar to learning-based compressive subsampling (LBCS) as presented in [23]. Consider the following mathematical model where a signal x ∈ R N is converted to its compressed representation y ∈ R M with M < N :…”