Search

Nonvolatile memory (NVM)‐based neuromorphic computing has been attracting considerable attention from academia and the industry. Although it is not completely successful yet, remarkable achievements have been reported pertaining to synaptic devices that can leverage NVM capable of storing multiple states. The analog synaptic devices performing computation similar to biological nerve systems are crucial in energy‐efficient analog neuromorphic computing systems. To use NVM as an analog synaptic device, researchers focus on improving device characteristics. Among various characteristics, the most challenging one is linearity and symmetry of synaptic weight update that is required for on‐chip training. In this regard, this review paper discusses recent synaptic device improvements focusing on novel schemes tailored for each NVM device to improve the linearity and symmetry. In addition to device‐level studies, recent research achievements are reviewed expanded up to chip‐level studies because in realizing neuromorphic hardware systems beyond a single synaptic device, several considerations and requirements are needed to confirm for high‐level design, and accordingly, cooptimize among synaptic devices, synapse arrays, electrical circuits, neural networks, algorithms, and implementation. Also, this review introduces various circuit and algorithmic approaches to compensate for the non‐ideality of the analog synaptic device. [ABSTRACT FROM AUTHOR]

This paper proposes a novel signal quality assessment method for quasi-periodic cardiovascular signals, chiefly focus on the photoplethysmogram (PPG). The proposed method utilizes the fact that most cardiovascular signals are slowly time varying and thus morphological aspects of the two adjacent beats are almost identical. In order to implement this idea, the method first identifies pulse onset to divide the signal into several segments each of which contains one period of the signal. The segmented pulse signals having different pulse durations are then temporarily normalized by resampling them at a specific rate. Finally, the quality of the signals is evaluated as the signal similarity between the two adjacent segments. Optimal thresholds for the classification between high-and low-quality PPG signals are determined using the equal training sensitivity and specificity criterion. The proposed method is evaluated using a database where PPG signals are collected during a variety of activities such as cycling exercise. It attains a sensitivity of 97.9%, a specificity of 85.3%, and an accuracy of 93.8%, compared to manually annotated results. The promising results indicate that the proposed method is affordable to simply determine the quality of quasi-periodic cardiovascular signals, particularly PPG signals. In addition, based on the quasi-periodic characteristics of cardiovascular signals, the proposed method can also be used to indicate the reliability and the availability of the collected signals.

Wearable devices for body-status monitoring require various signal processing such as ambient noise filtering and signal quality evaluation. While wearable devices are very limited in resource, conventional noise filtering and signal quality evaluating methods consume considerable amount of processing power. Moreover, these conventional methods are not suitable for processing bio-signals in terms of its performance. In this paper, we propose a novel method of preprocessing bio-signals. This preprocessing method includes distortionless noise filtering and signal quality estimation, where both parts are basically based on a simple combination of multiple low pass IIR filters.

Wearable devices are becoming a natural and economic means to gather biometric data from end users. The massive amount of information that they will provide, unimaginable until a few years ago, owns an immense potential for applications such as continuous monitoring for personalized healthcare and use within fitness applications. Wearables are however heavily constrained in terms of amount of memory, transmission capability and energy reserve. This calls for dedicated, lightweight but still effective algorithms for data management. This paper is centered around lossy data compression techniques, whose aim is to minimize the amount of information that is to be stored on their onboard memory and subsequently transmitted over wireless interfaces. Specifically, we analyze selected compression techniques for biometric signals, quantifying their complexity (energy consumption) and compression performance. Hence, we propose a new class of codebook-based (CB) compression algorithms, designed to be energy efficient, online and amenable to any type of signal exhibiting recurrent patterns. Finally, the performance of the selected and the new algorithm is assessed, underlining the advantages offered by CB schemes in terms of memory savings and classification algorithms.