Heterogeneous and Inexact: Maximizing Power Efficiency of Edge Computing Sensors for Health Monitoring Applications

Abstract

Abstract-In the Internet-of-Things (IoT) era, there is an increasing trend to enable intelligent behavior in edge computing sensors. Thus, a new generation of smart wearable devices for health monitoring is being developed, able to perform complex Digital Signal Processing (DSP) routines that extract features of clinical relevance from the acquired data. These new edge computing sensors for personalized healthcare must operate within a tight energy envelope; addressing the ensuing challenge, we herein introduce an inexact and heterogeneous edge computing architecture, specifically tailored to the bio-DSP domain. We observe that bio-signal analysis applications present task-level parallelism, intensive computational hotspots and a high degree of resilience towards errors. These characteristics drive our new bio-DSP edge node architecture design composed of multiple processing cores, a Coarse-Grained Reconfigurable Array (CGRA) accelerator, and hardware-software co-design support to become resilient to a non-zero probability of bit-flips at runtime. All these characteristics enable our new bio-DSP architecture to operate with an ultra-low voltage operating point. Indeed our results indicate that the energy benefits attained from the inclusion of all these characteristics in bio-DSP architectures are more than additive: task parallelism is harnessed both at the processor and the accelerator level, and the high tolerance of the CGRA towards voltage down-scaling is exploited to further decrease the IoT edge bio-DSP system energy envelope.

Edge computing in smart health care systems: Review, challenges, and research directions

“…The MEC system and attached decision support system therefore only has to process and give feedback for the abnormal values instead of analyzing a bulk of normal values. A reduction technique that lowers the computing complexity, called inexact computing, is used in conjunction with morphological filtering to reduce data processing for ECG data compared with zero‐error computing 114 . Data reduction is also needed for the diagnosis of medical images, which often contain too high a resolution to be sent for real‐time analysis.…”

“…A reduction technique that lowers the computing complexity, called inexact computing, is used in conjunction with morphological filtering to reduce data processing for ECG data compared with zero-error computing. 114 Data reduction is also needed for the diagnosis of medical images, which often contain too high a resolution to be sent for real-time analysis. A solution proposed for this problem is compressed cellular neural networks, 115 which are superior to CNN in cases involving image processing tasks on an edge device.…”

The Extreme Edge at the Bottom of the Internet of Things: A Review

“…What is a more common tendency is bringing more intelligence to the smart sensors, which is necessary for the global Internet of Things integration [38]. The sensor layer of the IoT is the one that will count on the highest amount of devices (billions in principle), and these devices will be deployed everywhere, most of the time unattended.…”

“…In the context of wearable devices for health care, Basu et al [38] present a solution for smart wearable devices. In this case, the authors take advantage of the inherent tasklevel parallelism of the bio-signals.…”