An On-Time Power-Aware Scheduling Scheme for Medical Sensor SoC-Based WBAN Systems

Tae-Ho Hwang,Jung-Guk Kim,Dong-Sun Kim

doi:10.3390/s130100375

Abstract

The focus of many leading technologies in the field of medical sensor systems is on low power consumption and robust data transmission. For example, the implantable cardioverter-defibrillator (ICD), which is used to maintain the heart in a healthy state, requires a reliable wireless communication scheme with an extremely low duty-cycle, high bit rate, and energy-efficient media access protocols. Because such devices must be sustained for over 5 years without access to battery replacement, they must be designed to have extremely low power consumption in sleep mode. Here, an on-time, energy-efficient scheduling scheme is proposed that performs power adjustments to minimize the sleep-mode current. The novelty of this scheduler is that it increases the determinacy of power adjustment and the predictability of scheduling by employing non-pre-emptible dual priority scheduling. This predictable scheduling also guarantees the punctuality of important periodic tasks based on their serialization, by using their worst case execution time) and the power consumption optimization. The scheduler was embedded into a system on chip (SoC) developed to support the wireless body area network—a wakeup-radio and wakeup-timer for implantable medical devices. This scheduling system is validated by the experimental results of its performance when used with life-time extensions of ICD devices.

Highlights

Owing to the increase in today’s aging population there has been a gradually increasing demand for medical implantable sensor devices such as pacemakers and defibrillators
Since implanted medical devices are designed so that they will not need to be accessed for maintenance over a long period of time, high-level constraints on power consumption are needed that can sustain a device over several years with a duty-cycle of under 1%
In order to achieve low-power and reliable on-time periodic operations of a medical sensor device, management of the power mode of the system based on the prediction of the timing behaviours of tasks is most important

Summary

Introduction

Owing to the increase in today’s aging population there has been a gradually increasing demand for medical implantable sensor devices such as pacemakers and defibrillators. A major issue in this field has been the development of technologies for wireless networking between implantable devices [1,2,3]. IEEE 802.15.6 [4] is a wireless body area network (WBAN) standard for communications between medical implantable devices [5]. It specifies a frequency band and a Medium Access Control (MAC) protocol as its Physical Layer (PHY) for in-body and on-body applications. Since implanted medical devices are designed so that they will not need to be accessed for maintenance over a long period of time, high-level constraints on power consumption are needed that can sustain a device over several years with a duty-cycle of under 1%. The design of an extremely low power RF transceiver [6,7] and a low-power system on chip (SoC) has been an important issue

Methods

Results

Conclusion