LiteMWBAN: A Lightweight Middleware for Wireless Body Area Network Agustinus Borgy Waluyo, Isaac Pek, Song Ying, Jiankang Wu, Xiang Chen, Wee-Soon Yeoh Institute for Infocomm Research (I 2 R), A*STAR, Singapore {awaluyo, ipek, sying, jiankang, xchen, wsyeoh} @i2r.a-star.edu.sg 1 Abstract— Healthcare is considered as one of the most important agenda to address by many countries worldwide. Thanks to the recent development in wireless sensor technology, a continuous medical monitoring of the patients under their natural physiological states has been made possible and this will certainly help to detect transient life threatening from daily activities. With the growing variety of healthcare sensor devices, it is of interest to deploy a middleware system that shields the underlying technology differences from the application layer. The aim of this paper is to propose LiteMWBAN, a lightweight middleware for wireless medical body area network. The main features of the proposed middleware include support on: (i) multiple sensors and applications, (ii) plug and play, (iii) resource management and alert, and (iv) mobile applications. The prototype of the middleware has been built and its effective uses when deployed with a mobile healthcare application are shown in this paper. Index Terms—lightweight middleware, lightweight body sensor middleware, mobile healthcare middleware. I. INTRODUCTION OST countries worldwide has considered healthcare as one of the most important agenda to be addressed. This is driven by aging populations worldwide, which causes a significant challenge to the healthcare providers and the respective government to provide the best way possible to care for their citizens [13]. The latest technology development in wireless sensor devices will have a significant role in the healthcare industry. A group of medical sensor devices, which works together and communicates is commonly called wireless body area network (WBAN). Due to the large variety of sensors deployed by many types of application, it is natural to implement a middleware that shield off the underlying sensor hardware or OS/protocol stack from the applications so as to simplify the application development by utilizing reusable codes. As a progression of our earlier effort in building a middleware for personal computer applications [12], this paper proposes LiteMWBAN, a WBAN middleware for mobile applications. The proposed middleware is designed to support the following features: Manuscript received Feb. 01, 2008. This work was supported by Science and Engineering Research Council (SERC) of the Singapore Agency for Science, Technology and Research (A*Star) on Embedded & Hybrid Systems II (EHS-II) programme under grant 052-118-0058. • Data acquisition The middleware is designed to receive the raw data readings from the sensor nodes, processes the data and transmits them to the application. • Multiple Sensor and Application A different type of sensor and application may take place in the same WBAN. For instance, vital signal monitoring (e.g., ECG) and behavior monitoring can be considered as two applications in the WBAN. • Plug and play function Detection of when the sensor joining in or leaving off the network are supported by middleware. • Push and pull communication The middleware supports downstream communication from the sensor nodes as well as upstream communication from the application. • Sensor battery level reading The middleware provides a method to check and alert the power level of each of the sensor mote. • Resource Control and Management Since wireless sensor devices inherent limited power capacity, it is important to implement resource control and management such as sensor sleep and activation, get and set sampling rate, and read sensor’s battery level. • Lightweight The middleware is built to be deployed on mobile devices. This requires an optimized system design while retaining the performance. The proposed middleware has been developed and its prototype when used with a healthcare monitoring system is presented. II. RELATED WORK One of the most popular middleware systems is CORBA [3], which is used for large distributed computing network without constraints on system resources. Unlike the wireless sensor network whereby energy constrains is of much concern, CORBA is not suitable in this domain. Edwards [5] proposed “plug and play” network of different devices, but they are both TCP/IP based, whose performance may be a significant issue when it comes to WBAN. Other middleware such as Limbo [5] and FarGo [8] have managed to adapt to the changing properties of the wireless M 978-1-4244-2253-1/08/$25 ©2008 IEEE Proceedings of the 5th International Workshop on Wearable and Implantable Body Sensor Networks, in conjunction with The 5th International Summer School and Symposium on Medical Devices and Biosensors The Chinese University of Hong Kong, HKSAR, China. Jun 1-3, 2008 141