Life Science Journal 2010;7(2) http://www.americanscience.org editor@americanscience.org 91 Reliability of Wireless Body Area Networks used for Ambulatory Monitoring and Health Care Ali Peiravi 1 , Maria Farahi 2 Department of Electrical Engineering, School of Engineerinf, Ferdowsi University of Mashhad Mashhad IRAN. 1 Ali_peiravi@yahoo.com , 2 mari.252@gmail.com Abstract: Ambulatory monitoring and health care using wireless sensor networks is an active area of applied research. The general network topology used for wireless body area networks is the star topology with the sensor nodes sending their data to a central processing node for data fusion. Reliability of these networks is very important since they deal with human life. Reported applications have had performance and reliability problems. In this paper, several reported applications of wireless body area networks are reviewed and the reliability of a sample WBAN is computed. [Life Science Journal 2010;7(2):91-97]. (ISSN: 1097-8135). Keywords: Reliability, wireless body area network, ambulatory monitoring, MEMS sensors Introduction Measurement of human position, balance, posture, orientation and body status is not only of interest to the medical scientists but also of importance in the entertainment field for computer generated especial effects. The complexities of motion analysis and the various parameters involved in the estimation or measurement of body position and orientation usually require the use of a complex set of wireless sensor nodes in the form of a wireless body area network and including accelerometers, gyroscopes, magnetometers, etc. plus the application of data fusion. In some applications, estimation may require the inclusion of some form of a Kalman filter or a particle filter. Data fusion may be defined as the use of techniques to combine data from multiple sources and gather that information in order to achieve inferences that are more efficient and potentially more accurate than if they were achieved by means of a single source. There are various fusion processes that are usually described as low, intermediate or high level that depend on the processing stage at which data fusion takes place (Mandic et al. 2005). In low level data fusion several sources of raw data are combined to produce new raw data that is expected to be more useful than the inputs. In intermediate level data fusion that may also be called feature level fusion, various features are combined into a feature map that may then be used by further processing. High level data fusion usually refers to a situation where decisions coming from several experts are fused together. These include voting methods, statistical methods, fuzzy logic, etc. Another potential application of data fusion is in capturing motion. Such areas as interactive game and learning, animation, film special effects, health- care and navigation may be named. Human motion capture techniques using multiple high resolution cameras in especial studios are highly costly and complex. With the recent developments in MEMS technology, micro inertial sensors-on-hip (MMocap), low cost real-time human motion capture systems have become possible. Earlier works An early application of ambulatory measurements was reported by Tanka et al. (1994) for long term monitoring of posture. They argued that human postures such as standing, sitting, lying, walking, etc. may be estimated from the angles corresponding to gravitational direction in three portions of the body, namely the chest, the thigh and the legs which may be measured using tri-axial accelerometers as shown in Fig. 1. Fig 1 – Accelerometers and gyroscope for measuring body posture and walking speed adopted from Motoi et al. (2003). The angles in question may be obtained from the low frequency signals of these accelerometers. Later on, Motoi et al. (2003) suggested the addition of a gyroscope on the thigh for measuring walking speed whereby the angular change in the saggital plane is obtained by integrating the gyroscope's signal during walking. They concluded that fairly good results with a reasonable degree of accuracy may be obtained for walking speeds of less than 0.6m/s. What has caused major progress in this area in recent years is mainly due to the developments in wireless sensor networks, nanoelectronics, MEMS technology and advances in data fusion. Solaiman et al. (1999) presented a monosensor/multiple source data fusion system for the detection of the esophagus in