Abstract
One of promising application offered by implant body area networks (BANs) is a capsule endoscope localization system. To begin with, this paper performs finite-difference time-domain (FDTD) simulations on implant BAN propagation with a numerical human model, and investigates the propagation characteristics of implant BAN signals at 400MHz medical implant communication service (MICS) band. Then, the paper presents a capsule endoscope localization system which utilizes only received signal strength indicator (RSSI) and two estimation methods, such as a maximum likelihood (ML) estimation method and a least squares (LS) method. Furthermore, we evaluate the two localization methods by two computer simulation scenarios. Our computer simulation results demonstrate that the ML localization can improve the location estimation accuracy as compared with the LS localization, that is, our performance comparison reveals that a careful consideration the propagation characteristics of implant BANs signals is efficient in terms of estimation performance improvement in capsule endoscope localization.
