1006 IEEE TRANSACTIONS ON ANTENNAS AND PROPAGATION, VOL. 57, NO. 4, APRIL 2009 An Efficient FDTD Algorithm Based on the Equivalence Principle for Analyzing Onbody Antenna Performance Andrea Sani, Student Member, IEEE, Yan Zhao, Member, IEEE, Yang Hao, Senior Member, IEEE, Akram Alomainy, Member, IEEE, and Clive Parini, Member, IEEE Abstract—In this paper, onbody antenna performance and its ef- fect on the radio channel is analyzed. An efficient numerical tech- nique based on the finite-difference time-domain technique and the equivalence principle is developed. The proposed technique be- gins with the problem decomposition by separately computing the wearable antennas and onbody propagation involving the digital human phantom. The equivalence principle is used as an inter- face between the two computational domains. We apply this tech- nique to analyze onbody antenna and channel characteristics for three different planar body-worn antennas operating at the indus- trial-scientific-medical frequency band of 2.4 GHz. Simulated re- sults are validated with measurement data with good agreement. Index Terms—Body-area networks, body-worn antenna, equiv- alence principle, finite-difference time domain (FDTD). I. INTRODUCTION W IRELESS body-area networks (WBAN) have recently received increasing attention due to their promising ap- plications in medical sensors and personal entertainment sys- tems [1]–[4]. Ubiquitous and wearable computing is regarded as the central component of fourth-generation communication systems [5]. In order to design a spectrum and power-efficient onbody communication system, it is very important to under- stand radio channel characteristics, which was obtained by using experiments in the past [6]–[12]. A computational model based on the finite-difference time domain (FDTD) [13] was devel- oped, aiming to provide a physical insight of an onbody radio channel [14], [15]; however, a point-source approximation was used in the simulation and, hence, it is not a true representation of body-worn antennas. The FDTD technique [13] is a method based on the direct solution of the Maxwell’s curl equations, and is ideal to sim- ulate stratified dielectric objects, such as a human body. How- ever, if practical antennas are considered in the onbody radio channel, much higher spatial resolution is needed to accurately encounter small geometrical features in the antenna. This, in turn, increases the computation time and the memory require- ments if a uniform mesh scheme is used. Although the use of a Manuscript received January 31, 2008; revised November 19, 2008. Current version published April 08, 2009 The authors are with Queen Mary College, University of London, London E1 4NS, U.K. (e-mail: y.hao@elec.qmul.ac.uk). Color versions of one or more of the figures in this paper are available online at http://ieeexplore.ieee.org. Digital Object Identifier 10.1109/TAP.2009.2014581 Fig. 1. Numerical problem is split in two subproblems: 1) the antenna is an- alyzed in free space with CST Microwave Studio and 2) the human body is modeled with FDTD. The equivalence principle is used to interface the two computational domains. nonuniform mesh and a subriding scheme in FDTD can be ap- plied, it may result in spurious solutions or even suffer from instability for the subgriding scheme [16]. It is also possible to combine the FDTD and other numerical schemes, for ex- ample, the frequency-domain method of moments (FD-MoM) to increase numerical efficiency. In [17] and [18], the equiv- alence principle has been used to divide the original problem into two subproblems: the radiating element, ideally, a metallic structure is modeled using the MoM, while the surrounding environment uses the FDTD. In [20] and [21], a time-domain MoM (TD-MoM) is used to analyze the antenna, but it suffers from numerical instability when applied to model planar an- tenna structures. In this paper, in addition to our previous work [22], a nu- merical technique based on the surface equivalence theorem (Huygen’s principle) is developed to characterize onbody antennas. The method begins with a division of the original problem into two subproblems as shown in Fig. 1. The antenna is analyzed in free space by using CST Mi- crowave Studio. Near fields on a closed surface surrounding the 0018-926X/$25.00 © 2009 IEEE Authorized licensed use limited to: Queen Mary University of London. Downloaded on June 17, 2009 at 11:31 from IEEE Xplore. Restrictions apply.