IEEE TRANSACTIONS ON MAGNETICS, VOL. 42, NO. 4, APRIL 2006 839
Numerical Modeling of an Indoor Wireless
Environment for the Performance
Evaluation of WLAN Systems
Theodoros T. Zygiridis , Elissavet P. Kosmidou , Konstantinos P. Prokopidis , Nikolaos V. Kantartzis ,
Christos S. Antonopoulos , Konstantinos I. Petras , and Theodoros D. Tsiboukis
Department of Electrical and Computer Engineering, Aristotle University of Thessaloniki, Thessaloniki GR-54124, Greece
HELETEL, Hellenic Electronic Communications S.A., GR-15562 Holargos, Greece
A site-specific numerical model, based on the finite-difference time-domain method, is developed in this paper for the indoor radio
channel. The scenario of interest is concerned with wave propagation in a typical office environment, for which several simulations are
performed considering different placements of the transmitting antenna. Both the 2- and 5-GHz bands are examined, where contempo-
rary wireless local area networks operate. Important channel characteristics are evaluated via the estimation of received power levels,
as well as the examination of small-scale fading and time dispersion.
Index Terms—Finite-difference time-domain (FDTD) method, indoor propagation, wireless communications.
I. INTRODUCTION
T
HE development and widespread utilization of wireless
local area networks (WLANs) has triggered intense re-
search concerning the study and prediction of indoor wave prop-
agation. This is due to the fact that the understanding of the
particular radio channel characteristics is considered essential
for designing and realizing efficient wireless communication
systems. However, in-building propagation is a highly complex
process, as it occurs within environments possessing a variety
of geometric and electromagnetic properties. Due to multiple in-
teractions with walls, furniture, equipment, and people, received
signals exhibit fading characteristics in space and spreading in
time [1]. For site-specific calculations, deterministic approaches
based on physical models that take into account all or most of the
environment’s details constitute a reliable choice. In this frame-
work, ray-tracing algorithms have been used extensively to per-
form indoor wireless studies [2]–[4]. On the other hand, there
also exist some applications of the finite-difference time-domain
(FDTD) method [5] for similar purposes. For example, the im-
pact of different wall models on several channel parameters was
pointed out via FDTD simulations in [6]. Moreover, FDTD re-
sults related to window stwere incorporated in a ray-tracing al-
gorithm in [7], while [8] introduced a hybrid ray-tracing FDTD
technique for accurate site-specific calculations.
This paper presents a set of numerical simulations regarding
WLAN operation in a typical three-room office, based on
a two-dimensional (2-D) FDTD approach. To incorporate
contemporary systems, both the 2.44- and 5-GHz bands are
considered, with the latter examined separately at 5.25 and
5.8 GHz. Via the second-order accurate solution of Maxwell’s
equations, field distributions are predicted with high spatial
resolution. Based on the received signal strengths, power-cov-
erage maps, as well as descriptions of local-field variations due
Digital Object Identifier 10.1109/TMAG.2006.871649
Fig. 1. Layout of the examined office area. Nine potential placements of the
transmitting antenna are indicated.
to multipath propagation, are provided. Statistical characteriza-
tion of small-scale fading is then performed by computing the
corresponding Rician factors. Further wide-band simulations
permit the study of the channel’s time dispersion, by analyzing
the recorded power delay profiles (PDPs). In essence, the nu-
merical results are exploited to assess fundamental properties
of the examined wireless channel. Therefore, the proposed
prediction tool can be practically utilized at the design stage
of WLANs to ensure reliable services, without resorting to
time-consuming or expensive measurement campaigns.
II. SITE DESCRIPTION AND THE FDTD MODEL
The indoor environment where WLAN operation will be
investigated is a typical office area comprising three rooms,
as depicted in Fig. 1. The examined region covers a space of
9.11 8.45 m. The walls are 20 cm thick, with conductivity
0018-9464/$20.00 © 2006 IEEE