The Internet Encyclopedia (Volume 3)

P1: JDW

Shankar WL040/Bidgolio-Vol I WL040-Sample.cls June 19, 2003 16:49 Char Count= 0

PROPAGATION OFSIGNALS 129

are available for loss prediction over short ranges (Har

et al., 1999; Harley, 1989; Ikegami et al., 1991).

Indoor Wireless Systems

The loss calculation in indoor wireless systems is less

straightforward than outdoor systems. It is possible to

use Equations 5 and 6 to calculate the received power or

predict the loss. Based on empirical measurements con-

ducted indoors and outdoors, the range of values ofnhas

been proposed by several researchers. Note that these val-

ues depend on the environments in which the wireless sig-

nal is propagating, and in indoor propagation, the values

ofnare strongly influenced by factors such as the building

materials used, floor arrangement, location of the trans-

mitting antenna (inside the building or outside the build-

ing), height of the transmitting antenna, frequency used.

The values ofnrange from 2 to 4 as one moves from an

open space where free space LOS propagation is possible

to urban areas with tall buildings and other structures.

In indoor environments, the value ofnless than 2 has

been observed in grocery stores and open-plan factories

(Dersch & Zollinger, 1994; Durgin et al., 1998; Rappaport

& Sandhu, 1994). This low value (lower thann=2 in free

space) has been attributed to the strong reflections con-

tributed to the received signal by the metallic structures

in those places, resulting in higher power levels compared

with a completely open space (n=2). Inside the buildings,

ncan take values in the range of 1.5– 4 depending on the

number of floors, location of the transmitter, and type of

partition (hard vs. soft) used.

A typical indoor propagation scenario is shown in

Figure 6. The base station is outside the building. It is

also possible to have the base station inside the building

depending on the complex. It is easy to see that the signal

may have to travel through multiple floors and multiple

Base station

User

Figure 6: An indoor wireless system with the base station

(BS) serving all the users inside. The BS may also be inside the

building. The signals between the BS and the users may have

to travel through several floors and partitions.

walls, each of which may be of a different type, to reach

the receiver. The loss calculations are therefore compli-

cated, and the wireless systems would have to be designed

specifically for a given complex or building. It is nonethe-

less possible to write a general equation for the calculation

of the loss. The loss at a distanced,Lpcan be expressed as

Lp(d)dB=Lp(d 0 )dB+ 10 nefflog 10

(

d

d 0

)

+

∑K

k= 1

Pk, (14)

wherenef fis the effective loss exponent taking into ac-

count the multiple floors and walls in the building through

which the signal traverses (Durgin et al., 1998; Rappaport

and Sandhu, 1994). The other factor in Equation 14,Pk,

is the specific material attenuation indBsuffered by the

signal as it traverses throughKfloors/walls. The loss at

a distance ofd 0 =1 m is given byLp(d 0 ).

Signal Variability and Fading

I now briefly review the origins of signal variability seen

in wireless systems. The signal variability may be caused

by short-or long-term fading, or both. Short-term fading

may result from multipath fading and Doppler fading. The

next sections explores these fading mechanisms and diver-

sity techniques used to mitigate the problems caused by

fading.

Multipath Fading

The second characteristic of the wireless signal is the

signal variability seen in wireless systems (Feher, 1995;

Hashemi, 1993; Kennedy, 1969; Pahlavan & Levesque,

1995; Rappaport, 2002). The signal variability is the lack

of predictability of the loss or received power. We had seen

in Figure 3 that the power loss fluctuates as the distance

increases, with the mean or median fluctuations obeying

thenth power of the distance. This random nature of the

wireless signals is termedfading. This can lead to occa-

sional loss of the signal and network break down because

the systems require a minimum amount of power (thresh-

old) to perform satisfactorily and power fluctuations may

bring the power below this threshold. This can be taken

into consideration by providing a power margin.

Even though the obvious effect of the fading is the

random fluctuations of the received power, fading also is

responsible for limiting the bandwidth capability of the

wireless systems. I first attempt to explain the reasons for

the fluctuation in power and effects of fluctuations on data

transmission and then explain why the fading limits the

bandwidth.

In a typical wireless environment, the signal reaches

the receiver after being reflected, scattered, diffracted, or

refracted from a number of objects in its path (Jakes, 1974;

Shankar, 2001). Thus, the signal does not take a single

path to reach the receiver. Instead, the signal reaches the

receiver through multiple paths, as shown in Figure 7.

These signals with various amplitudes and phases

combine at the receiver. This multipath phenomenon is

responsible for the fluctuations in the signal power ob-

served in Figure 3c. This fluctuation in power or fading

is short-term fading, of which there are two major con-

sequences. First, the random nature of these fluctuations