33

CHAPTER 3

WIRELESS SENSORS FOR AIRCRAFT INDOOR

COMMUNICATION

In this chapter, aircraft wireless communication model that replaces

existing indoor wired sensor communication in aircrafts by wireless sensor
architectures to reduce electrical wiring issues is discussed. Aircraft wireless
communication model has been developed for aircraft indoor wireless
networks. In the proposed wireless system, radio signals from multiple
wireless sensors using the space division multiple access technique increases
the strength of transmission power and reduces electromagnetic signal
interference. The performance of proposed aircraft wireless communication
model is investigated and validated for safe indoor propagation and enhance
the overall capacity of the aircraft wireless communication system.

3.1 AIRCRAFT COMMUNICATION ARCHITECTURE

Existing wired aircraft communication architecture is shown in

Figure 3.1(a), Replacing the aircraft sensor wiring by wireless
communications reduces the weight of wiring in the aircraft and also
improves the safety of the aircraft. Moreover, minimizing aircraft wiring
structure as shown in Figure 3.1(b) reduces the sensor installation and
maintenance costs. Aircraft wireless communication model considers multiple
reflections and transmissions of the propagated waves and predicts the
propagated signal with accuracy. Time, frequency and multipath channel
responses are analysed at different positions in indoor aircraft environment. A
 34

centralised data processing centre consisting of many access points (antenna)

transmit signals to different node points (sensors) as well as receives signals
from node points. This arrangement is used for simulation.

Central
Central computer
computer

Ethernet based
CDPC/HUB
Backbone bus

Sensors Wireless
Sensors

(a) (b)
Figure 3.1 (a) Wired sensor architecture (b) Wireless sensor architecture

3.1.1 AWC System Model

The characteristics of the propagation channel in any radio system

are influenced by the design of the system. The design also determines signal
distortions, power loss, temporal, spatial and harmonic dispersions of radio
signals that are propagated from transmitter to receiver. The proposed aircraft
wireless communication system model is shown in Figure 3.2. It is
characterized by strong multipath propagation for communication link. The
fixed wireless access systems provide data rates up to 30 Mbps with coverage
radius of 2 km and its channel provides adequate spatial diversity. This
system can be used for aircraft wireless local area network.

The indoor propagation channel involves multipath propagation,

which leads to distortion of the received signals. The magnitude of this
distortion increases with an increasing data rate. Restrictions on bandwidth
 35

and transmit power limit the capacity of the transmission (Nowak & Mitra
2003). The data rate is obtained based on transmission techniques and the
spectral efficiency. The spectral efficiency is increased by better coding and
higher transmitter power.

Central Sensor data Transmitter

computer processing centre

Channel

Channel

Control unit data

Sensor
processing centre Receiver
(CDPC/Hub)
Control
Unit Noise

Figure 3.2 Aircraft wireless communication system

The space division multiple access technique has been proposed for
aircraft wireless communication for two reasons. First, for reuse of bandwidth
within short range to attain high spectral efficiency. Second, to separate
spectrally overlapping sensor nodes based on their positions and by the
processing method of the signals received at an antenna. The space division
multiple access technique also offers efficient solutions to channel estimation,
synchronization and power control in any practical wireless communication
system.

3.2 MULTIPATH PROPAGATION IN AIRCRAFTS

Metallic materials consisting of magnesium, aluminium, steel,

titanium, alloys, and non-metallic materials consisting of reinforced plastic,
transparent plastic, composite, and carbon-fibre are used in an aircraft
 36

construction. These materials cause signal loss during propagation. The

Boeing 787's aircraft is made of 50% composite, 20% aluminium, 15%
titanium, 10% steel, and 5% other materials. Aluminium and composite
materials constitute more than 20% of the Airbus A380's airframe (Airbus-
A380 2012). Metallic and non-metallic materials have different influences on
the electromagnetic waves.

A mathematical model for aircraft wireless communication system

has been developed by considering the signals loss during propagation in
aircrafts. The spatial characteristics of multipath channel are modelled,
verified, and analysed by Rayleigh distribution. It considers the multipath
propagation reflections of electromagnetic waves on the walls of aircraft and
other objects in the indoor aircraft environment.

The indoor propagation channel considers reflection and diffraction

of radio waves on various obstacles during transmission. The propagation
channel produces attenuated and delayed pulses for each transmitted pulse.
The transmitted signal reaches to the receiver by multiple paths or through
multipath propagation. The component such as delay, attenuation and phase
shift of the received signals are determined by obstacles encountered and the
length of travel of the electromagnetic waves. Multipath channel impulse
response at time instant T and position P are determined by Equation (3.1).

(3.1)

where represent the number of multipath components.

refers to the amplitude of the signals, refers to the phase shift of the
signals, is a delta function and refers to the time delay of the
multipath component. The number of physical paths is theoretically infinite.
A practical model considers the amplitude of the signals in physical paths as
well as thermal noise power in the channel. When indoor propagation velocity
 37

of the objects is low, the multipath parameters change slowly with respect to
the data rates.

3.2.1 Multipath Channel and Time Response

The power delay ( ) refers to the time distribution of the received

signal power when a pulse is transmitted over the channel. is obtained as
the squared value of the impulse response for a multipath channel and this is
given in Equation (3.2). For a sample case, excess delay values are assumed
in multiples of and the corresponding received power are
computed based on Rayleigh distribution. This is presented in Table 3.1. The
graphical representation of the data in Table 3.1 is presented in Figure 3.3. It
is observed that the received power varies in a random manner for different
values of excess delay.

(3.2)

The time dispersion in a multipath channel is denoted by mean

excess delay and mean square delay spread . The mean excess delay
and delay spread are defined as first and second central moment of the power
delay and are given in Equations (3.3) and (3.4).

Mean excess delay and mean square delay values for the selected
case study are respectively 30 ns and 11 ns that is one-tenth of the inverse of
the delay spread or 10 Mbps. Transmitter transmits data at the data rate of 10
Mbps without equalization. The maximum excess delay is defined as the
delay of the latest multipath component of the power, which is above a
threshold. The threshold value of power is chosen based on the consideration
that the effect of attenuation is negligible below that value.
 38

Table 3.1 Excess delay vs. signal attenuation

S. S.
S.No
No (µs) (dB) (µs) (dB) No (µs) (dB)
1 0.004 -39 11 0.044 -77 21 0.084 -91
2 0.008 -58 12 0.048 -59 22 0.088 -68
3 0.012 -68 13 0.052 -37 23 0.092 -95
4 0.016 -30 14 0.056 -60 24 0.096 -32
5 0.020 -66 15 0.060 -43 25 0.100 -89
6 0.024 -33 16 0.064 -34 26 0.104 -31
7 0.028 -41 17 0.068 -68 27 0.108 -91
8 0.032 -85 18 0.072 -80 28 0.112 -86
9 0.036 -31 19 0.076 -66 29 0.116 -39
10 0.040 -62 20 0.080 -62 30 0.120 -47

-30

-40

-50

-60

-70

-80

-90

-100
0 0.02 0.05 0.07 0.10 0.12

Excess delay (µs)

Figure 3.3 Channel power delay profile

 39

(3.3)

(3.4)

It is assumed that the threshold value is 60dB for the case study.
From the Figure 3.3, it can be seen that corresponding to excess delay value
of 102 ns, the attenuation is above the threshold value.

3.2.2 Multipath Channel and Frequency Response

The frequency response of the channel, R[f], can be obtained from

the Fourier transform of the complex baseband impulse response as given in
Equation (3.5). The magnitudes of the baseband frequency response of
channel corresponding to the channel power delay profile in Figure 3.3 are
shown in Figure 3.4. It can be seen that the channel attenuation is frequency
dependent. External objects add unwanted frequencies to the multipath
components signals and their phase shifts are frequency dependent. In
Equation (3.6), the autocorrelation function and the channel frequency
e shift over a frequency
c Hz and the correlation of the channel attenuation.

(3.5)

(3.6)

In the coherence bandwidth of the channel, when the degree of

correlation is decreased the frequency separation increases. From the case
study Figure 3.4, it can be seen that bandwidth separation occurs when the
correlation is decreased by around 3 dB. It is also seen that when the
coherence bandwidth is 70 MHz, there is no need for any equalization.
 40

Aircraft wireless communication system can transmit signals over a

coherence bandwidth and the channel response is smooth compared to that of
200MHz bandwidth shown in Figure 3.4. Further, frequency selective fading
occurs when the channel response shows significant variations.

-40

-50

-60

-70

-80

-90

-100
0 50 100 150 200
Frequency (MHz)

Figure 3.4 Channel frequency responses

3.2.3 Spatial Domain and Channel Response

The power received by the antennas (angular power) is a function of

the incident angle. Figure 3.5 shows the channel angular power and the angle
spread of the channel is the angular depressiveness of the channel. Angle
spread refers to the second central moment of the angular power and the mean
angle-of-arrival is given in the Equations (3.7) and (3.8).

When the angle spread is greater than zero, the received signal
arrives from different directions. Transmitted signals can interfere with other
signals depending on the location of the receiving antenna. The angle-spread
 41

produces space-selective fading depending on the coherence distance of the

channel. Coherence distance refers to the spatial separation which the
correlation of the channel response is decreased by 3 dB approximately. The
angle spread is inversely proportional to the coherence distance due to the
delay spread and coherence bandwidth.

(3.7)

(3.8)

The fading rate of the channel is defined as the maximum rate at

which the channel response is changed. The fading rate of the channel is
expressed in Hertz. This is the ratio between the maximum velocities that any
object in the propagation environment is expected to move the wavelength of
the carrier frequency.

For example when velocity of 150m/s and a carrier frequency of

2 MHz are used for the aircraft local area network application and the channel
fading rate is minimum. The coherence time of the channel, is defined as
the minimum period of time or the inverse of the fading rate divided by two.

The sensors allow the entire frame to be transmitted. For the chosen
case study, the difference in channel response at the beginning and end of
such transmission takes place over a distance of 0.023 wavelengths. Figure
3.6 shows simulation results that confirm that the channel responses are
correlated. Graphical representation of the channel response for different
wave lengths and correlation parameters is given in Figure 3.6.
 42

900
- 50 dB
- 60 dB
-70 dB
- 80 dB

0
1800 0

2700

Figure 3.5 Channel angular power profile

0.8

0.6

0.4

0.2

0
0 0.5 1 1.5 2 2.5 3

Figure 3.6 Channel response correlation

 43

3.2.4 Channel Modelling

Channel modelling presents the channel frequency response based

on the Rayleigh distribution and investigates how the probability density
functions of the channel responses. A mathematical model is developed for
channel responses based on the propagation environment, reflection,
transmission and path loss coefficients. The channel responses are used to
measure the performance of a communication system in propagation
environment. A channel frequency response model which describes the
characteristics of the indoor propagation channel accurately is needed.
Graphical representation of channel frequency response is shown that the
average power delay varying exponentially based on the excess delay. The
impulse amplitude at excess delay is approximately Rayleigh distributed and
its phase is uniformly distributed. Channel frequency response which exhibits
these characteristics is mathematically expressed in Equation (3.9).

(3.9)

refers to the tap spacing, refers to the phase corresponding

th
to tap and L refers to uniformly spaced taps. The channel response is
modelled as a Rayleigh process at particular frequency. The channel
responses (impulse amplitudes) for different time instants for the case study
are given in Table 3.2. The Rayleigh distribution is overestimating the
probability density function in the lower amplitude range in Table 3.2.
Figure 3.7 shows that the fading of signals is more at high amplitudes.
 44

Table 3.2 Channel frequency response vs. Time

S.No t H(t) S.No t H(t) S. No t H(t)

1 0 0 12 0.99 0.734646 23 1.98 0.045281
2 0.09 0.210986 13 1.08 0.642860 24 2.07 0.030753
3 0.18 0.410009 14 1.17 0.548027 25 2.16 0.020451
4 0.27 0.586232 15 1.26 0.455600 26 2.25 0.013318
5 0.36 0.730915 16 1.35 0.369674 27 2.34 0.008495
6 0.45 0.838126 17 1.44 0.292949 28 2.43 0.005308
7 0.54 0.905101 18 1.53 0.226851 29 2.52 0.003249
8 0.63 0.932232 19 1.62 0.171734 30 2.61 0.001948
9 0.72 0.922720 20 1.71 0.127147 31 2.70 0.001145
10 0.81 0.881963 21 1.80 0.092092 32 2.789 0.000659
11 0.90 0.816789 22 1.89 0.065272 33 2.879 0.000372

0.8

0.6

0.4

0.2

0
0 0.5 1 1.5 2 2.5 3 3.5

Amplitude

Figure 3.7 Rayleigh fading vs. Amplitude

 45

3.2.5 Simulation Results and Discussion

The simulation is carried out for a carrier frequency of 2 MHz and

with an attenuation of 45 dB below the strongest multipath component. The
power delay profile corresponding to the above parameters is obtained from
Figure 3.3. The frequency response of Figure 3.4 provides the normalized
channel frequency response in MHz. The angular power profiles of
Figure 3.5, illustrates the variation of the channel response over frequency
and space. Figure 3.6 shows the spatial and expected correlation of the
channel response over a certain distance. Rayleigh fading vs. probability
density function is graphically shown in Figure 3.7. It is observed that
probability of fading is high when amplitude is maximum. This result is
obtained from average of multiple channel realizations. The simulation result
demonstrates that the indoor channel (frequency) is spatially selective, due to
the rich multipath propagation. Another observation from the study of
simulation results is that the delay spread of the channel realizations is 40 ns.
The frequency response is frequency selective with coherence band width
between 50 and 70 MHz.

3.3 CONCLUSION

The characteristic of the indoor propagation channel is investigated

for ALAN application. The Rayleigh model is used to model the channel in
analytical derivations. The SDMA techniques offer solutions for ALAN. This
technique can be used for efficient communication of wireless signals in
AWC system. Multipath propagation can be used to determine frequency,
spatial selectivity for the bandwidth and distance. There is strong multipath
propagation between the wireless transmitter and receiver. Therefore, AWC
system offer solutions for high data rate ALAN and helps to manage the
channel frequency and spatial selectivity. AWC model also provides solutions
for real time performance such as channel estimation, synchronization and
power control and it is efficiently communicate the wireless signals.