A Telematics System using In-Vehicle UWB Communications

197

(a) (b)
Fig. 1. A highway communication infrastructure (a) and an in-vehicle femtocell (b)
High capacity, Quality of Service (QoS) and data transfer communications are in demand.
Methods to meet these demands include techniques that range from the use of smaller
radius cells and the use of sectorisation, to the most recent use of multiple antennas in
spatial and/or frequency diversity configurations. For these systems it is desirable to have
easy integration and small package antenna units. Multiple Input Multiple Output (MIMO)
Antennas composed of three 120

sectors exploiting frequency diversity gain have been
reported in (Garcia et al., 2008). In the same way, wireless systems devices for the vehicle to
road intercommunication are usually of small profile. A tri-sector configuration antenna to
meet this demand has been reported in (Garcia et al., 2010).
Low cost RoF distribution deployed in-vehicles is of great interest, Fig. 1b. Unlike coaxial,
fibre lines are capable of providing electromagnetic interference free transmissions over the
distance. Centralisation of transceivers to a system backbone might complement with lower
engineering costs, both in deployment and maintenance.
Antennas play a key role in wireless transceivers as they are the last step to radiating energy
into the space. Miniature antennas are required for the in-vehicle application; the prototypes
will have to ensure a good system performance and be platform tolerant.
In the next section, the motorway-to-vehicle wireless communication system, the in-
vehicle wireless communication system including a UWB in-car wireless channel
analysis, and finally the design of an UWB antenna for in-vehicle applications are
introduced.
2.1 Motorway-to-vehicle communication system
In Telematics applications, an autonomous data collection and processing systems can

The signal generated is carried over the downlink SMF to a remote RAU site which converts
this optical signal to an RF signal by positive-intrinsic-negative photodiode with a
transimpedance amplifier (PIN-TIA) receiver with high sensitivity. For uplink transmission,
the wireless signal received from an RAU is amplified, converted into an optical form by a
DFB laser (chosen as an ideal device for SMF and long haul communications), and
transported uplink by the SMF to the CO where another PIN-TIA receiver in BIDI TRX
demodulates the optical signal into the RF.
2.2 In-vehicle communication system
UWB technology has gained huge interest globally due to its potential to deliver high data
rate and spatial capacity, with multipath immunity (Das et al., 2006; Way, 1989) and low
power, low cost design. The deployment of this wireless technology in vehicles will provide

A Telematics System using In-Vehicle UWB Communications

199
mobility and connectivity to a host of passenger devices while reducing significantly the
costs associated with wiring.
In addition, large vehicles can benefit from the use of low cost optical fibre communications.
A RoF system can be used as part of a distributed antenna system (DAS), (with centeralised
control) which supports the deployment of femtocellular access networks at 480Mbps
within airplanes, buses, coaches, cars, lorries, trains, trams and other transport vehicles.
Such a system can assist in the minimisation of radio frequency (RF) inference when
compared to coaxial cable links, simplifies the infrastructure and reduces engineering cost.
A high-level block diagram for the in-car system is depicted in Fig. 3.
Next, a study of a UWB system over RoF in an in-vehicle scenario is described in Section
2.2.1. Experimental results of the radio propagation within the car in a realistic environment
validate the system and are described in Section 2.2.2. Fig. 3. The in-vehicle distributed antenna system


Fig. 4. The complete UWB RoF femtocell
The measured output power was -41.3dBm; this agreed with the maximum UWB Effective
Isotropic Radiated Power (EIRP) allowance and the receiver sensitivity was -79.55 dBm. The
system Transmit (Tx) power budget was measured near 0dBm after compensating for a
power penalty (attributed to optical and RF devices loses) of 8.26dB. Fig. 5. Transmitted/received UWB MB-OFDM signal
A Bit Error Rate (BER) test evaluated the performance of the system. The transmitter
produced OFDM UWB symbols that a receiver was capable of analysing; with/without

A Telematics System using In-Vehicle UWB Communications

201
multipath over an in-door channel. The resulting test for the down link is depicted in Fig. 6,
where an irrelevant BER difference between a referenced transmitter and the UWB RoF
full-duplex system is observed. Fig. 6. The system Bit Error Rate (BER)
The results show the feasibility of the RoF system; this would allow extending UWB radio
signals over hundreds of meters distances, well enough for in-vehicle applications.
2.2.2 UWB in-car wireless propagation
Based on the MultiBand technique, the multiband UWB (MB-UWB) splits the spectrum into
sub-bands and uses conventional narrow band techniques, such as Orthogonal Frequency-
Division Multiplexing (OFDM), to transmit the information in each sub-band (Elmirghani et
al., 2006).
In this section, the propagation of a MB-UWB wireless system is studied within the in-car
environment.

are
measured at a speed of 120km/h.
The vehicle in motion (system closed) affects the BER results. The ISI mainly arising from
high reflections within the small car metallic chamber is conjectured to be aggravated by the
antenna instability due to the mobile vehicle vibration and interaction through the restricted
window area. This is translated into a collection of received variations in the amplitudes and
phases of differently delayed waves caused by further fading and multipath. The
interference of direct path and the reflected waves results in higher BERs. (a) (b)
Fig. 7. BER as a function of distance (a) and BER as a function of mobility (b). (a) (b)
Fig. 8. Capacity vs. distance (a) and Capacity vs. mobility (b).
An average of 115Mbps throughput is measured when stationary and up to 102Mbps at a
speed of 120km/h. The vehicle in motion affects the BER results due to the antenna
instability that is created while in motion. There was high multipath in-vehicle, and this, in

A Telematics System using In-Vehicle UWB Communications

203
the moving vehicle, resulted in ISI and caused higher BER measured at higher speeds. In the
same way, a lower data rate was achieved in motion.
3. UWB Antenna for in-vehicle applications
Planar Inverted-F Antennas (PIFAs) are well suited for integration inside vehicles. Their
chasses may contain large steel plates and antennas over ground planes are favoured for
ceiling mounts (Garcia et al., 2009). A UWB PIFA incorporating two shorting posts with
coupling gaps is presented. The antenna operates at the lower UWB band (3.168-4.752 GHz)

attached to a grounding strip, D, electrically connected to B. The total volume of the antenna
is 19.58 x 15.75 x 5.53 mm
3
. The maximum dimension is smaller than 0.21λ at the lowest
frequency of operation.
A simulated parametric study of the capacitive gap is reported in (Garcia et al., 2010b)
where decreasing the gap between a and a' (or b and b') tended to improve the band-notch
depth and impedance roll-off. Therefore, adjusting the gap capacitance of the electrically
unconnected shorting posts allows a BPF like characteristic to be defined. An optimum
length value of a = b = 2.9 mm and, a' = b' = 1.45 mm was found to give a band-notch at
5.5GHz, a return loss (RL) of -5dB, roll-off of 0.18 and 0.03 dB/MHz and a -5dB S11
fractional bandwidth of 40%. The optimal value of the gap corresponds to 1.18mm. Fig. 10. The UWB PIFA antenna
The reflection coefficient of the UWB antenna is shown in Fig. 10 compared to a standard
frontend BPF (RFlambda, n.d.). Compared to the commercially available BPF (2441MHz
pass-band rejection and 2dB insertion loss; roll-offs of 0.050dB/MHz and 0.031dB/MHz for
the lower and upper bands respectively), the UWB antenna has a lower 1108MHz pass-band
rejection and improves the roll-offs to 0.024dB/MHz and 0.030dB/MHz.
An antenna having a VSWR of 3.57 (5dB RL) can be calculated to present an equivalent
mismatch loss of 1.65dB (Kraus & Marhefka, 2001). Therefore, if the BPF and its associated
mismatch loss of 2dB was removed, then there will still be an overall reduction in loss of
0.8dB.
The measured 5dB return loss bandwidth of the proposed PIFA is 42.15% for the
3.168-4.860 GHz FCC UWB. To investigate the effect of attaching the antenna to a large
conducting plate in a car chassis, a larger ground plane of dimensions 510 x 800 x 0.75 mm
3

was placed ¼ wavelength below the PIFA. Fig. 10 shows the S11 response.

Telematics for the delivery of enhanced services to highways users.
A Telematics system has been described based on a high performance WiMAX spectrum
using an unlicensed band (5.470-5.725 GHz) and recent developments in RoF systems as a
base for the delivery of wireless communications to motorway-to-vehicle applications. This
approach results in a relative low cost deployment and maintenance, extends the radio over
long distances and delivers peak rates of at least 1 Gbit/s to fixed users and 100Mbit/s to
mobile users over micro-cells.
Cost-effective while efficient narrow band tri-sector antenna units have been assessed for
Intelligent Transport Systems (ITS) in the presented highway scenario. The antennas served
the WiMAX standard over full-duplex bi-directional optical links (Garcia et al., 2008). These
antennas seem to be reasonably proficient for use in ITS due to their potential higher gains
and reduced spatial limitations. The low VSWR performance achieved by the use of these
narrow band antennas can improve the system link budget, which is translated into a
relatively higher coverage/throughput.
The Robustness to multipath interference offered by the unlicensed lower band
(3.168-4.752 GHz) UWB communication is to be exploited for in-vehicle communications.
Within this work a promising low cost RoF link to extend the UWB radio over relatively long
distances (i.e.: trains, trams and airplanes) has been introduced. The transmission was assessed
using a relatively inexpensive multimode RoF link. The transmission network was capable of
providing high data rates of 400-480Mbps at picocells of about a metre radius with
inconsiderable SNR degradation performance over fibre links of several hundred of meters.
In addition, a wireless propagation of UWB radio inside a vehicle is analysed. The analysis
of the UWB radio channel in-vehicles demonstrates that UWB is a very suitable and
promising technology for transmission networks able to provide high data rates of 400Mbps
within cars. Path loss was not of a significant level due to the short ranges that are
encountered within cars. However, the main attenuation might perhaps be due to
shadowing effects. High data rates were achieved in closed environment scenarios (Garcia et
al., 2009). As many new cars include air conditioning, it is not unreasonable to expect the
environment to be closed for the majority of the time.
A UWB antenna design example for an in-vehicle application has been introduced. The

Loughborough, UK, 17-18 March, 2008.
García Zuazola, I.J., Elmirghani, J.M.H. & and Batchelor, J.C. (2009). High-speed ultra-wide
band in-car wireless channel measurements, IET Communications., pp. 1115–1123,
Volume 3, Issue 7, 2009.
García Zuazola, I.J., Batchelor, J.C. & Elmirghani, J.M.H. (2010). Sectorized WIMAX Antenna
for future Vehicular Communications Systems, Microwaves, Antennas & Propagation,
IET, pp. 210 – 218, Volume 4, Issue 2, Feb. 2010.
García Zuazola, I.J., Batchelor, J.C., Elmirghani, J.M.H. & Gomes, N.J. (2010b). UWB PIFA
Antenna for simplified transceivers, Electronics Letters, pp. 116–118, Volume 46,
Issue 2, January 2010.
Gunter, Y. & Grosmann, H.P. (2005). Usage of Wireless LAN for Inter-Vehicle
Communication, Proceedings of the 8th International IEEE Conference on Intelligent
Transportation Systems, pp. 296-301, Vienna, Austria, September 13-16,
2005.
Kerner, B. S., Rehborn, H., Aleksi, M. & Haug, A. (2005). Traffic Prediction Systems in
Vehicles, Proceedings of the 8th International IEEE Conference on Intelligent
Transportation Systems, pp. 251-256, Vienna, Austria, September 13-16,
2005.
Kraus, J. D. & Marhefka R. J. (2001). Antennas for all applications, 3rd edition, McGraw-Hill,
ISBN 0072321032, Boston, 2001.
Lee, K.F. & Williams, D. B. (2000). A Space-Frequency Transmitter Diversity Technique for
OFDM Systems, IEEE Globecom, pp. 1473-1477, Volume 3, San Francisco, Nov.
2000.
Mohammad, N.H. & Ismail, W. (2008) System-level integration and simulation of ultra
wideband receiver front-end, Communications, Propagation and Electronics, MIC-CPE
Mosharaka International Conference, pp. 1-6, Jordan, 6-8 March 2008
RFlambda (n.d.) Available from: www.rflambda.com

Novel Applications of the UWB Technologies


1. Introduction
In this chapter we present UWB communication as a potential candidate for cognitive radio
technology. Cognitive radios are intelligent radios that could adopt itself by sensing and
learning the radio environment and optimize its transmission strategies to maximize the
utilization of the scarce radio resources such as the radio spectrum. This has been motivated
by the radio regulatory bodies around the world (EC, 2007; FCC, 2003) to utilize unused
radio spectrum known as white space in the spatio-temporal domain. In the recent years
UWB communication has emerged as a potential candidate for the CR technology due to
its ability to share the spectrum with others for short range wireless communications. In
this context we present the concept of cognitive radios and the necessary techniques to
adopt UWB as cognitive radios in this chapter. Especially, we enhance on the fundamentals
of cognitive radios and spectrum sensing which enable the UWB radio to learn the radio
environment. We also touch upon other cognitive radio related topics that are related to UWB
communications such as dynamic spectrum access, interference mitigation and localization
techniques. Furthermore, we present some potential applications for the use of UWB based
cognitive radios which are derived from the European Union funded projects EUWB (EUWB,
2008) which is one of the biggest UWB projects that the world has seen so far, and the
C2POWER project (C2POWER, 2010) which is related to energy efficiency in short range
wireless communications with the use of cognitive radios. In this chapter we do not consider
the technological aspects related to the use of cognitive radios for energy efficiency but only
consider the use of cognitive radios for dynamic spectrum access. However, at the end of
the chapter we present a scenario for the use of cognitive radios for energy efficiency derived
from (C2POWER, 2010).
In the material presented in this chapter we mainly consider the high data rate UWB
radios based on the Multi-Band Orthogonal Frequency Division Multiplexing (MB-OFDM)
technique following the Wimedia specifications (Wimedia-PHY, 2009). The OFDM based
transceiver design makes it feasible for the UWB radio to sense the radio environment
and dynamically change the transmission parameters accordingly. This makes the UWB
11
2 Will-be-set-by-IN-TECH

at a particular location over time. Therefore it is necessary to sense and learn the radio
environment in order to maximize the spectral usage opportunistically. In other words
212
Novel Applications of the UWB Technologies
UWB Cognitive Radios 3
detecting ’spectrum holes’ as it is termed is quite crucial for dynamic spectrum access. Figure-
1 depicts the concept of ’spectrum hole’ evolution in the spatio-temporal domain.
2.2 Spectrum sharing in cognitive radio networks: ’Underlay’ and ’Overlay’ techniques
With cognitive radio technology the concept of ’primary users’ and ’secondary users’ of the
spectrum are developed. The primary users are the incumbent users with the exclusive rights
to use the spectrum at anytime and the secondary users, also known as the cognitive radio
users, are the users that use the spectrum without interfering with the primary users. There
are basically two spectrum sharing techniques considered for cognitive radio networks for
maximizing the spectral efficiency between the primary and the secondary users. First is the
’spectrum underlay’ technique and second is the ’spectrum overlay’ technique.
Fig. 2. Spectrum sharing in cognitive radio networks, with (a) overlay and (b) underlay
sharing techniques.
In the ’spectrum underlay’ method the secondary users can utilize the spectrum
simultaneously with the primary users without exceeding a predefined interference level to
the primary users. Secondary users in this case can share the spectrum such that the total
interference power from the secondary users to the primary users are controlled below the
interference limit set by the relevant regulatory authorities. The characterization of such
interference limit is given in the next subsection. Figure-2 depicts the concept of spectrum
underlay technology. UWB radio technology due to the low powered transmissions in the
ultra wide band frequency range is therefore a potential candidate for deploying spectrum
underlay technology for spectral sharing. Using the low powered transmissions and making
sure that the interference limit is not exceeded UWB radios can potentially share the spectrum
with the primary users and coexist.
In the ’spectrum overlay’ method the cognitive radios can identify the spectrum holes in the
spatio-temporal domain and opportunistically utilize them by giving higher priority to the

where k
= 1.38 ×10
−23
is the Boltzmann’s constant, B (Hz) is the receiver operating frequency
bandwidth and T
N
(in degrees Kelvin unit) is the noise temperature. Likewise the total
interference power P
I
due to the transmissions of wireless devices and natural interferences
at a particular point in space can be characterized by,
P
I
= kT
I
B (2)
The interference temperature limit T
max
I
therefore is an upper limit on the value of T
I
that
can be used to control and limit the interference in the radio environment. Such limits for the
interference temperature can be used to enable the underlay spectrum sharing technique by
coordinating or policing the interference level in the environment generated by the secondary
users to the primary users.
2.4 The cognitive cycle
The cognitive cycle is the term describing the activities involving the intelligence of the radio
device such as sensing, learning and adopting. In (Mitola, J. & Maguire Jr. G.), Mitola had
presented a generic cognitive cycle that corresponds to his view of ideal cognitive radios.

radio in the environment is not always feasible given the fact that the localization task needs
to be performed blindly. Once the cognitive radio nodes have a good understanding of
the radio environment it would then perform power control with appropriate interference
mitigation techniques in the spatio-temporal domain to transmit its data. Furthermore, other
functionalities also can be added into the cognitive engine depending on the applications and
any specific requirements appropriately.
215
UWB Cognitive Radios
6 Will-be-set-by-IN-TECH
3. Dynamic spectrum access
The radio spectrum can be utilized by considering various access strategies, methodologies
or policies. In this section we provide a quick background on the spectrum access models to
explain how cognitive radios are used for dynamic spectrum access. Spectrum access models
can be classified as command and control model, exclusive-use model, commons model and
the shared model (Hossain, E. et. al.). In the shared use model the secondary user of the
spectrum will opportunistically access the spectrum without interfering with the primary user
of the spectrum, in the exclusive use model the unlicensed secondary user can be granted
access to the spectrum by the licensed primary user, and in the commons model the secondary
user can access the spectrum without any restrictions. In Figure-5 we present a taxonomy of
the different spectrum access models.
Fig. 5. Classification of spectrum access models
For a detailed description of the different access models the reader is referred to (Hossain,
E. et. al.). In the previous section we briefly described the access model that is of interest to us
which is the shared spectrum access model that includes the spectrum underlay and overlay
techniques. In the ’shared’ model the concept of primary and secondary users of the spectrum
are derived and the spectrum can be shared simultaneously between the primary and the
secondary users of spectrum. The primary users are the incumbent users of the spectrum
however the secondary radios also can use the spectrum. In this case the secondary radios
need to make sure that they do not interfere with the primary radio transmissions, and as long
as the interference constraint is met the secondary users can use the spectrum transparently

interfered. Spectrum sculpting techniques are used for shaping the spectrum in UWB radios
(Wang, Z.; Yamaguchi, H.). The two most common spectrum sculpting methods are the
spectrum shaping in time domain using shaped pulses and spectrum shaping in the frequency
domain using tone nulling (in OFDM systems). The time domain method in general may
not be possible to shape the spectrum in all the cases, the frequency domain tone nulling
method on the hand can provide better performances in terms of shaping the spectrum. The
tone nulling technique can cause spectral overshoots in the transmission band and hence
various derivatives of this method are also considered such as enhanced active interference
217
UWB Cognitive Radios
8 Will-be-set-by-IN-TECH
cancelation as proposed in (Wang, Z.). The example shown in Figure-6 clearly depicts how the
spectrum sculpting technique is used in UWB radios in order to coexist and share the radio
spectrum with the primary user radios in the environment.
4.2 Power control
Power control in wireless and mobile communications is a well studied topic for more than
twenty years. It has attained more attention in the recent years for potential spectrum sharing
in cognitive radio networks. Traditionally power control was considered for maximizing
the transmission rate with fare-scheduling without degrading the QoS of the other users in
the environment. In a similar context power control is also considered for cognitive radio
networks as presented in (Gu, H.; Radunovic, B.; Xing, Y.; Zhang, L.). Here we briefly explain
the concept on power control for dynamic spectrum sharing with underlay technology in
cognitive radio networks by having the total interference power as a constraint.
Suppose P
I
is the interference power limit corresponding to the interference temperature T
I
as explained in (2). If there exist K number of cognitive radios in the environment sharing the
spectrum with the incumbent users, then the total interference caused to the l
th

K
∑
k=1
h
kl
P
k
≤ P
I
∀l (4)
The cognitive radio nodes on the other hand would like to achieve the highest possible
transmission rate which is related to the received signal to interference ratio γ
km
at the m
th
secondary receiver where m = 1 K and m = k,givenby,
γ
km
=
h
km
P
k
∑
K
u
=1,u=k,m
h
um
P

l
≤ P
I
(6)
It might be difficult to measure the interference power at the primary user node unless the
primary user cooperates. In such situations there can be a power controller or a monitor
serving the purpose of controlling the power by measuring the total interference power at
some central location. In literature one could find various cooperative and distributed power
controlling methods using game theoretic approaches which we do not cover in this chapter.
218
Novel Applications of the UWB Technologies
UWB Cognitive Radios 9
5. Spectrum sensing
Spectrum sensing is one of the crucial functionalities of a cognitive radio in order to learn
the radio environment. Various spectrum sensing techniques exist (Kandeepan, S. et. al;
Yucek, T. and Arslan,H.) and in general could be classified as 1) energy based sensing, 2)
cyclostationary feature based sensing and 3) matched filter based sensing. The energy based
sensing is the simplest method to sense the environment in a blind manner, the cyclostationary
based sensing may require some information about the spectral-user signal characteristics,
and the matched filter based sensing requires the complete information of the spectral-user
signal. In this section we elaborate in detail on the various spectrum sensing techniques
and their related detection performance for MB-OFDM based sensing. Moreover, we present
collaborative sensing techniques in order to address the ’hidden node problem’.
Let us provide some background on spectrum sensing prior to presenting the related
techniques. Spectrum sensing and detecting the presence of a radio in the environment
is treated as a classical statistical detection problem (Kay, S.). We define the two binary
hypotheses H
0
and H
1

1
] (9)
P
FA
= Pr[d = 1|H
0
] (10)
The probability of miss detection on the other hand is defined by Pr
[d = 0|H
1
],andthusis
given by P
M
= 1 − P
D
. In general the detection threshold λ is chosen in order to trade off
between the detection and false alarm probabilities. Different criteria can be used in order to
find the optimal threshold which is a well treated topic in the literature of statistical detection,
which we do not present in this chapter. In the subsequent sections we provide various ways
to derive the test statistic ξ used for the detection of primary users.
5.1 The hidden terminal problem
Prior to presenting the spectrum sensing techniques we present why spectrum sensing is
treated as an important topic in cognitive radio literature. We mentioned that the detection
performance is characterized by the probability of successfully detecting the radio and the
probability of false alarm. In cognitive network applications the regulatory bodies are quite
strict on secondary nodes causing any interference to the primary users, in this sense the
primary users need to be reliably detected by the secondary users with a high detection
probability (close to 100% or P
D
 1). The detection probability usually depends on the

statistic, where T
= NT
s
and T
s
is the signal sampling period. The test statistic at the base
band considering the complex envelope of the received signal is therefore given by,
ξ
=

t
2
t
1
r(t)
˜
r
(t)dt (11)
where,
˜
r
(t) is the complex conjugate of r(t). The signal to noise ratio (SNR) is then defined
based on the received signal s
(t) for t
1
< t ≤ t
2
for some t
1
, t

ξ
≈ T
s
N
−1
∑
n=0
r[n]
˜
r
[n] (13)
where, N is the total number of complex samples and is also known as the time-bandwidth
product (Urkowitz, H.). Note that in (12) there are essentially N number of real component
samples and N number of imaginary component samples. Considering the discrete domain
test statistic the detection criteria is then given by,
d
=

0; ξ
< λ
1; ξ
≥ λ
(14)
In order to compute the detection probability and the false alarm probability we consider the
distribution of the test statistic ξ. The energy based test statistic ξ follows a non-central and a
central chi-sqaure distribution under H
0
and H
1
respectively with 2N degrees of freedom.

(.) is the Gamma function, Q
N
(a, b)=

∞
b
u
N
exp(−(u
2
+ a
2
)/2)I
N−1
(au)/a
N−1
du
is the generalized Marcum Q-function, and I
N−1
(.) is the modified Bessel function of first kind
with order N
−1.
Let us look at some results for the detection performance of the energy detector in the additive
Gaussian noise channel by plotting the complementary receiver operating characteristics (C-
ROC) curve. The C-ROC depicts the probability of false alarm in the x-axis and probability
of miss detection in the y-axis. Figure-8 shows the C-ROC curves for the energy detector
for various values of signal to noise ratio levels ρ.Asweobservefromthefigure,the
detection performance improves with increasing values of ρ by achieving lower miss detection
probabilities for lower false alarm probabilities when ρ increases. Figure-9 on the other hand
shows the C-ROC curves for various values of N, and again we observe that the detection