On the coexistence of ultra wideband (uwb) wireless communication …summit.sfu.ca/system/files/iritems1/9729/etd1819.pdf · 2020-06-18 · on the coexistance of ultra wideband (uwb)

ON THE COEXISTANCE OF ULTRA WIDEBAND (UWB) WIRELESS COMMUNICATION SYSTEMS WITH

NARROWBAND INTERFERENCE

Amir Hosain Jodar B .A.Sc. (Honors), Simon Fraser University, 2003

THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF

MASTER OF APPLIED SCIENCE

In the School of

Engineering Science

O Amir Hosain Jodar 2005

SIMON FRASER UNIVERSITY

Summer 2005

All rights reserved. This work may not be reproduced in whole or in part, by photocopy

or other means, without permission of the author.

APPROVAL

Name:

Degree:

Title of Thesis:

Arnir Hosain Jodar

Master of Applied Science

ON THE COEXISTANCE OF ULTRA WIDEBAND (UWB) WIRELESS COMMUNICATION SYSTEMS WITH NARROWBAND INTERFERENCE

Examining Committee:

Chair: Dr. Ljiljana Trajkovic Professor of Engineering Science

Dr. Dong In Kim, Senior Supervisor Associate Professor of Engineering Science

Date Approved:

Dr. Paul Ho, Supervisor Professor of Engineering Science

Dr. Steve Hardy Professor of Engineering Science

July 28,2005

SIMON FRASER UNIVERSITY

PARTIAL COPYRIGHT LICENCE

The author, whose copyright is declared on the title page of this work, has granted to Simon Fraser University the right to lend this thesis, project or extended essay to users of the Simon Fraser University Library, and to make partial or single copies only for such users or in response to a request from the library of any other university, or other educational institution, on its own behalf or for one of its users.

The author has further granted permission to Simon Fraser University to keep or make a digital copy for use in its circulating collection.

The author has further agreed that permission for multiple copying of this work for scholarly purposes may be granted by either the author or the Dean of Graduate Studies.

It is understood that copying or publication of this work for financial gain shall not be allowed without the author's written permission. \

Permission for public performance, or limited permission for private scholarly use, of any multimedia materials forming part of this work, may have been granted by the author. This information may be found on the separately catalogued multimedia material and in the signed Partial Copyright Licence.

The original Partial Copyright Licence attesting to these terms, and signed by this author, may be found in the original bound copy of this work, retained in the Simon Fraser University Archive.

W. A. C. Bennett Library Simon Fraser University

Burnaby, BC, Canada

ABSTRACT

Ultra wideband (UWB) communication is based on the transmission of short

pulses on the order of nanoseconds. As a result, UWB systems operate over dedicated

spectrum bands already occupied by other narrowband systems. Because a UWB device

spreads its energy over a large frequency range, it radiates a small percentage of its total

power in the operating spectrum of narrowband receivers. Consequently, UWB signals

appear as low-power noise and have little impact on underlying operating devices.

However, the interference coming from narrowband systems operating in the spectrum

allotted for UWB use could be problematic. This thesis contains two main investigations.

In the first investigation, we analyze the coexistence of time-hopping pulse amplitude

modulation UWB systems in the presence of narrowband interference exploiting additive

white Gaussian noise and flat-fading channels. The second investigation deals with the

coexistence of UWB systems with narrowband interference based on experimentation.

l o my h a r parents Rosa andM@id

Mom your l f e is a perJect refkction of your ayelic soul

I hue you!

ACKNOWLEDGEMENTS

I express my deepest gratitude to my senior supervisor, Professor Dong In Kim,

for his guidance throughout the course of my project. Thank you for your insight,

kindness, and your confidence in me.

I would also like to thank Dr. Paul Ho and Dr. Stephen Hardy for serving on my

examining committee and Dr. Ljiljana Trajkovic for chairing the thesis defense. Dear

Ljiljana, I am ever grateful for all your kindness. Thanks for changing your flight

schedule in the midst of the night just for making it to my thesis defense.

I would not have accomplished my goals had there not been all the love and

support of my family. I give you all my gratitude and love. Dad you're the Champ. Mom

I don't know what to say, thank you thank you thank you and I love you. Doostetoon

daaram va mokhlesetoonam hastam.

My especial thanks to my friends Mrs. Raj Pabla and Dr. Andrew Rawicz for all

the wonderful time we had at the Renaissance Cafe.

Finally, my sincere thanks to my colleagues in the RFMcrowave Mobile

Communications Laboratory especially: Serhat, Pejman, Thomas, Wan-Jong, Kyoung-

Joon, Tao, Dr. Cavers, Dr. Vaughan, and the rest of the group.

TABLE OF CONTENTS

. . Approval .......................................................................................................................... 11

... Abstract ......................................................................................................................... 111

Dedication ......................................................................................................................... iv

Acknowledgements ............................................................................................................ v

Table of Contents ............................................................................................................. vi ... List of Figures ................................................................................................................. vlll

List of Tables ...................................................................................................................... x

Glossary ......................................................................................................................... xi

Chapter 1 Introduction to Ultra Wideband (uwb) Signals and Systems ...................... 1 1.1 Brief History of Ultra Wideband ..................................................................... 1 1.2 Overview of UWB and the Coexistence Issue ................................................ 3 1.3 UWB Modulation and Multiple Access Formats ............................................ 6

1.3.1 UWBPulseShape ........................................................................................ 7 1.3.2 Time-Hopping UWB (TH-UWB) Transmission Scheme ......................... 10 1.3.3 Direct-Sequence UWB (DS-UWB) Transmission Scheme ....................... 13

............................................................................. 1.4 UWB Signal Propagation -13 1.5 UWB Benefits and Applications ................................................................... 14

................................................................................................. 1.6 Project Goals 15

Chapter 2 On the Performance of TH-UWB Systems in the Presence of Narrowband Interference ............................................................................. 17

................................................................................................... Introduction 17 ................................................................................................ System Model 18

......................................................... Transmitted TH-PAM UWB Signal 18 ................................................................... Transmitted Interferer Signals 19

........................................................................................ Received Signal -19 ........................................................................................... UWB Receiver 21

TH-PAM Performance in AWGN Channel .................................................. 23 TH-PAM Performance in Flat-Fading Channel ............................................ 26

........................................................................... Unfaded-Faded Scenario 26 Faded-Faded Scenario ............................................................................... 27

Results on the Coexistence of TH-PAM UWB Systems with ............................................................................... Narrowband Interference 28

........................................................................... Unfaded-Faded Scenario 29 ............................................................................... Faded-Faded Scenario 32

2.6 Comparison of TH-PAM and TH-PPM Systems Based on Power ................................................................................. Spectral Characteristics 36

Chapter 3 Experimental Analysis on the Performance of TH-PAM UWB Radio in Coexistence with Narrowband Interference .............................. 47

.................................................................................................. Introduction -47 UWB Radio Operation .................................................................................. 47

System Overview ....................................................................................... 47 .......................................... UWB Radio Synchronization and Modulation 51

Narrowband Tone Interferer .......................................................................... 54 Performance of TH-PAM UWB Radio without Interference ........................ 56

UWB System in a LOS Environment ........................................................ 56 UWB System in a NLOS Environment ..................................................... 59

Performance of TH-PAM UWB Radio in the Presence of ........................................................................... Narrowband Interference 6 1

LOS Scenario ............................................................................................. 61 .......................................................................................... NLOS Scenario 64

Effect of the Tone Interferer's Frequency on TH-PAM UWB .................................................................................................. Performance -66

Chapter 4 Conclusions ..................................................................................................... 68

References ......................................................................................................................... 70

vii

LIST OF FIGURES

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Figure 1 1

Figure 12

UWB signal design points representing the concept of partial bandwidth with cut-off frequencies at -10 dB from the PSD mask limit. ............................................................................................................... 4 FCC radiation limits for indoor UWB communication applications

Gaussian pulse corresponding to the generated waveform in the UWB transmitter. ........................................................................................... 8 First derivative of the Gaussian pulse corresponding to the transmitted waveform (The UWB antenna is modelled as a differentiation block in the time domain due to its high pass behavior). ....................................................................................................... .9 Second derivative of the Gaussian pulse corresponding to the received waveform at the UWB receiver. ...................................................... 9

Time-hopping system concept in which each bit time is divided into Ns disjoint frames each lasting Tf seconds. Each frame is further divided into Nc bins each lasting Tc seconds with the monocycle pulse placement chosen by means of a time- hopping code. ....................... 11 (1) PAM modulation where bit 0 is transmitted by inverting the pulse. (2) PPM modulation where bit 0 is transmitted by shifting the pulse by a small quantity S in time. ............................................................ 12 Matched filter receiver showing the AWGN, UWB signal, and the

........................................................................ narrowband tone interferers. 18 BEP for the TH-PAM system in the unfaded-faded scenario. The performance is evaluated for different values of SIR ratios. The tone interferer has frequency 4.5 GHz. ................................................................ 30 BEP as a function of tone interferer's frequency for the TH-PAM system in the unfaded-faded scenario at fixed SNR=12 dB and SIR=-10 dB. ................................................................................................. 31 BEP for the TH-PAM system in the faded-faded scenario. The performance is evaluated for different values of the fading parameter m. The tone interferer has frequency 4.5 GHz with SIR =- 10 dB. ........................................................................................................... 33 BEP for the TH-PAM system in the faded-faded scenario. The performance is evaluated for different values of the fading parameter m. The tone interferer has frequency 4.5 GHz with SIR =- 15 dB. .............. 34

... V l l l

Figure 13 BEP as a function of tone interferer's frequency for the TH-PAM system in the faded-faded scenario at fixed SNR=12 dB and SIR=-

Figure 14

Figure 15

Figure 16 Figure 17 Figure 18

Figure 19

Figure 20

Figure 21

Figure 22 Figure 23 Figure 24

Figure 25

Figure 26

Figure 27 Figure 28

Figure 29 Figure 30

Figure 3 1

Figure 32

10 dB. ........................................................................................................... 35

PSD of UWB TH-PAM system with Tw=0.227 ns, Tc=2 ns, Tf50 ns, Nc=25, and Ns=8. .................................................................................... 38 PSD for the code portion of the UWB TH-PAM system with Tw=0.227 ns, Tc=2 ns, Tf50 ns, Nc=25, and Ns=8. ..................................... 39 TH-PAM code PSD in the spectrum band 0 to lGHz with Nc=lO. ..... .. . . .. ..40 TH-PAM code PSD in the spectrum band 0 to lGHz with Nc=25. ............. 41 BEP as a function of tone interferer's frequency for the TH-PAM system in the unfaded-faded scenario at fixed SNR=12 dB and SIR=-10 dB. Ns=8, Nc=8. ............................................................................ 42 BEP as a function of tone interferer's frequency for the TH-PAM system in the unfaded-faded scenario at fixed SNR=12 dB and SIR=-10 dB. Ns=8, Nc=25. ......................................................................... 43 PSD of UWB TH-PPM system with Tw=0.227 ns, Tc=2 ns, Tf50 ns, Nc=25, 6=0.15 ns, and Ns=8. ................................................................. 44 PSD for the code portion of the UWB TH-PPM system with Tw=0.227 ns, Tc=2 ns, T ~ 5 0 ns, Nc=25, 6=0.15 ns, and Ns=8. ................ 45 PulseOnTM UWB radio's Transmitter top level schematic [37] ................... 48 PulseOnTM UWB radio's receiver top level schematic [37] .. .. .. ... .. .. .. .. .. .. .. ..50 An UWB radio packet consists of an Acquisition preamble and a Payload. The Acquisition preamble allows the receiver to synchronize to the transmitter. . ..... .. .. ... .. ... .. .. . .. .. . .. .. . .. .. ... .. .. . .. .. .. .. . . ... .. .. ... . . . .5 1 UWB radio's synchronization mechanism. The receiver steps its receive time window until it reaches a synchronization threshold. In this illustration, the receiver needs to further step in time since the threshold value is not satisfied. .................................................................... 52 UWB radio's synchronization mechanism. In this illustration, the receiver has met the synchronization threshold and the transmitter and the receiver are synchronized. ............................................................... 53 Schematic of the Discone antenna ................................................................ 55

Bit error rate (BER) measurement configuration. A bit error pattern known both to the transmitter and receiver is used for calculating the BER. ....... .. . .... .. .. ... .. .. ..... .. .. .. . .. .. .. ... .. .. .. . .. .. ..... .. . .. .. . .. .. ... .. ... .. .. .. . . ..... .. .. .. .. .. .. ..56 BER measurement setup for a LOS environment. ....................................... 57 Performance measurement as a function of UWB transmitter- receiver separation in a LOS scenario for Ns=8 and Ns=16. ...... .. .. .. .. . . . . .. .. . .58 Experiment setup for performance measurement of UWB link in a NLOS environment. .. .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . S 9

Performance measurement as a function of UWB transmitter- receiver separation in a NLOS scenario for Ns=8 and Ns=16. ..................... 60

Figure 33 Experiment setup for the unfaded-faded scenario. The UWB link has clear LOS while the tone interferer undergoes Rayleigh fading. ................. 62

Figure 34 Performance of TH-PAM as a function of SIR with SNR=12dB. The tone interferer has frequency 4.5 GHz. The UWB link has clear LOS while the tone interferer undergoes Rayleigh fading. .................................. 63

Figure 35 Experiment setup for the faded-faded scenario. In this case neither the UWB transmitter nor the tone interferer has a LOS to the UWB receiver. ....................................................................................................... .64

Figure 36 Performance of TH-PAM as a function of SIR with SNR=15dB. The tone interferer has frequency 4.5 GHz. ........................................................ 65

Figure 37 BEP as a function of tone interferer's frequency for the TH-PAM system in the faded-faded scenario at fixed SNR=12 dB and SIR=- 10 dB. N,=511, N,=16. ................................................................................. 67

LIST OF TABLES

Table 1 UWB radio's specifications. ........................................................................ 48

GLOSSARY

AWGN BAN BEP BER CDMA DS FCC FM IEEE IR LOS MF NLOS PAM PAN PN PPM PRF PSD PTD RF RX S/H S INR SIR SNR SS TH-PAM TH-PPM TX UHF USB UWB VHF WLAN WPAN

Additive White Gaussian Noise Body Area Network Bit Error Probability Bit Error Rate Code Division Multiple Access Direct Sequence US Federal Communications Commission Frequency Modulation The Institute of Electrical and Electronics Engineers Impulse Radio Line-of-Sight Matched Filter Non-Line-of-Sight Pulse Amplitude Modulation Personal Area Network Pseudo-random Noise Pulse Position Modulation Pulse Repetition Frequency Power Spectral Density Programmable Time Delay Radio Frequency Receiver Sample and Hold Signal-to-Interference-Plus-Noise Ratio Signal-to-Interference Ratio Signal-to-Noise Ratio Spread Spectrum Time-Hopping Pulse Amplitude Modulation Time-Hopping Pulse Position Modulation Transmitter Ultra High Frequency Universal Serial Bus Ultra Wideband Very high frequency Wireless Local Area Network Wireless Personal Area Network

CHAPTER 1 INTRODUCTION TO ULTRA WIDEBAND (UWB)

SIGNALS AND SYSTEMS

1.1 Brief History of Ultra Widebarnd

As it happens, wireless communications started out as ultra wideband and later

marched towards narrowband usage of the radio spectrum. In 1864, James Clerk Maxwell

formulated the concept of electricity and magnetism using mathematics in his equations

of electromagnetism. Maxwell predicted that energy could be transported through space

at the velocity of light by the action of electric and magnetic waves. This concept linked

light and electromagnetic waves as being the same phenomenon.

It was not until 1886 when Heinrich Rudolf Hertz put into practice what Maxwell

had proposed with mathematics. Hertz calculated that an electric current oscillating in a

conducting wire would radiate electromagnetic waves into the surrounding space. In his

laboratory, using a spark-gap apparatus, Hertz generated radio energy and detected such

oscillations over a distance of several meters [23]. The era of wireless had begun and in

fact, the spark-gap apparatus used in the first: wireless transmitters is classified as ultra

wideband.

The transmitters of the time utilized spark gap apparatus that emitted wideband

noisy signals. The receivers in use were simple amplitude detectors that could not

effectively gather the wideband energy. This resulted in poor signal to noise (SNR)

performance, thus requiring large transmitter powers to achieve longer ranges of

coverage. Excessively wide signal bandwidths and high transmitter power meant

significant spectrum sharing problems and plenty of interference. These two issues drove

wireless communications towards narrower and narrower bandwidths. In the United

States, the Radio Act of 23 July 1912 stepped to resolve the interference issue by

mandating the narrowest bandwidths possible and codified the separation of wireless

services by wavelength [17].

In the United States the Federal Communications Commission (FCC) that has

broad regulation powers in both wire-line-based communications and radio-based

communications, had traditionally favored narrowband radios that concentrate their

power in fairly narrow channels within the radio frequency spectrum. However, the

number of available channels became limited as the number of users sharing the spectrum

increased.

In 1948, Claude Shannon offered a new paradigm redefining the relationship

among noise, power density, and information capacity. Shannon stated that under certain

specific conditions, the more an information signal is spread in bandwidth in a way that

makes the information signal appear like background noise, the more information it is

capable of holding. Consequently, an alternative to transmitting a low bandwidth signal

with a high power density would be to use a wide bandwidth signal and low power

density [35]. Shannon's view led to spread spectrum (SS) communications in which the

signals are intentionally spread using a special family of digital codes to many times their

information bandwidth.

By 1985, the FCC began allowing spectrum technology in which multiple users

would be separated by means of direct-sequence codes rather than by discrete frequency

channels which started development of CDMA cellular systems. At the same time,

throughout the last half of the twentieth century much experimentation took place for

impulse radar transmission with primary focus on military and tracking investigations.

The experiments led to "impulse radio (IR)" later nicknamed UWB radio, presented in a

paper by R.A. Scholtz at MILCOM '93 [I]. Impulses are short pulses in time domain that

transform into wideband signals in frequency domain. Commercial experiments,

inventions and petitions before the FCC in the 1980s and 1990s led to the landmark FCC

regulation of 2002 that permits low-power UWB technology for commercial

development. Under the new FCC regulations, multiple users could share spectrum

previously allocated to other users. The frequency band allocated by FCC for indoor

UWB communications is between 3.1 and 10.6 GHz which makes the coexistence

between UWB devices and existing narrowband one of the most important UWB

research topics.

1.2 Overview of UWB and the Coexistence Issue

Ultra wideband radio is based on the radiation of waveforms produced by a

sequence of nanosecond pulses. Due to the jmpulse characteristic of UWB pulses, the

transmitted UWB signal occupies a wide bandwidth as large as 7.5 GHz. Contrary to the

usual narrowband transmission where the baseband signal is modulated to a reference

radio frequency, in UWB no carrier frequency is used for transmission; hence, the term

'carrierless' is often associated with UWB [24].

The Federal Communications Commission (FCC) regulations shape the general

characteristics of UWB signals and systems. To define what is exactly meant by UWB

signal, the following fractional bandwidth definition has traditionally been employed

where fL and fH correspond to lower and higher cutoff frequencies of the -10 dl3

emission points. The fractional bandwidth is effectively the ratio of signal bandwidth (10

dl3) to the center frequency. UWB signals are those signals that have fractional

bandwidths greater than 25 percent. The minimum measured bandwidth at the 10 dl3

points below the peak emission level is 500 MHz [28]. Figure 1 shows the notion of

fractional bandwidth.

Figure 1 UWB signal design points representing the concept of partial bandwidth with cut- off frequencies at -10 dB from the PSD mask limit.

Wideband signals are defined as signals with fractional bandwidths between one

and twenty five percent, while narrowband signals have fractional bandwidths less than

one percent [I 81.

The permissible emission levels for UWB signals in the UWB band are set at -41

dBm1MHz in the frequency band between 3.1 and 10.6 GHz. The FCC emission mask

limit for the operation of UWB devices for indoor use is shown in Figure 2 [28].

. , . .

: - Indoor L

Frequency in GHz

Figure 2 FCC radiation limits for indoor UWB communication applications [28].

The emission mask specifies limits on effective isotropically radiated power

(EIRP), that is, power radiated by an antenna having a gain of 1. Therefore, the total

emitted power is found by computing the area under the curve described by the power

spectral density (PSD) and multiplying it by the FCC mask limit. If all available spectrum

from 3.1 to 10.6 GHz were perfectly filled with the maximum allowed signal PSD, the

total EIRP for a transmitted signal would be

This represents the absolute maximum possible EIRP limit for UWB under these

particular regulations. The FCC regulated power level for indoor wireless

communications is very low (0.556 mW/MHz), which allows UWB technology to

overlay already available services such as the lEEE 802.1 1 wireless local area networks

(WLANs) without causing harmful interference in the 3.1-10.6 GHz band. However, the

interference from narrowband transmitters to UWB receivers could be problematic. First,

the total power of a narrowband transmission generally will fall within the UWB

passband. Second, a wide UWB passband (several hundred MHz or more) may span

multiple narrowband transmitters, some of which may be very powerful or may be

operating in close proximity to the UWB receiver [36]. This necessitates extensive

research in the area of coexistence, especially the issue of interference coming from

narrowband systems operating in the allotted spectrum for UWB devices.

1.3 UWB Modulation and Multiple Access Formats

UWB transmission is by means of radiating pulses of short duration. Depending

on the UWB system, each information bit is transmitted using one or more pulses. A

number of modulation schemes may be used with UWB systems. Two widely used

modulation schemes include pulse position modulation (PPM) and pulse amplitude

modulation (PAM). Furthermore, UWB communications systems for the most part utilize

either time-hopping (TH) or direct-sequence (DS) techniques for multiple-access and

spectrum spreading. Since modulation and multiple access are independent processes,

TH-UWB and DS-UWB can adopt either PPM or PAM for data modulation [30]. An

overview of these techniques is presented in the following subsections.

1.3.1 UWB Pulse Shape

There are many signals that satisfy the fractional bandwidth definition for UWB

signals. However, the most commonly adopted pulse for UWB communications is the

monocycle Gaussian pulse with temporal extensions of fractions of one nanosecond. The

Gaussian pulse has a duration of T, seconds and is essentially zero out of the time span

[o,T,]. The pulse duration is less than one nanosecond which sets the system bandwidth

defined approximately as l/T, , to a few GHz. The received pulse waveform is the second

derivative of a Gaussian monocycle pulse mathematically defined as

w ( t ) = - 1-4n - exp -2n - K[ i;l] [ [;I) In deriving the expression for the received waveform, each antenna is realized as a

differentiator [3] [4]. Furthermore, Ew = 3Tw/8 is the energy of the second derivative

Gaussian monocycle and the factor ensures that the signal is normalized to unit

energy. The Fourier transform of w ( t ) is given by

The time domain representation for the monocycle Gaussian, first derivative

(transmitted), and second derivative (received) waveforms are illustrated in Figures 3,4,

and 5, respectively. The pulse transmission through the antenna requires the use of a

pulse with electromagnetic zero. In carrier modulation, this requirement is satisfied by

use of the sinusoidal carrier frequency. In UWB transmission the electromagnetic zero is

implemented by differentiating the monocycle pulse so that the waveform can radiate

through the antenna. The antenna's filtering effect can be modeled as a differentiator.

Figure 3 Gaussian pulse corresponding to the generated waveform in the UWB transmitter.

Figure 4 First derivative of the Gaussian pulse corresponding to the transmitted waveform (The UWB antenna is modelled as a differentiation block in the time domain due to its high pass behavior).

Figure 5 Second derivative of the Gaussian pulse corresponding to the received waveform at the UWB receiver.

1.3.2 Time-Hopping UWB (TH-UWB) Transmission Scheme

The continuous pulse generation in UWB systems leads to strong spectral lines in

the transmitted signal at multiples of the pulse repetition frequency. Data modulation

typically occurs in three stages. First, a pulse train is generated. Second, a randomizing

technique is applied to break up the spectrum of the pulse train. Third, the data

modulation is applied to carry the information. Two main approaches for randomizing the

pulse train in addition to accommodating multiple users, are time hopping and direct

sequence techniques [29]. Time hopping and direct sequence are well known techniques

used in spread spectrum (SS) communication systems such as code division multiple

access (CDMA).

In a TH-UWB system the waveform construction is as the following: each bit

time T, is divided into N, disjoint frames, each lasting Tf seconds ( T , = N,Tf ). In each

frame, a monocycle is placed with its position being chosen randomly inside the frame.

The frame is further divided into N, disjoint bins (chips) each lasting T,seconds

(Tf 2 N,T, ). The monocycle is placed inside a randomly chosen bin by means of a time-

hopping code C, . The time-hopping code is a pseudo-random noise (PN) sequence which

is uniformly distributed in [o,. . . , N, - 11. The waveform construction for a TH-UWB is

illustrated in Figure 6.

Figure 6 Time-hopping multiple-access mechanism in which each bit time is divided into N, disjoint frames each lasting Tf seconds. Each frame is further divided into Nc bins each lasting Tc seconds with the monocycle pulse placement chosen by means of a time- hopping code.

Two of the commonly used modulation schemes with TH-UWB systems include

pulse position modulation (PPM) and pulse amplitude modulation (PAM). In a binary

PAM system, the pulses corresponding to a 1-bit remain unchanged while the pulses

corresponding to a 0-bit are inverted. In the case of binary PPM system, the pulses

corresponding to a 1-bit remain unchanged while the pulses corresponding to a 0-bit are

shifted in time by a small quantity 6 . The PAM and PPM modulation schemes are

illustrated in Figure 7.

(2) PPM

Figure 7 (1) PAM modulation where bit 0 is transmitted by inverting the pulse. (2) PPM modulation where bit 0 is transmitted by shifting the pulse by a small quantity 6 in time.

In TH systems, each user is assigned a different PN code. For instance, c?'

corresponds to the PN code associated to the mth user. Additionally, c:"' takes on integer

values from 0 to Nc -1 for each frame and each frame should see independent TH shift

dictated by 12:"'. The information signal for the mth user with PPM modulation is

- Ns-1

dm) ( t ) = x x w(t - iWV,Tf - jTf - c?'T, -dim'@ k=- j=O

for PAM modulation as follows

where dim' corresponds to the information bit sequence of the mth user and w(t)

represents the monocycle Gaussian waveform carrying each bit. In a binary PAM and

PPM system the information bit sequence d p ) are mapped to {-1,1} and {0,1}

respectively.

1.3.3 Direct-Sequence UWB (DS-UWB) Transmission Scheme

Direct sequence can also be used with UWB for multiple access either with PAM

or PPM modulation since modulation and multiple access are independent processes. In

this technique, initially the transmitted binary sequence is coded with a pseudo-random

noise (PN) sequence. Input symbols are modulated onto either the amplitude or the

relative positions of each sequence of pulses. The transmitted direct-sequence UWB

signal with PAM for the mth user has the following form [6]

m N,v -1

dm) (t) = dc x x cy'dp)w(t - kNsTf - jTf )

1.4 UWB Signal Propagation

A clear line-of-sight (LOS) rarely exists between a transmitter and a receiver in an

indoor environment. As a result, multiple time-delayed versions of a signal will arrive at

the receiver. The time delay is due to the different paths that the signal travels. The

channel response to a narrowband signal is different from an UWB signal. The fading

characteristics observed in narrowband systems which is due to time harmonics is not

present in UWB systems since the signals are short in time and have high time resolution.

Narrowband systems have low time resolution, which causes large number of

multipath components to interact leading to the Gaussian distribution modeling of fading.

However, the Gaussian approximation does not apply to UWB channels due to the high

time resolution feature of UWB. Much experimentation has been done confirming the

fact that the arriving multipath components in UWB channel are resolvable [4].

Consequently, most previous work on narrowband channel modeling does not apply to

UWB channels since the central limit theorem, which is the basis for the Gaussian

approximation in narrowband channels, does not hold for UWB. In this thesis, we

consider the channel model proposed in [16] with details provided in Chapter 2.

1.5 UWB Benefits and Applications

UWB technology has many potential advantages due to its wide transmission

bandwidths including: 1) no significant multipath fading due to fine time resolution [9];

2) accurate position location and ranging, due to fine time resolution [8]; 3) multiple

access due to wide transmission bandwidths [lo]; 4) possibility of extremely high data

rates [ l l ] ; 5) covert communications due to low transmission power operation [18]; and

6) possible easier material penetration due to low frequency components.

The UWB applications are distributed amongst three categories [17]

Communications and sensors

Position location and tracking

Radar

An important application for UWB technology is in the Wireless Personal Area

Networks (WPANs), where data are transmitted over distances of 10 m or less. The

applications for WPANs are classified into two areas, low or high data rates [7] . Both

require low power and high capacity, which are the prominent qualities of UWB. The

high-data-rate applications (100 Mbitls and up) are mainly related to consumer

electronics (digital TV) and computer networks (wireless USB), while low-data-rate

applications can include other consumer-electronics applications (e.g. audio streaming),

as well as tasks that were traditionally treated by Bluetooth and infrared devices.

Irrespective of the data rate considerations, the envisioned environments for UWB

are mainly office and residential structures, with distances between 1 and 10 m. Both the

case of fixed-location devices (e.g. mounted on a TV or PC), and of person-held or body-

worn devices are of interest. A special case of WPANs are "Body Area Networks"

(BAN), where the communication is between two body-worn devices [12]. UWB Sensors

also find applications in medical situations in order to free the patient from the tangle of

wired sensors.

1.6 Project Goals

The goal of this project is to investigate the coexistence of UWB systems with

other narrowband systems operating in the same bandwidth. In particular, we analyze the

effect of narrowband interference on the performance of UWB systems by means of

theoretical and experimental analysis. Based on the lack of published work on the

coexistence of UWB systems via experimentation, we decided to look into the issue of

coexistence more closely by means of actual experimental measurements.

In the first stage of the project, presented in Chapter 2, we focus on analytical

analysis of the coexistence issue. The analytical analysis involves the derivation of

closed-form expressions for the bit error probability of TH-PAM UWB systems in the

presence of narrowband interferers for AWGN and flat-fading channels. We present the

results of our performance analysis for the TH-PAM UWB system in different operating

scenarios. In particular, we evaluate the performance of the TH-PAM UWB system for a

range of operating frequencies of the narrowband system and investigate the regions that

produce the worst-case performance.

Moreover, we compare the power spectral densities of TH-PAM and TH-PPM

UWB systems with regards to the coexistence issue and based on our results, we suggest

strategies for making the UWB system less susceptible to interference coming from

narrowband systems operating in the same bandwidth and vice versa.

In the next stage of the project, presented in Chapter 3, we investigate the

coexistence issue by means of experimentation. In particular, we examine the

performance of a TH-PAM UWB radio in different operating scenarios.

The uniqueness of the experimental aspect of the project lies on the challenges

that are encountered when dealing with a new technology such as UWB. As a result, a

considerable amount of time and effort was invested for planning a test bed for the

experimentation. I strongly hope that the efforts put into building the UWB laboratory

will promote further research in the area of UWB systems.

CHAPTER 2 ON THE PERFORMANCE OF TH-UWB SYSTEMS

IN THE PRESENCE OF NARROWBAND INTERFERENCE

2.1 Introduction

In this chapter, our goal is to analyze the performance of TH-PAM UWB systems

in the presence of narrowband interference. In particular, we derive closed-form bit error

probability (BEP) expressions for the performance of TH-PAM UWB systems in the

presence of narrowband interference in different operating scenarios.

The narrowband interferers are approximated as independent asynchronous tone

interferers and the UWB fading channel is modeled as having Nakagami-m distribution

as suggested in [16]. We present the system model for general matched filter (MF)

reception in the next section. With the subsequent sections extending the analysis to deal

with performance analysis of TH-PAM UWB transmission scheme in the presence of

narrowband interference in additive white Gaussian noise (AWGN) channels and flat-

fading channels. The theoretical analysis presented in this chapter is based on [14] in

which the performance of TH-PPM UWB systems in the presence of narrowband

interference is evaluated and [15] in which a general framework for analyzing the

performance of UWB systems in presence of interference is evaluated. In section 2.5, we

investigate the issue of coexistence of TH-PAM UWB systems in the presence of

narrowband interference and in section 2.6, we present a comparison between TH-PAM

and TH-PPM modulation schemes based on their spectral characteristics in view of the

coexistence issue.

2.2 System Model

In this section, we consider the reception of a general binary UWB system in the

presence of independent asynchronous tone interferers with arbitrary frequencies. The

block diagram of the system for a single user scenario is presented in Figure 8.

Figure 8 Matched filter receiver showing the AWGN, UWB signal, and the narrowband tone interferers.

In the following subsections, we analyze the elements of the presented system

including the UWB signal, tone interferers, and matched filter reception.

2.2.1 Transmitted TH-PAM UWB Signal

The transmitted TH-PAM UWB signal is given by

where E,represents the energy of each transmitted bit, T, is the bit duration with 1/T,

denoting the bit rate, N, is the number of pulses required to transmit a single information

bit, Tf is the frame length, {c,} is the TH sequence, and T, is the chip time . The

information bits di E {0,1} are assumed to be independent and equiprobable. The

information bits are assumed to be mapped to the set {-1,1} intended for antipodal

modulation. Furthermore, w(t) is the monocycle Gaussian pulse normalized so that

2

[w(t)] d(t) = 1 . We now consider the narrowband tone interferers.

2.2.2 Transmitted Interferer Signals

As shown in Figure 8, the narrowband interferers are modeled as asynchronous

tone interferers expressed as

where p, is the average transmitted interferer power, f n is the interferer carrier

frequency and &, is the phase modeled as a random variable uniformly distributed over

[O, 2n) .

2.2.3 Received Signal

The communication channel is considered to be time-invariant with impulse

response bB (t) for the UWB signal and hIn (t) for the N, interferer signals. The overall

received signal r(t) due to the UWB signal, N, independent asynchronous tone

interferers, and the additive white Gaussian noise (AWGN) is

where n(t) is the AWGN with two-sided power spectral density equal to N0/2. The term

rwB(t;di) is the channel response to the transmitted UWB signal and is mathematically

expressed by the convolution of the UWB channel impulse response and the transmitted

TH-PAM UWB as

in which, w(t) is modeled as the second derivative of the monocycle Gaussian pulse

taking into account the antenna effects and is given in equation (1.4). Furthermore,

without loss of generality it is assumed that the N, narrowband interferers experience

flat-fading, i.e., hIn = aIn S(t - z,) , n = 1,. . . , N, . Consequently, the interference term in

equation (2.3) is expressed as

Each narrowband interferer has channel gain aIn , frequency fn , time shift rn , and phase

qjn . The interferers are assumed to be asynchronous, hence @, are modeled as

independent identically distributed (i.i.d.) random variables, uniformly distributed over

[o, 2n) . Moreover, without loss of generality it is assumed that the average channel gains

for the UWB signal and the interferers are normalized, i.e., ~[a;,] = 1 and at ] = 1

for n = 1,. . . , N , where the UWB channel gain is represented by the random variable

2.2.4 UWB Receiver

The optimum receiver in the presence of AWGN is a matched filter (MF) or

equivalently a correlator followed by an integrator [22]. Following the same approach as

in [14] and [15], the UWB receiver is considered to be a MF with template waveform

given by

where rwB (t; 0) and rwB (t; 1) denote the two received UWB waveforms corresponding

to bits '0' and '1' at the receiver given in equation (2.4). Note that, the template

waveform estimation affects the performance of the system. One method for estimating

the template waveform is by sending a reference signal for each pulse transmission.

Another method is by using a set of reference pulses at the receiver which correspond to

different channel environments.

Using Fourier transform,3{-) , the transfer function of the matched filter at the

appropriate sampling time to is expressed as

The magnitude of the MF transfer function in equation (2 .7) is

I H ( f ) I = IH0( f ) I .19{&B (t)}I with I H , ( f )I denoting the magnitude of the transfer

function due to the transmitted UWB signal and the second term dependent on the UWB

channel impulse response. The magnitude of the matched filter transfer function due to

the transmitted TH-PAM UWB signal is computed as

where W ( f ) is the Fourier transform of the received UWB pulse given in equation (1.4).

We assume that the UWB pulses introduce negligible intersymbol interference

and the receiver is perfectly synchronized with the transmitter. The output of the matched

filter at sampling instance to is expressed as

In equation (2.9), I H ( fn)l is the matched filter's impulse response to the interference

signal and @n is the random phase uniformly distributed over [ 0 , 2 n ) . We have

represented the phase term 2 n fn(t -z,) + a r g { H ( f n ) } with a slight abuse of notation as

No @,, . Also, no denotes sample noise with zero mean and variance o2 = (-1 f r v 2 ( t ) d t . 2

We next consider the performance of the described system in AWGN, and frequency flat

fading.

2.3 TH-PAM Performance in AWGN Channel

With the system model represented in the previous section, we consider the

detection of bit d = 0 in AWGN and derive the BEP of the matched filter receiver. In this

scenario, the UWB channel response is simply bB = S(t) and the channel gains of the

interferers are = 1. With this in mind, the output of the matched filter can be written

where so is the desired received UWB signal and is evaluated as

Here, is the received bit energy of the UWB signal and p is the correlation

coefficient between the two received UWB waveforms corresponding to bits 0 and 1.

Note than for TH-PAM waveform the correlation value is given by -1.

The transmitted tone interference signals are assumed to be asynchronous with

arbitrary frequencies which results in the output of matched filter due to the interferers to

be expressed as

Furthermore, the no term is the noise sample with zero mean and variance equal

Referring to equation (1.16), the BEP of the matched filter is expressed as

Equation (2.14) could be represented likewise as the cumulative distribution

function (CDF) of 6 = I +no as P, = Fc(-so) . Furthermore, noting that the interferers

and noise are independents, the BEP is evaluated easily using the characteristic function

of the random variable 6 . The characteristic function of 6 is evaluated as

Qc ( w ) = ~ [ e ' ~ ~ ] = @, ( w ) . @ , ( w ) (2.15)

The characteristic function due to the zero-mean noise term is easily calculated as

-3oX

@, ( w ) = e [13] where o2 = No(l - p ) . Also, the characteristic function, @, ( w ) , is

calculated using the 0" order Bessel function of the first kind. In deriving this

characteristic function it is important to note that the phase term of the interferers, @, , is a

random variable with uniform dstribution over the interval [o, 2 n ) . The corresponding

characteristic function is

The BEP is expressed using the inversion theorem [31] by noting that J is an

even random variable:

1 1 w sin w w 2 d p e 2 ? T = - 10, (--)Texp[-F)dw

Substituting the equation (2.16) into equation (2.17), the BEP can be written as a function

c Eb of signal-to-noise ratio (SNR) and the signal-to-interference ratios (SIRS), - = - , as en enT,

We note that in absence of interference we have I = 0 and consequently

0, (o) = 1 which suggests the BEP can be evaluated easily as Q (dw) where Q (.)

is the Gaussian Q-function. In addition, expression (1.24) could be rewritten with

0 , ( w ) = l as

Using equations (2.1 8) and (2.19) the BEP as derived in [14] is expressed as

sin w w2 No 1 exp(--.--

w 2 E, 1 - p )dm

2.4 TH-PAM Performance in Flat-Fading Channel

In this section the performance of the matched filter receiver is evaluated in the

presence of independent Rayleigh fading on multiple interferers, Nakagami-m fading on

the UWB signal, and AWGN. Initially we consider the scenario in which the UWB signal

is unfaded and then continue with the scenario in which the UWB signal undergoes

Nakagami-m fading.

2.4.1 Unfaded-Faded Scenario

This scenario represents the case of short-range transmissions that is characterized

by line-of-sight (LOS) propagation with fixed positions for both the transmitter and the

receiver. The general form of the received signal is expressed as

where E, is the average received energy per bit, pin is the average received power of the

nth interferer, a," are independent Rayleigh distributed random variables with unit power

and amB is the is the UWB fading amplitude. In this scenario the amplitude of the UWB

signal remains constant.

The output of the matched filter at the appropriate sampling time to is given by

In this expression, each ct;" COS(@~) term is a zero-mean Gaussian random variable with

variance 0.5. This is due to the fact that aIn are Rayleigh distributed and the @,,are

uniformly distributed over the interval [0,2n). Therefore, the interference term

2 a," & 1 Ho ( fn)l COS(@~) becomes a zero-mean Gaussian random variable with n =l

Nl

variances; = XP," IH0( fn)12 . The sum of noise and interference can be expressed as a n=l

random variable 5 with Gaussian distribution and variance a2 +a:. Since the total

disturbance has Gaussian distribution, we can express the BEP as

where y is the average signal-to-interference-plus-noise ratio (SINR) as a function of

Eb/No and the average signal-to-interference (SIR) ratios corresponding to N,

interferers. The SINR as derived in 1141 is given by

2.4.2 Faded-Faded Scenario

In this scenario both the UWB signal and the narrowband interferers are affected

by fading which could be due to non-line-of-sight (NLOS) propagation. As in the

unfaded-faded scenario the narrowband interferers experience Rayleigh fading while it is

assumed that the UWB signal follows the Nakagarni-m distribution according to the

channel model presented in 1161. The probability density function for Nakagami-m

distribution is given by

where s-2 is the mean of the normalized received energies s-2 = EraUWB] = l . The

parameter m is the fading parameter that controls the severity of the fading conditions.

Note that m=l corresponds to Rayleigh fading and in the worst case when m=0.5, the

distribution follows a one-sided Gaussian distribution. The BEP for this scenario is

expressed by conditioning the probability of error on the random variable aUWB as

The BEP can be evaluated by averaging the conditional probability over the

random variable aUWB which has a Nakagami-m distribution given in equation (2.25).

The result is obtained with the use of hypergeometric function as derived in [14] and [15]

where the term ,F,(., .; .; .) in equation (2.27) is the hypergeometric function. A detail

analysis of the hypergeometric function is provided in [41]. In the next section we

present the performance of TH-PAM for different values of the fading parameter m.

2.5 Results on the Coexistence of TH-PAM UWB Systems with Narrowband Interference

In this section, we investigate the performance of TH-PAM UWB systems in the

presence of narrowband interference based on the results obtained in sections (2.4.1) and

(2.4.2), which correspond to the unfaded-faded and faded-faded scenarios, respectively.

In particular, we investigate the BEP performance behavior of the TH-PAM UWB

system by varying the frequency of the tone interferer and observing the worst-case

performance of the system.

In both scenarios, we consider a TH-PAM system in the presence of a tone

interferer with the frame length Tf = 104.1 ns, the chip time Tc = 0.6 ns, N, = 8 , Nc = 8 ,

and Tw = 0.227 ns. The modulation format uses an antipodal signal with correlation value

p=-1.

2.5.1 Unfaded-Faded Scenario

In this scenario, we consider the case in which the interferer undergoes Rayleigh

fading while the UWB signal is unfaded. This scenario represents the case of short-range

transmissions with LOS propagation and fixed positions for both the UWB transmitter

and the receiver. The BEP for this scenario is evaluated using equation (2.23) for a tone

interference operating at frequency f, = 4.5 GHz. The BEP as a function of E , / N , for

different values of SIR is shown in Figure 9 .

Figure 9 BEP for the TH-PAM system in the unfaded-faded scenario. The performance is evaluated for different values of SIR ratios. The tone interferer has frequency 4.5 GHz.

As can be seen in Figure 9, the performance of the TH-PAM UWB becomes

worse as the value of SIR is decreased. Additionally, the performance curves follow a

waterfall shape at high values of SIR but the form of the curves are altered at low SIR

values. The performance is severely degraded at low SIR values which in this example

correspond to SIR=-25 dB and SIR=-30 dB. Consequently, the TH-PAM UWB system

can experience severe performance loss when operated in the presence of a narrowband

interferer with high transmit power even when the UWB transmitter and receiver are

operating in a LOS scenario.

We next, investigate the performance of the TH-PAM UWB system in the

unfaded-faded scenario when the tone interferer's frequency is varied from 1.5 GHz to

6.9 GHz at a 1 MHz resolution. We consider the case when the SNR and SIR ratios are at

12 dB and -10 dB, respectively. The BEP as a function of the tone interferer's frequency

was evaluated using equation (2.23) and the result is shown in Figure 10.

3 3.5 4 4.5 5 5.5 Tone Interferer's Frequency [GHz)

Figure 10 BEP as a function of tone interferer's frequency for the TH-PAM system in the unfaded-faded scenario at fixed SNR=12 dB and SIR=-10 dB.

As shown in Figure 10, the BEP for the TH-PAM system fluctuates for different

values of the tone interferer's frequency. The variation in the BEP is due to the term

containing the TH code in equation (2.8) which is directly proportional to the PSD for the

TH-PAM modulation format. The worst performance is observed when the tone

interferer's frequency is operating close to the frequencies which are integer multiples of

the chip time i.e., k/Tc Hz. In this example, these critical frequencies are at 1.6 GHz, 3.3

GHz, and 5 GHz which correspond to 1/T, Hz, 2/Tc Hz, and 3/Tc Hz, respectively.

From the coexistence point of view, the TH-PAM UWB system would experience severe

performance loss when a narrowband system operates at the critical frequencies of the

TH-UWB system. We will investigate this issue in more detail in section 2.6 of this

chapter.

2.5.2 Faded-Faded Scenario

In this scenario, we consider the case in which the UWB and the interferer both

undergo fading. As described in section (2.4.2), the interferer undergoes Rayleigh fading

while the UWB experiences Nakagami-m fading. This scenario represents the case of

short-range transmissions with NLOS propagation for the UWB link. The BEP for this

scenario is evaluated using equation (2.27) for a tone interference operating at frequency

f, = 4.5 GHz. The following two figures show the BEP as a function of E , / N , for

different values for the Nakagami-m fading parameter, m, and at a fixed SIR value.

1U . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . - - - - - - - - - - - - - 1 - - - - - - - - - - - - - - r - - - - - - - - . - - - - - - - - I - Tone Interferer. SIR=-lOdB 11 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Figure 11 BEP for the TH-PAM system in the faded-faded scenario. The performance is evaluated for different values of the fading parameter m. The tone interferer has frequency 4.5 GHz with SIR =-I0 dB.

Figure 12 BEP for the TH-PAM system in the faded-faded scenario. The performance is evaluated for different values of the fading parameter m. The tone interferer has frequency 4.5 GHz with SIR =-I5 dB.

As can be seen from the above two figures, the performance of the TH-PAM

UWB system is degraded at lower levels of SIR. Consequently, the TH-PAM UWB

system can experience severe performance loss when operated in the presence of a

narrowband interferer with high transmit power. Furthermore, the performance of the

TH-PAM UWB system is poorer in the faded-faded scenario compared to the unfaded-

faded scenario which is due to the fading of UWB signals.

We next, investigate the performance of the TH-PAM system in the faded-faded

scenario when the tone interferer's frequency is varied in order to address the issue of

coexistence. We consider the case when the SNR and SIR ratios are at 12 dB and -10 dB

respectively and the fading parameter has a value of m=2. In addition, the tone

interferer's frequency sweeps over the frequency band from 1.5 GHz to 6.9 GHz at a 1

MHz resolution. The BEP as a function of tone interferer's frequency was evaluated

using equation (2.27) and the result is shown in Figure 13 .

Tone Interferer's Frequency (GHz)

Figure 13 BEP as a function of tone interferer's frequency for the TH-PAM system in the faded-faded scenario at fmed SNR=12 dB and SIR=-10 dB.

As shown in Figure 13, the BEP for the TH-PAM system fluctuates as the tone

interferer's frequency changes. The variation in the BEP is due to the term containing the

TH code in equation (2.8) which is directly proportional to the PSD for the TH-PAM

modulation format.

As in the unfaded-faded scenario, the worst performance is observed when the

tone interferer's frequency is operating in the vicinity of the frequencies which are

integer multiples of the chip time i.e., k/T, Hz. In this example, these critical frequencies

are at 1.6 GHz, 3.3 GHz, and 5 GHz which correspond to 1/q Hz, 2/T, Hz, and 3/q

Hz, respectively. The BEP of the TH-PAM UWB system in the faded-faded scenario

exhibits less fluctuations compared to the unfaded-faded scenario. This is due to the

performance behavior of the UWB system at the specified SNR and SIR value as shown

in Figures 9 and 11. From the coexistence point of view, the TH-PAM UWB system

would experience severe performance loss when a narrowband system operates at the

critical frequencies of the TH-UWB system. We will investigate this issue in more detail

in the next section.

2.6 Comparison of TH-PAM and TH-PPM Systems Based on Power Spectral Characteristics

In this section, our goal is to address the coexistence issue concerning the

spectrum shape of the UWB modulation format. From spectral point of view, the

transmitted UWB monocycle pulse train produces energy spikes (peaks) at regular

intervals. Therefore, the low power level of the UWB is spread among the spikes. These

spikes cause harmful interference to already existing narrowband systems. In the same

way, the UWB system can suffer from narrowband interference operating over the region

where the spikes are present. In order to minimize the spectral spikes of the UWB

system, a randomization is utilized which varies the pulse-to-pulse time interval. This

randomization is achieved by the use of the PN code in TH systems. Moreover, the UWB

modulation format influences the spectral shape of the transmitted UWB signal. In

particular, we analyze the spectral shape for TH-PAM and TH-PPM modulation formats.

Following the notation set in section 1.3.2, we can express the PSD of the TH-

PAM signal derived in [20] as

N, sin (%NCf )' 1 sin (~~T,N,N, f )' prx-~*a(f)=pwB~w(f)I' +- I (2.28)

Nc sin(fl,f )' N: sin(flcf)'

where PwB is the total signal power, W( f ) is the Fourier transform of the monocycle

pulse and the term in the parenthesis is the code spectrum. We evaluated the PSD of a

TH-PAM UWB system with the frame length Tf = 50 ns, the chip time Tc = 2 ns,

N, = 8 , N, = 25 and T, = 0.227 ns . The TH code was generated using a pseudorandom

number generator and the peak value is normalized to 0 dBm. The PSD of the TH-PAM

UWB system is shown in Figure 14.

-50 0 1 2 3 4 5 6 7 8 9 10

Frequency [Hz) x l og

Figure 14 PSD of UWB TH-PAM system with T,=0.227 ns, Tc=2 ns, T ~ 5 0 ns, Nc=25, and N,=8.

The spectral shape of the TH-PAM is determined by the Fourier transform of the

monocycle pulse and the code spectrum. We notice spectral spikes occurring at integer

multiples of l/T, as evident in Figure 14. The spectral spikes are due to the TH-PAM

code spectral shape. To better clarify this point, we evaluate the TH code spectral shape

for the TH-PAM modulation format as presented in Figure 15.

Frequency (Hz)

Figure 15 PSD for the code portion of the UWB TH-PAM system with T,=0.227 ns, Tc=2 ns, T ~ 5 0 ns, Nc=25, and N,=8.

As apparent in Figure 15, the code spectrum exhibits peaks at critical frequencies

that are integer multiples of 1/? (500 MHz for this example). From the coexistence point

of view, when the frequency of the interfering tone is equal or close to a multiple of 11Tc

Hz, its effect on the performance of the TH-PAM system can be dramatic as reported in

sections (2.51) and (2.52). Furthermore, the spectral peaks cause harmful interference to

narrowband systems operating in the same spectrum band.

As reported in 1201, the height of the peaks at the critical frequencies are

proportional to N,' with their width proportional to l/NsNcq and one way of

eliminating the peaks is by setting in order to guarantee that the peaks fall where there

are no sensitive signals. Another way of spreading the power in the peaks is by increasing

N , for a fixed value of N, , i.e., Nc O N, . This way the power in the peaks is spread and

the spurs around the peaks are eliminated. To illustrate this approach, we evaluated the

code PSD for two cases. In the first case, the frame time Tf = 50 ns, the chip time T, = 2

ns, N , = 8 , and Nc = 10 while in the second case, the frame time Tf = 50 ns, the chip

time T, = 2 ns, N , = 8 , and N , = 25. The shape of the spectral peaks in the 0 to 1 GHz

spectrum band is illustrated in Figures 16 and 17, corresponding to N , = 10 and N , = 25,

respectively.

Frequency (Hz) x lo8

Figure 16 TH-PAM code PSD in the spectrum band 0 to 1GHz with N,=10.

40

-45 L 0 1 2 3 4 5 6 7 8 9 10

Frequency (Hz) x lo8

Figure 17 TH-PAM code PSD in the spectrum band 0 to lGHz with Nc=25.

Comparing the results of the TH-PAM code spectrum for the case when Nc = 10

to Nc = 25, we notice that increasing Nc for a fixed value of N, , has the effect of

smoothing the spurs around the peaks at l/c and decreasing the average power contained

in the peaks. Therefore, the effect of an interferer hitting the peak will be lowered but the

main peak at the integer multiples of 1/c can not be eliminated which has also been

reported in 1421. Furthermore, increasing Nc would increase the processing gain in the

spread spectrum sense, and therefore increase the interference rejection of the UWB

system as Nc is increased.

To this end, we evaluate the performance of the TH-PAM system described in

section 2.5.1 for two cases. The first case has Ns = 8 and Nc = 8 while the second case

has Ns = 8 and Nc = 20. The BEP as a function of the tone interferer's frequency was

evaluated using equation (2.23) when the tone interferer's frequency is varied from 1.5

GHz to 6.9 GHz at a 1 MHz resolution. The results for these two cases are shown in

Figures 18 and 19. As evident from the results, the UWB system will be less prone to

interference especially at the critical frequencies which are 1.6 GHz, 3.3 GHz, and 5GHz.

3 3.5 4 4.5 5 5.5 Tone Interferer's Frequency (GHz)

Figure 18 BEP as a function of tone interferer's frequency for the TH-PAM system in the unfaded-faded scenario at fixed SNR=12 dB and SIR=-10 dB. Ns=8, Nc=8.

Tone Interferer's Frequency (GHz)

Figure 19 BEP as a function of tone interferer's frequency for the TH-PAM system in the unfaded-faded scenario at fixed SNR=12 dB and SIR=-10 dB. Ns=8, Nc=25.

The PSD for a TH-PPM system is derived in [34] and given by

sin ( f lCNc f )' sin (flf Ns 12) 2 cos(z6f2 sin (fl'~,f)~ + PwB IWf )I

sin (%f )2 sin (flf f )2 N: sin(flCf )' '

where PwB is the total signal power, W( f ) is the Fourier transform of the monocycle

pulse and the term in the parenthesis is the code spectrum. The PSD of the TH-PPM is

composed of a continuous part and a discrete part involving a Diract delta pulse

sequence. We evaluated the PSD of a TH-PPM UWB system with the frame length

Tf =50 ns, the chip time T, = 2 ns, S=0.15 ns, N, = 8 , N, =25, and T, =0.227ns.

The TH code was generated using a pseudorandom number generator and the peak value

is normalized to 0 dBm. The PSD of the TH-PPM UWB system is shown in Figure 20.

1 2 3 4 5 6 7 8 Frequency (Hz)

Figure 20 PSD of UWB TH-PPM system with Tw=0.227 ns, Tc=2 ns, T ~ 5 0 ns, Nc=25, 6=0.15 ns, and Ns=8.

Similar to the TH-PAM spectral shape, the TH-PPM PSD is determined by the

Fourier transform of the monocycle pulse and the code spectrum. The spectral spikes for

the TH-PPM UWB system occur at integer multiples of 1/T as evident in Figure 20. The

spectral spikes are due to the TH-PPM code spectral shape. To better examine the TH

code, we evaluated the code spectral shape for the TH-PPM modulation format in Figure

21.

-55 I I I I I I I I I I 0 1 2 3 4 5 6 7 8 9 10

Frequency (Hz) x 10'

Figure 21 PSD for the code portion of the UWB TH-PPM system with Tw=0.227 ns, Tc=2 ns, T ~ 5 0 ns, Nc=25,6=0.15 ns, and Ns=8.

As shown in Figure 21, the PSD of the TH-PPM code exhibits discrete spectral

lines at multiples of 1/T, in addition to the continuous part of the code. The discrete PSD

components have a cosine envelope as also shown in equation 2.29. The discrete spectral

lines of the TH-PPM UWB system are undesirable from the coexistence point of view

since the spikes generate interference to existing narrowband systems. Moreover, the

existence of spectral lines makes the TH-PPM modulation format more susceptible to the

interference coming from narrowband systems.

CHAPTER 3 EXPERIMENTAL ANALYSIS ON THE

PERFORMANCE OF TH-PAM UWB RADIO IN COEXISTENCE WITH NARROWBAND

INTERFERENCE

3.1 Introduction

In this chapter, our goal is to present the performance analysis of UWB

communication systems based on experimentation. In particular, we analyze the

performance of a TH-PAM UWB radio in different operating scenarios with the goal of

examining the issue of the coexistence. The organization of this chapter is as follows:

Initially, the key concepts regarding the operation of the UWB radio is described

followed by the experiment setup and measurement results for each scenario.

3.2 UWB Radio Operation

3.2.1 System Overview

The PulseOn 210TM (P210) UWB radio is made by the Time ~ o m a i n @

Corporation. In this thesis, the term UWB radio is used instead of PulseOn 210TM UWB

radio. The UWB radio's intentional emissions meet the FCC part 15 mask for indoor

ultra wideband devices. The main specifications of the UWB radios are presented in

Table 1.

Table 1 UWB radio's specifications.

I Bandwidth (1 0 dB radiated) 1 3.2 GHz I

Pulse Repetition Frequency (PRF)

Center frequency

I Power Consumption 1 6.5 W I

9.6 MHz

approximately 4.7 GHz

The basic element of the UWB radio is the Gaussian monocycle pulse as

described in Chapter 1. As shown in Table 1, the UWB radio has a pulse repetition

frequency (PW) of 9.6 MHz. Consequently, the number of pulses making up a bit is

governed by the data rate of the radio. The top-level schematics for the transmitter and

receiver of the UWB radio are presented in Figures 22 and 23, respectively.

Pulse Generator

L

Figure 22 PulseOnTM UWB radio's Transmitter top level schematic [37]

Data In

Clock Oscillator

Code Generator

Modulation +

J

t Programmable Time

Delay

The pulse generator generates the monocycle Gaussian waveform and the

waveform is then provided to the transmitting antenna. The pulse transmitting time is

controlled by a programmable time delay (PTD), which uses the signal coming from the

clock oscillator to create a timing signal for the pulse generator. The programmable time

delay gets its control signal from the modulator and code generator. The modulator

modulates the incoming data depending on the selected modulation scheme, and the code

generator gives an individual PN code for the modulated data. As shown, the transmitter

does not contain a power amplifier. Instead, the transmitted pulse is generated by the

pulse generator at a specified power level.

Baseband Signal Pulse Generator

Acquisition & Processing

Correlator

Control I

Programmable Time Delay

S/H Multiplier

Clock Oscillator

C

Integrator

-I Code Generator

Data Out

Figure 23 PulseOnTM UWB radio's receiver top level schematic [37].

The receiver is based on the correlation technique. The received signal is

multiplied with a template waveform generated in the receiver, and the result is then

integrated over several received periods of the received pulse train. The correlator

converts the received signal directly into a baseband signal, which is further processed

using a signal processor. The signal processor also provides acquisition and tracking

control for the programmable time delay block [18]. The detected bits are DC voltages

which are integrated depending on the number of pulses making up a symbol.

3.2.2 UWB Radio Synchronization and Modulation

The UWB radio is packet-based which transmits or receives data in bursts. Each

packet consists of an acquisition preamble and a payload as illustrated in Figure 24.

Acquisition Preamble

Payload

Figure 24 An UWB radio packet consists of an Acquisition preamble and a Payload. The Acquisition preamble allows the receiver to synchronize to the transmitter.

The radio can only transmit or receive at any moment in time. The receiver loses

synchronization between packets therefore each packet needs to have a synchronization

preamble. The acquisition preamble allows the receiver to be synchronized to the

transmitter. The payload portion of the packet carries the user specified data.

The UWB radio is based on TH architecture and uses a PN code to determine the

position andlor polarity of each pulse. Varying the position and polarity of each pulse

spreads the spectral features, resulting in lower spectral peaks as discussed in Chapter 1.

The acquisition preamble and the payload use different PN codes, where acquisition

codes specify the pseudorandom time position and polarity for each pulse while payload

codes only specify the time position of each pulse.

The synchronization process in the UWB radio is as follows: A receiver varies the

time base of the sampling window across a frame. For each time shift, the receiver

integrates the sampled energy and compares this value to a user defined synchronization

threshold value. The receiver's sample window is stepped in time when the integrated

energy does not satisfy the threshold. Synchronization is achieved when the threshold

value is satisfied. The concept of synchronization is illustrated in Figures 25 and 26. In

Figure 25, synchronization is not achieved and the receiver's sample window needs to be

shifted in time since the integrated energy does not satisfy the threshold value. However,

synchronization is achieved in Figure 26 since the integrated energy has satisfied the

threshold value.

Transmitted Pulse Position

Receiver Sampling Position

Integrated Energy

Time

Figure 25 UWB radio's synchronization mechanism. The receiver steps its receive time window until it reaches a synchronization threshold. In this illustration, the receiver needs to further step in time since the threshold value is not satisfied.

Transmitted Pulse Position ) I ! ! . . . . . . . . . ! . . . . I ! ! . .

. . . . . . . . . . . . . . . . Receiver Sampling

Position

Integrated Energy

Time

Figure 26 UWB radio's synchronization mechanism. In this illustration, the receiver has satisfied the synchronization threshold and the transmitter and the receiver are synchronized.

The synchronization mechanism implementation in the UWB radio is very

simple, which from the perspective of synchronization could cause problems at large

transmitter-receiver separation andlor in close proximity of interferers. Due to the large

bandwidth of UWB systems, a dense channel multipath profile is observed where many

components can be distinguished from the received signal. The multipath channel then

introduces more than one correct synchronization cell. The energy of the signal is spread

over many multipath components and the energy of each path is very low. Therefore, the

paths are difficult to acquire and the synchronization threshold becomes very significant.

Consequently, the performance of the UWB radio is dependent on the threshold value set

by the user. There is no automatic adjustment of the threshold value implemented for the

UWB radio. A better solution would be to implement a similar synchronization scheme

as in CDMA systems [38] in which, an uncertainty regain is defined which is divided into

a number of cells. Once the algorithm finds one of the possible synchronization cells, an

additional sweep is performed for finding the strongest multipath component.

The UWB radio uses PAM for modulating the information bits. It is important to

note that neither the pulse amplitude nor pulse shape of the transmitted pulse varies to

modulate data. The transmitted power remains constant throughout operation.

3.3 Narrowband Tone Interferer

For the purpose of generating narrowband tone interference, a discone antenna

was built. The discone (disk-cone) antenna is in the family of monopole antennas but has

a much broader frequency band than the ordinary quarter-wavelength monopoles. They

are a combination of the two basic antennas namely, the monopole/dipole antenna and the

biconical antenna. The discone antennas find wide application in the VHF (30-300 MHz)

and UHF (300 MHz - 3 GHz) spectrum for FM broadcast, television and mobile

communications but can be designed to operate in higher frequencies. The discone

antenna was first designed by Kandoian in 1945 [39]. The schematic of the discone

antenna is shown in Figure 27.

Figure 27 Schematic of the Discone antenna.

The design guidelines for the antenna construction are based on the reference [40]

with typical dimensions for the antenna given by

The design parameters for the discone antenna were selected based on the

operating bandwidth of the UWB radio. The discone antenna was made from a 0.003

thick 40 guage, thin copper sheet that can easily be shaped and cut into a cone . The feed

cable used in the antenna construction is a 50 0 rigid coax cable which is able to

withstand the weight of the antenna and hold it in a fixed position in space. The other end

of the rigid coax cable was terminated with an SMC connector and connected to

Agilent's PSG Vector Signal Generator.

3.4 Performance of TH-PAM UWB Radio without Interference

The first step in our experimentation concerns the performance of the TH-PAM

UWB radio without the presence of intentional interference. This scenario would serve as

a basis for analyzing the effect of interference on the performance of UWB system in

later sections and also serve as a link budget analysis concerning the range of operation

of the UWB radio. Measurements were made in two scenarios at different ranges and

data rates. The first scenario covers the LOS path between two UWB radios at different

distances. The second scenario covers a NLOS environment for the link.

The real BEP of the link is calculated by means of transmitting a 32 Kbit, Bit

Error Pattern which is known both to the UWB transmitter and the receiver. The receiver

compares the received data sequence with its local Bit Error Pattern in order to calculate

the BEP of the link. This concept is shown in Figure 28.

Bit Error Transmitter Pattern

Bit Error x Figure 28 Bit error probability (BEP) measurement configuration. A bit error pattern

known both to the transmitter and receiver is used for calculating the BEP.

3.4.1 UWB System in a LOS Environment

In this scenario, a direct path is present between the UWB radio transmitter and

receiver. Figure 29 shows the block diagram for the measurement setup of this scenario.

UWB Transmitter

UWB Receiver

Figure 29 Experiment setup for performance measurement of UWB link in a LOS environment.

The UWB radios were each placed on a table 135 cm from the floor with their

antenna's boresight facing each other. It was made sure that no object obstructed the path

between the two UWB radios. The BEP of the link was measured for different

separations between the UWB transmitter and receiver. In the following sections, we

refer to the distance between the UWB transmitter and receiver as the transmitter-receiver

(T-R) separation. The link performance in the LOS scenario as a function of UWB T-R

separation is shown in Figure 30 for two different data rates. The parameter Ns

represents the number of pulses composing a single bit transmission. The values N, = 8

and Ns = 16 correspond to data rates 1.2 Mbps and 600 Kbps, respectively.

I V

4 4.5 5 5.5 6 6.5 7 7.5 8 Distance (m)

Figure 30 Performance measurement as a function of UWB transmitter-receiver separation in a LOS scenario for Ns=8 and Ns=16.

As expected, the link BEP degrades as the UWB T-R separation is increased. This

is due to path loss the UWB signal experiences, which directly affects the SNR and the

BEP. Decreasing the data rate from 1.2 Mbps to 600 Kbps has the effect of improving the

performance since the number of pulses making a bit is doubled and as a result, the SNR

increases.

3.4.2 UWB System in a NLOS Environment

In this scenario, a direct path does not exist between the UWB radio transmitter

and receiver. Figure 31 shows the block diagram for the measurement setup of this

scenario.

UWB Transmitter

Figure 31 Experiment setup for performance measurement of UWB link in a NLOS environment.

The UWB radios were each placed on a table 135 cm from the floor with their

antenna's boresight facing each other. A metal cabinet obstructed the LOS between the

UWB radios. The BEP of the link was measured for different UWB T-R separations. The

link performance in the NLOS scenario as a function of UWB T-R separation is shown in

Figure 32 for two different data rates. The parameter N, represents the number of pulses

composing a single bit transmission. The values N, = 8 and N, = 16 correspond to data

rates 1.2Mbps and 600 Kbps, respectively.

5.5 6 6.5 Distance (m)

Figure 32 Performance measurement as a function of UWB transmitter-receiver separation in a NLOS scenario for Ns=8 and Ns=16.

As expected, the link BEP degrades as the UWB T-R separation is increased. This

is due to path loss the UWB signal experiences, which directly affects the SNR and the

BEP. Also, note that decreasing the data rate from 1.2 Mbps to 600 Kbps has the effect of

improving the performance since the number of pulses making a bit is doubled and as a

result, the SNR increases. Therefore, there is a trade-off between the data rate of the

UWB system and the coverage range and BEP namely; when the data rate is increased,

the BEP degrades and the coverage is reduced. The acceptable BEP for a communication

link is application dependent. Therefore, the design of the UWB system is dependent on

the application and the operating environment. Furthermore, comparing the results of the

NLOS scenario to the LOS scenario, a severe degrade in BEP is observed which is

caused by the fading that the UWB signals undergo in a NLOS environment.

3.5 Performance of TH-PAM UWB Radio in the Presence of Narrowband Interference

The goal of this experiment was to compare the analytical results with

experimental measurements made for the performance of TH-PAM UWB radio in the

presence of a tone interferer. The first scenario covers a LOS path between the two UWB

radios in the presence of a tone interferer. This scenario corresponds to the unfaded-faded

scenario described in Chapter 2. The second scenario covers a NLOS path between the

two UWB radios in the presence of a tone interferer, which corresponds to the faded-

faded scenario described in Chapter 2.

3.5.1 LOS Scenario

The UWB radios were each placed on a table 135 cm from the floor with their

antenna's boresight facing each other. The transmitter-receiver separation was 5 meters

with a received SNR value of 12dB at the transmitter. The interference source had a

NLOS to the UWB receiver. Figure 33 shows the block diagram of the measurement

setup for this scenario.

UWB Transmitter

UWB Receiver

Tone

Figure 33 Experiment setup for the unfaded-faded scenario. The UWB link has clear LOS while the tone interferer undergoes Rayleigh fading.

Figure 34 shows the link performance in the unfaded-faded scenario at different

SIR values. The TH-PAM UWB radio had the following configurations: frame length

Tf = 104.1 ns, N, = 16, T, = 0.2 ns, and T, = 0.227 ns. Also the antipodal signaling

results in a correlation value of p = -1. The tone interferer had frequency f, = 4.5 GHz.

lo i I I I - I I I

I i - Analytical Performance I I I I I : o Measured Data

SIR [dB)

Figure 34 Performance of TH-PAM as a function of SIR with SNR=12dB. The tone interferer has frequency 4.5 GHz. The UWB link has clear LOS while the tone interferer undergoes Rayleigh fading.

Both the analytical and measured data are shown in Figure 34. As can be seen

from the measurement results, the performance degrades as the SIR is reduced. The

performance mismatch between the analytical and experimental results is due to the fact

that true LOS conditions are seldom possible and the UWB signals experience fading.

Another source of mismatch between the two results is because we considered perfect

synchronization in the analytical analysis while the UWB radios suffer from

synchronization errors especially at very low values of SIR were the UWB receiver could

not synchronize to the transmitter. The results indicate that a TH-PAM UWB experiences

severe performance loss especially when operated in the presence of a narrowband

interferer with high transmit power.

3.5.2 NLOS Scenario

The UWB radios were each placed on a table 135 cm from the floor with their

antenna's boresight facing each other. The transmitter-receiver separation was 4 meters

with received SNR value of 15dB at the receiver. Figure 35 shows the block diagram of

the measurement setup for this scenario.

UWB Transmitter

Tone

UWB Receiver

Figure 35 Experiment setup for the faded-faded scenario. In this case neither the UWB transmitter nor the tone interferer has a LOS to the UWB receiver.

Figure 36 shows the link performance in the faded-faded scenario at different SIR

values. The TH-PAM UWB radio had the following configurations: frame length

Tf = 104.1 ns, N, = 16, T, = 0.2 ns, and T, = 0.227 ns. Also the antipodal signaling

results in a correlation value of p = -1. The tone interferer had frequency f, = 4.5 GHz.

A value of m=4 was used for the fading parameter of the Nakagami-m distribution.

Figure 36 Performance of TH-PAM as a function of SIR with SNR=lSdB. The tone interferer has frequency 4.5 GHz.

Both the analytical and measured data are shown in Figure 36. As can be seen

from the measurement results, the performance degrades as the SIR is reduced.

Comparing the results to the unfaded-faded scenario, we see that the performance is

severely degraded especially when the interferer power is strong and the UWB receiver

experiences detection errors. As in the unfaded-faded scenario, one source of mismatch

between the analytical and experimental results is because we considered perfect

synchronization in the analytical analysis while the UWB radios suffer from

synchronization errors especially at very low values of SIR where the UWB receiver

could not synchronize to the transmitter. The results indicate that a TH-PAM UWB

experiences severe performance loss especially when operated in a NLOS link and in the

vicinity of a narrowband interferer with high transmit power.

3.6 Effect of the Tone Interferer's Frequency on TH-PAM UWB Performance

In this section, our goal is to address the coexistence issue by investigating the

effect of the tone interferer's operating frequency on the BEP performance of the TH-

PAM UWB radio. In section 2.5 of Chapter 2, we concluded that the TH-PAM UWB

system suffers BEP performance around the critical frequencies which are at integer

multiples of the chip time, i.e., k/T, Hz. Referring to Table 1, the operating bandwidth of

the UWB radio is in the range 3.1 GHz to 6.3 GHz. Furthermore, The UWB radio utilizes

a TH code which is 51 1 in length with chip time T, = 0.2 ns. Consequently, the only

critical frequency in the bandwidth of the UWB radio is around 5 GHz.

In order to investigate the effect of the tone interferer's operating frequency on

the BEP performance of the TH-PAM UWB radio, a set of measurements were

performed by sweeping the tone interferer's frequency over the bandwidth 4.8 GHz to 5.2

GHz and measuring the BEP of the UWB radio. The transmitter-receiver separation was

4 meters with the experiment setup as in the NLOS scenario described in section 3.5.2.

Tone Interferer's Frequency [GHz)

Figure 37 BEP as a function of tone interferer's frequency for the TH-PAM system in the faded-faded scenario at fixed SNR=12 dB and SIR=-10 dB. Nc=511, Ns=16.

As evident in Figure 37, no significant performance loss is observed at or around

the critical frequency which is at 5 GHz. The interference rejection of the TH-PAM radio

is due to the smooth spectral shape that the TH-PAM format offers and the fact that the

radio utilizes a TH code which is 51 1 in length. As discussed in section 2.6, the spectral

peaks are eliminated when N, O N, which is the case in this scenario.

CHAPTER 4 CONCLUSIONS

This thesis investigated the issue of coexistence of UWB with narrowband

systems operating in the same bandwidth. In particular, the performance of UWB systems

in the presence of narrowband interference was investigated both through analytical

analysis and through experimentation. The basic aspects of UWB communications were

presented in Chapter 1. Of particular interest are the results on analytical evaluation of

matched filter reception in the presence of narrowband interference presented in Chapter

2. Moreover, the closed-form BEP evaluation of UWB reception of Chapter 2 constitutes

a valid reference for performance evaluation of more complicated UWB signal formats.

In Chapter 2, we presented the results for the performance of TH-PAM UWB

system in different scenarios following with a comparison of TH-PPM and TH-PAM

modulation formats in view of the coexistence issue. In particular, we investigated the

effect of the tone interferer's frequency on the BEP performance of the TH-PAM UWB

system. It was shown that the TH-UWB system will experience performance loss when

the tone interferer hits the spectral components of the TH-UWB signal at critical

frequencies which are at integer multiples of the chip time, i.e., 1/q Hz. One method for

improving the performance of TH-UWB system is by setting N , O N, which causes the

spurs around the critical frequency to flatten and improve the BEP performance.

Furthermore, it was shown that the TH-PAM exhibits a smoother spectrum shape

compared to the TH-PPM which has discrete spectral components. The discrete spectral

components in the TH-PPM make the UWB system susceptible to narrowband

interference and also cause interference to the narrowband systems operating in the same

spectrum bandwidth.

The experimentation consisted of measuring the BEP performance of the TH-

PAM radio in different operating scenarios. The architecture of the UWB radio is based

on a single correlator receiver, which is envisioned as a potential solution for low-cost

transmission systems such as wireless sensor networks where energy and space

constraints necessitate the use of simple receivers. We concluded that the TH-PAM radio

will suffer performance when operated near a narrowband interference with high transmit

power. Furthermore, the coexistence of the TH-PAM UWB radio was investigated by

varying the frequency of the tone interferer. The measured BEP performance of the TH-

PAM UWB system in this case, confirmed the fact that the performance loss around the

critical frequencies could be avoided by setting N, O N , .

The novelty of this research is that most of the available research publications in

the field of coexistence are based on simulation and real world experiments provide

insight for better design of UWB systems. A substantial amount of time and effort was

put into planning and setting up the test and measurement apparatus for UWB

experimentation. I hope my work in this respect would be beneficial to future research

projects.

REFERENCES

1. R.A. Scholtz, "Multiple Access with Time-Hopping Impulse Modulation", Proc. MZLCOM 1993, pp. 447450, Bedford, MA, October 11-13, 1993.

2. M. Z. Win and R. A. Scholtz, "Impulse radio: How it works," ZEEE Commun. Lett., vol. 2, no. 2, pp. 36-38, Feb. 1998.

3. - , "Ultra -wide bandwidth time -hopping spread-spectrum impulse radio for wireless multiple -access communications," IEEE Trans. Commun., vol. 48, no. 4, pp. 679-691, Apr. 2000.

4. - , "Characterization of ultra -wide bandwidth wireless indoor communications channel: A communication theoretic view," IEEE J. Select. Areas Commun., vol. 20, no. 9, pp. 1613-1627, Dec. 2002, Invited Tutorial.

5. J. D. Choi and W. E. Stark, "Performance analysis of ultra-wideband spread-spectrum communications in narrowband interference," in Proc. Military Comm. Cant, vol. 21, no. 1, Anaheim, CALIFORNIA, Oct. 2002, pp. 1075-1080.

6. J. R. Foerster, "The performance of a Direct Sequence spread ultra-wideband system in the presence of multipath, narrowband interference and multi-user interference," in Proc. of IEEE Conference on Ultra Wideband Systems and Technologies (UWBST), Baltimore, MD, May 2002, pp. 87-91.

7. D. Porcino and W. Hirt, "Ultra-wideband radio technology: Potential and challenges ahead," IEEE Communications Magazine, pp. 66-74,2003.

8. S. Gecizi, Z. Tian, G. B. Giannakis, Z. Sahinoglu, H. Kobayashi, A. F. Molisch, and H. V. Poor, "Localization via ultra-wideband radios,"IEEE Communications Magazine, in press.

9. M. Z. Win and R. A. Scholtz, "On the energy capture of ultra -wide bandwidth signals in dense multipath environments," IEEE Comm.Lett., vol. 2, pp. 245-247, Sept 1998.

10. F. Ramirez-Mireles, "Performance of ultrawideband ssma using time hopping and m- ary ppm," ZEEE J. Selected Areas Comm., vol. 19, pp. 1 186 -1 196, Jun 2001.

11. Batra et al., "Multi-band OFDM physical layer proposal," 2003. Document IEEE 802.15-031267r2.

12. I. Kovacs, G. Pedersen, P. Eggers, and K. Olesen, "Ultra wideband radio propagation in body area network scenarios," in Proc. ZEEEZSSSTA 04,2004.

13. A. Papoulis, S. Unnikrishna Pillai, Probability, Random Variables and Stochstic Processes, 4th ed. New York, NY, 10020: McGraw-Hill, Inc., 2002.

14. A. Giorgetti , M. Chiani and M. Z. Win, "Performance of TH-PPM Systems with Narrowband Interferers," ZEEE Znt. Con$ on Commun., (ICC704), Paris, June 2004.

15. A. Giorgetti, Integration of Wireless and Satellite Networks, Ph.D. thesis, University of Bologna, ITALY, March 2003.

16. D. Cassioli, M. Z. Win, and A. F. Molisch, "The ultra -wide bandwidth indoor channel: from statistical model to simulations," ZEEE J. Select. Areas Commun., vol. 20, no. 6, pp. 1247-1257, Aug. 2002.

17. K. Siwiak , D. McKeown, Ultra-wideband Radio Technology, John Wiley & Sons, Inc., 2004.

18. Oppermann, M. H&nal%nen, J. Iinatti, UWB: Theory and Applications, John Wiley & Sons, Inc., 2004.

19. Taha and K. M. Chugg, "A theoretical study on the effects of interference UWB multiple access impulse radio," in Proc. 36th Asilomar Con$ on Signals, Systems and Computers, vol. 1, California, Nov. 2002, pp. 728-732.

20. L. Piazzo, "UWB EM compatibility and coexistence issues," in Proc. of thefirst Znt. Workshop "Networking with UWB", Rome, ITALY, Dec. 2001.

21. G. Yue, L. Ge, and S. Li, "Performance of ultra wideband impulse radio in the presence of jamming," in Proc. of the Znternational Workshop on Ultra Wideband Systems (ZWUWBS), Oulu, FINLAND, June 2003.

22. G. Proakis, Digital Communications, 4th ed. New York, NY, 10020: McGraw-Hill, Inc., 2001.

23. J. H. Bryant, Heinrich Hertz - The Beginning of Microwaves, New York: IEEE Press, 1988.

24. J. D. Taylor, Zntroduction to ultra wideband radar systems, Boca Raton, FL USA, CRC Press, Inc., 1995

25. M. M. Hamalainen, V. Hovinen, R. Tesi, J. H. J. Iinatti, and M. Latva-aho, "On the UWB system coexistence with GSM900, UMTSIWCDMA, and GPS," IEEE J. Select. Areas Commun., vol. 20, no. 9, pp. 1712-1721, Dec. 2002.

26. M. M. Hamalainen, R. Tesi, and J. Iianatti, "On the UWB system performance studies in AWGN channel with interference in UMTS band," in Proc. of ZEEE Con5 on Ultra Wideband Systems and Technologies (UWBST), May 2002, pp. 321-325.

27. M. M. Hamalainen, R. Tesi, J. Iianatti, and V. Hovinen, "On the performance comparison of different UWB data modulation schemes in AWGN channel in the presence of jamming," in Proc. ZEEE Radio and Wirless Confi, Aug. 2002, pp. 83- 86.

28. US 47 CFR Part 15 Ultra-Wideband Operation FCC Report and Order, 22 April 2002.

29. I. Guvenc and H. Arslan, "On the Modulation Options for UWB Systems", in Proc. ZEEE Milcom, Boston, MA, 2003

30. M. G. Di Benedetto and B. Vojcic, "Ultra Wide Band Wireless Communications: A Tutorial", Journal Of Communication and Networks., vol. 5, no. 4, pp.290-302, Dec. 2003.

31. J. Gil-Pelaez, "Note on the inversion theorem," Biometrika, vol. 38, no. 314, pp. 481- 482, Dec. 195 1.

32. D.A. Cummings, "Aggregate ultra wideband impact on global positioning system receivers", ZEEE RAWCON 2001, Radio and Wireless Conference 2001, pp. 101-104.

33. Assessment of Compatibility between Ultrawideband (UWB) Systems and Global Positioning System (GPS) Receivers, NTIA Special publication 01-45, Feb. 2001

34. L. Piazzo: Some basic facts about UWB EM compatibility and UWB spectrum, Whyless internal report, May 2001.

35. R. Malik, Spread Spectrum, Secrete Military Technology to 3G, (online) www.ieee.ordornanizations1history centerlcht papers/SpreadSpectrum.pdf, 20 Dec, 2004.

36. Padgett., Coexistence of UWB and Legacy Narrowband Systems, www.darpa.mil/ato/pro~ramslnetex/coexistf, 15 Jan, 2005.

37. Time Domain Corporation, PulsOn technology overview, (online) www.timedomain.com, 10 Dec, 2003.

38. J. Iinatti and M. Latva-aho, "A modified CLPDI for code acquisition in multipath channel", in Proc. of Personal, Indoor, and Mobile Radio Communications (PIMRC'OI) San Diego, CA, PP. F6-F10,2001.

39. A.G. Kandoian, "Three new antenna types and their application" in Proc. IRE, vol. 34, pp. 70W-75W, Feb. 1946.

40. J.J. Nail, "Designing discone antennas", Electronics, Vol. 26, pp. 167-169, Aug. 1953.

41. S. Gradshteyn and I. M. Ryzhik, Tables of Integrals, Series, and Products, 4th ed. San Diego, CA: Academic Press, Inc., 1980.

42. C. Miiller, S. Zeisberg, H. Seidel and A. Finger, "Spreading Properties of Time Hopping Codes in Ultra Wideband Systems", ISSTA 2002, Prague, Czech Republic

43. L. Zhao and A.M. Haimovich, "Performance of Ultra-Wideband Communication in the Presence of Interference," IEEE Journal on Selected Areas Special issue on Ultra Wide Band Radio in Multi-Access Wireless Commun. vol: 20, Issue: 9, Dec 2002, pp. 1684-1692.

44. N. Lehmann and A. M. Haimovich, "The Power Spectral Density of a Time Hopping UWB Signal: A Survey," in Proc. IEEE Conference on Ultra Wideband Systems and Technologies (UWBST'03), Reston Virginia, Nov 2003.