DRA T Design of Indoor Positioning System Based on

DRA T  Design of Indoor Positioning System Based on
Design of Indoor Positioning System Based on
IEEE 802.15.4a Ultra-wideband Technology
JINKANG CEN
DRA T
TRITA-ICT-EX-2013:225
Master of Science Thesis
Stockholm, Sweden 2013
Design of Indoor Positioning System Based on
IEEE 802.15.4a Ultra-wideband Technology
JINKANG CEN
Master of Science Thesis performed at
MarnaTech AB, Stockholm, Sweden
June 2013
Examiner: Lirong Zheng
Supervisor: Peter Reigo
i
KTH School of Information and Communications Technology (ICT)
System On Chip Design
C
Jinkang Cen, June 2013
ii
Abstract
Global Positioning System (GPS) has revolutionized the way we navigate and get locationbased information in the last decade. Unfortunately the accuracy of civilian GPS is still
remaining in meter level and it does not work well in indoor environment, which is a major
drawback for applications such as autonomous vehicle, robot machine and so on. UWB
(Ultra-wideband) is one of the most promising technologies to solve this problem. The UWB
technology has large bandwidth and it is quite robust to fading and multipath effect. Therefore,
it is capable of high accuracy down to centimeters for positioning in both outdoor and indoor
scenarios.
The IEEE 802.15.4a was released in 2007, which adopted UWB in this standard and specified
its physical layer for accurate positioning in WPAN (Wireless Personal Area Network). Apart
from the capability of accurate positioning, solutions based on this standard will have quite
low power consumption and low cost.
In this thesis work a positioning system based on IEEE 802.15.4a has been designed. A few
practical constrains have been taken into account in designing the system, such as
performance, cost, power consumption, and governmental regulations and so forth. To reduce
the system complexity and communication channel occupancy, TDOA (Time Difference of
Arrival) has been chosen as the ranging protocol. The system has been designed accordingly.
Main components have been selected and PCBs (Printed Circuit Board) has been designed as
well. The design work covered both hardware and software. The proposed system is believed
to be able to achieve a positioning accuracy of ±20 centimeters.
iii
Acknowledgements
First of all, I would like to thank Peter Reigo of MarnaTech AB for giving me the opportunity
to carry out this thesis work on a cutting edge technology. His knowledge and enthusiasm on
this project have inspired me a lot. I would like to express my appreciation for his supervisor
and support on the thesis work.
I would also like to thank Irfan M. Awan, Binyam S. Heyi, Alessandro Monge, and Yuefan
Chen, who have helped me during this time in MarnaTech AB.
In addition, I want to thank a few friends of mine who have been helping me for the past three
years, no matter materially or spiritually. They are Juelin Wang, Juhua Liao, Chuang Zhang,
Zhihao Zheng, Yanpeng Yang, and Yalin Huang.
Special thanks to my family, including my brother, my sister-in-law and especially my mother.
They are always standing by my side supporting me and encouraging me. The most gratitude
shall be given to them.
Finally, I would like to dedicate this work to my father who left us in 2009. May he rest in
peace and love.
iv
v
Contents
Abstract .................................................................................................................................................. iii
Acknowledgements ................................................................................................................................ iv
Contents .................................................................................................................................................. vi
List of Figures ...................................................................................................................................... viii
List of Tables ........................................................................................................................................... x
Abbreviations ......................................................................................................................................... xi
Chapter 1 ................................................................................................................................................. 1
Introduction ............................................................................................................................................. 1
1.1 Background ................................................................................................................................... 1
1.2 Related Work................................................................................................................................. 3
1.3 Problem Definition ........................................................................................................................ 4
Chapter 2 ................................................................................................................................................. 5
Ultra-wideband Positioning System ........................................................................................................ 5
2.1 Fundamentals of UWB .................................................................................................................. 5
2.1.1 Definition of UWB Signal ...................................................................................................... 5
2.1.2 Impulse Radio UWB Signal ................................................................................................... 5
2.1.3 International Regulations........................................................................................................ 6
2.2 The IEEE 802.15.4a Standard ....................................................................................................... 9
2.2.1 Operating Frequency and Channel Allocations ...................................................................... 9
2.2.2 PHY Specifications .............................................................................................................. 10
2.3 Time-based Ranging Protocols.................................................................................................... 12
2.3.1 Two-way Time-of-Arrival .................................................................................................... 12
2.3.2 Symmetric Double-Sided Two-Way Time-of-Arrival ......................................................... 13
2.3.3 Time Difference of Arrival ................................................................................................... 14
Chapter 3 ............................................................................................................................................... 16
System Design ....................................................................................................................................... 16
3.1 System Requirements .................................................................................................................. 16
3.2 Protocol Selection........................................................................................................................ 17
3.3 System Setup ............................................................................................................................... 18
3.4 Solutions Selection ...................................................................................................................... 20
vi
3.4.1 UWB Transceiver ................................................................................................................. 20
3.4.2 Wireless Link ....................................................................................................................... 21
3.4.3 Microcontroller ..................................................................................................................... 22
3.4.4 DC/DC Regulator ................................................................................................................. 23
3.4.5 Connectivity ......................................................................................................................... 23
3.5 Software Architecture.................................................................................................................. 24
3.5.1 Software Architecture of Tag ............................................................................................... 25
3.5.2 Software Architecture of Reference Node............................................................................ 26
Chapter 4 ............................................................................................................................................... 27
System Implementation ......................................................................................................................... 27
4.1 Hardware Design ......................................................................................................................... 27
4.1.1 Block Diagram ..................................................................................................................... 28
4.1.2 Microstripe Design ............................................................................................................... 31
4.1.3 Grounding............................................................................................................................. 33
4.3 Software Design .......................................................................................................................... 35
4.3.1 Flow Chart ............................................................................................................................ 35
4.3.2 TDOA Algorithm ................................................................................................................. 37
Chapter 5 ............................................................................................................................................... 39
Performance Evaluation ........................................................................................................................ 39
5.1 Functionality Verification ........................................................................................................... 39
5.1.1 DC/DC Regulators ............................................................................................................... 39
5.1.2 Microcontrollers ................................................................................................................... 39
5.1.3 UWB Transceiver ................................................................................................................. 40
5.1.4 Wireless Link ....................................................................................................................... 41
5.1.5 Ethernet ................................................................................................................................ 41
5.2 Performance Evaluation .............................................................................................................. 41
5.2.1 RF Performance .................................................................................................................... 41
5.2.2 Wireless Link Performance .................................................................................................. 43
5.2.3 UWB Transceiver Performance ........................................................................................... 46
Chapter 6 ............................................................................................................................................... 47
Conclusion and Future Work ................................................................................................................ 47
6.1 Summary ..................................................................................................................................... 47
6.2 Conclusion ................................................................................................................................... 47
6.3 Future Work ................................................................................................................................ 48
Reference ............................................................................................................................................... 49
vii
List of Figures
Figure 1.1 Outline of Wireless Positioning Technologies ............................................................... 2
Figure 2.1 A UWB Pulse with the Pulse Width of Around 1ns ...................................................... 6
Figure 2.2 FCC Regulations on EIRP Emission for Indoor UWB Devices .................................... 7
Figure 2.3 FCC Regulations on EIRP Emission for Outdoor UWB Devices ................................. 7
Figure 2.4 EC Regulations on EIRP Emission for UWB Devices without Mitigation Techniques 8
Figure 2.5 EC Regulations on EIRP Emission for UWB Devices with Mitigation Techniques ..... 9
Figure 2.6 PHY Data Flow for the IEEE 802.15.4a Standard ....................................................... 10
Figure 2.7 Frame Format for the IEEE 802.15.4a Standard .......................................................... 11
Figure 2.8 Illustration of Two-way Time-of-Arrival Ranging Protocol........................................ 13
Figure 2.9 Illustration of Symmetric Double-Sided Two-Way Time-of-Arrival Ranging Protocol
............................................................................................................................................... 14
Figure 2.10 Illustration of Time Difference of Arrival Ranging Protocol .................................... 15
Figure 3.1 System Setup of Positioning System Based on the IEEE 802.15.4a............................ 19
Figure 3.2 Architecture of a Tag in the Proposed Positioning System .......................................... 19
Figure 3.3 Architecture of a Reference Node in the Proposed Positioning System ...................... 20
Figure 3.4 Architecture of a Communication Device in OSI Model ............................................. 24
Figure 3.5 Software Architecture of a Tag .................................................................................... 25
Figure 3.6 Software Architecture of a Reference Node ................................................................ 26
Figure 4.1 Stackup of a 6-layer FR4 PCB ..................................................................................... 28
Figure 4.2 Stackup of a 4-layer FR4 PCB ..................................................................................... 29
Figure 4.3 Block Diagram of Base Board ..................................................................................... 29
Figure 4.4 Finished PCB of a Base Board ..................................................................................... 30
Figure 4.5 Block Diagram of Connection Board ........................................................................... 30
viii
Figure 4.6 Finished PCB of a Connection Board .......................................................................... 31
Figure 4.7 Finished PCB of a Reference Node ............................................................................. 31
Figure 4.8 Block Diagram of Tag.................................................................................................. 32
Figure 4.9 Finished PCB of Tag .................................................................................................... 32
Figure 4.10 Layout of High Frequency Signal Tracks in Base Board........................................... 33
Figure 4.11 Structure of a Microstripe .......................................................................................... 33
Figure 4.12 Microstrip Calculation with AppCAD ....................................................................... 34
Figure 4.13 Illustration of the Impact of Transmission Line Effect on Grounding ....................... 35
Figure 4.14 Placements of Ground Vias to Avoid Transmission Line Effect ............................... 35
Figure 4.15 Program Flow Charts in Tag, Base Station and Coordinator ..................................... 37
Figure 5.1 UWB Signals from the UWB Transceiver ................................................................... 41
Figure 5.2 S11 Measurements for the RF Part of Base Board ...................................................... 43
Figure 5.3 S11 Measurements for the RF Part of Tag ................................................................... 43
Figure 5.4 Line-of-Sight Tests on the Performance of Wireless Link .......................................... 45
Figure 5.5 Non-Line-of-Sight Tests on the Performance of Wireless Link .................................. 46
ix
List of Tables
Table 2.1 The IEEE 802.15.4a Characteristics ................................................................................ 9
Table 2.2 UWB Channel Allocations for the IEEE 802.15.4a ...................................................... 10
Table 3.1 Typical Errors in Time-of-Flight Estimation Using TW-TOA ..................................... 17
Table 3.3. Unlicensed Frequency Bands for Non-specific Short Range Devices in Sweden........ 21
Table 3.4 Requirements for Microcontrollers ............................................................................... 23
Table 5.1 Packet lost rate for Line-of-Sight .................................................................................. 45
Table 5.2 Packet lost rate for Non-Line-of-Sight .......................................................................... 46
x
Abbreviations
2D
3D
AC
AGC
A-GPS
AOA
API
APP
APS
AWGN
BPM
BPSK
CSS
DAA
DC
EC
EIRP
FCC
GPS
IC
IEEE
IR
LDC
LINK
MAC
NET
OSI
PCB
PER
PHR
PHY
PHY
PPM
PSDU
QFN
RDEV
RFRAME
RMII
RSS
SDSTW-TOA
SFD
SHR
SPI
SRD
Two-Dimensional
Three-Dimensional
Alternating Current
Automatic Gain Control
Assited Global Positioning System
Angle of Arrival
Application Programming Interface
Application Layer
Application Support Layer
Additive White Gaussian Noise
Burst Phase Modulation
Binary Phase Shift Keying
Chirp Spread Spectrum
Detect and Avoid
Direct Current
European Commission
Equivalent Isotropically Radiated Power
Federal Communications Commission
Global Positioning System
Integrated Circuit
The Institute of Electrical and Electronics Engineers
Impulse Radio
Low Duty Cycle
Link Layer
Media Access Layer
Network Layer
Open Systems Interconnection
Printed Circuit Board
Packet Error Rate
PHY Header
Physical Layer
Physical Layer
Parts Per Million
PHY Service Data Unit
Quad-Flat No-Leads
Ranging Device
Ranging Frames
Reduced Media Independent Interface
Received Signal Strength
Symmetric Double-Sided Two-Way Time of Arrival
Start-of-Frame Delimiter
Synchronization Header
Serial Peripheral Interface
Short Range General Purpose Device
xi
TCP/IP
TCXO
TDOA
TG4a
TOA
TOF
TW-TOA
UART
USA
UWB
WPAN
Transmission Control Protocol and the Internet Protocol
Temperature Compensated Crystal Oscillator
Time Difference of Arrival
Task Group 4a
Time of Arrival
Time of Flight
Two-Way Time of Arrival
Universal Asynchronous Receiver/Transmitter
United States of America
Ultra-wideband
Wireless Personal Area Network
xii
Chapter 1
Introduction
Position acquisition is one of the key features of navigation application. With the advent of Global
Positioning System (GPS) in the 1990s the way we acquire position information has been
revolutionized rapidly, thus enhancing the navigation technique dramatically. As civilian GPS can
only be used in outdoor environment and its accuracy is only in meter level, further development of
positioning system is in higher and higher demand for better accuracy and better performance in
indoor scenario. Ultra-wideband (UWB) is one of the most promising technologies to meet this
demand. The UWB technology has large bandwidth of more than 500 MHz and it is quite robust to
fading and multipath effect. This technology makes centimeter level accuracy for positioning to be
possible and it is one of the most promising alternatives for positioning systems in the near future.
1.1 Background
Since the fully operational in 1993, GPS has been the most important positioning system in the world.
Its usage covers highly varied applications, from terrestrial navigation, to maritime and aeronautic
location and guidance systems, to cellular emergency assistance and varies kinds of lost-and-found
applications.
GPS is a Time-of-Arrival (TOA, which will be discussed in Chapter 2) distance measuring system
which requires at least three satellites (or so called transmitters) in order to calculate latitude, longitude
and height. It operates at 1575.42 MHz, which is referred to as L1, and 1227.6 MHz, which is referred
to as L2 [1]. For both frequency bands, GPS signal is sensitive to multipath effect which means the
signal can easily be blocked by roof, wall, or tree. Therefore, usage of GPS in indoor environment is
not possible. Even though civilian GPS works well for outdoor scenario, its accuracy is still remaining
in meter level. Field measurements have shown that the positioning accuracy of civilian GPS can be as
good as ±8 meters [1], which becomes a bottleneck for many potential localization applications such
as goods and items tracking, precision landing, autonomous vehicle, robot guidance and so on.
To overcome the problems that GPS is facing with, many other solutions have been introduced in
recent years. Wireless Assisted GPS (A-GPS) [2] is a straight-forward solution to extend the usage of
GPS from environments with good satellite signals to those with poor reception of satellite signals,
including indoor environments. A-GPS technology uses a location server with a reference GPS
receiver that can simultaneously detect the same satellites as the wireless handset (or mobile device)
with a partial GPS receiver, to help the partial GPS receiver find weak GPS signals [3]. In this case,
many functions of a full GPS receiver in a wireless handset are performed by the location server.
Preliminary results (such as satellite orbit, clock information, initial position, time estimate, and
1
position computation and so on) are transmitted from the server to the wireless handset via cellular
network accordingly. One study [4] has shown that the accuracy of A-GPS can get down to ±5 meters.
Another alternative that has been widely discussed and developed is WLAN (Wireless Local Area
Network). By making usage of the widely existing WLAN infrastructure, positioning system based on
RSS (Receiving Signal Strength) measurement can be easily developed. The accuracy of a typical
WLAN positioning system is approximately 3 to 30 meters [3], depending on the complexity of indoor
environment. The major detrimental factor influencing accuracy is the propagation attenuation caused
by different objects (such as wall, table, roof), which is a common problem with positioning solutions
based on RSS measurement.
Bluetooth is also a popular solution for positioning system as many portable devices support this
technology. Positioning system based on Bluetooth usually employs RSS measurement, the same
method as WLAN. Its accuracy can get down to 2 meters with 95% reliability [3]. The bottleneck is
still the propagation attenuation. Besides, it requires users to install Bluetooth Beacons around the
measuring area and the range of those Beacons are usually quite short (10-15 meters).
Other positioning technologies are also available, such as cellular based network, Zigbee, ultrasonic
and infrared and so on. Figure 1.1 [3] shows a comparison of resolution between different wireless
positioning technologies. Among these technologies, UWB, which is the main focus of this thesis
report, is believed to provide the best performance, down to centimeter accuracy.
Figure 1.1 Outline of Wireless Positioning Technologies
2
1.2 Related Work
The initial concepts and patents for UWB technology, which was alternatively referred to as baseband,
carrier-free or impulse, originated in the late 1960’s in the United States of America (USA) [5]. Early
development of UWB technology focused on impulse radar mainly for the purpose of defense.
Civilian usage didn’t get started until early this century when the Federal Communications
Commission (FCC) in the USA announced its “First Report and Order” in 2002, which approved and
regulated the unlicensed use of devices based on UWB technology [6]. After that, similar actors were
also taken in Europe and Asia to authorize the use of UWB devices under certain restrictions [7-8][19].
After the FCC regulated the use of UWB devices, standardization efforts were taken by the IEEE (The
Institute of Electrical and Electronics Engineers) to employ the UWB technology for low-rate Wireless
Personal Area Networks (WPANs) that focus on low power and low complexity devices. The task
group 4a (TG4a) for an amendment to the IEEE 802.15.4 standard for an alternative physical layer
(PHY) was formed by IEEE in 2004. And the IEEE 802.15.4a standard was first introduced in 2007
[9]. The IEEE 802.15.4a has specified the use of UWB technology for its PHY which provides highprecision ranging/localization capability, high throughput and ultra-low-power consumption. This
standard is studied in Chapter 2.
Since the IEEE 802.15.4a was released, many researches [10-16] have been carried out to implement it.
In [10] a modular architecture of UWB transmitter based on IEEE 802.15.4a was proposed, which
gave an insight on how to implement the multi-channel, multi-band UWB transmitter for high design
flexibility. But no modeling or implementation work was done to prove their concepts in this study.
Another study in [11] developed a high-level MATLAB model of UWB PHY for the IEEE 802.15.4a
standard. With the performance evaluation of the proposed model by adding additive white Gaussian
noise (AWGN), the rationality of the model has been verified.
The first implementation of the IEEE 802.15.4a standard on IC (Integrated Circuit) level was done in
2008 by a group in Singapore [12]. An UWB transceiver capable of both communication and
localization based on this standard was implemented on a 0.18μm CMOS technology. It supports 12
channels from 3 to 9 GHz and variable data rates. The system can achieve 0.2ns resolution which
corresponds to two-way ranging accuracy of 3cm [12]. The power consumption is relatively low as it
achieves 0.74nJ/pulse for transmission and 6.5nJ/pulse for reception.
There was another group in Korea who implemented a transceiver based on the IEEE 802.15.4a
standard on IC level in 2009 [13]. The transceiver was manufactured on a 0.13μm CMOS technology.
It supports three channels at 3494.4 MHz, 3993.6 MHz and 4492.8 MHz with bandwidth of 499.2
MHz. It is capable of data rates up to 850 kbps with a communication range of 20 meters. A low level
ranging protocol/architecture was also implemented with a ranging accuracy of ±30 centimeters in
the multipath indoor shadowing environment. However, the power consumption of the chip was not
reported. Part from their work on the transceiver, a packet-based ranging system was also built by the
same group in [14] and it had achieved a ranging accuracy of ±30 centimeters as well.
3
1.3 Problem Definition
So far, several prototyping systems have been implemented for the IEEE 802.15.4a standard, as well
as transceivers based on CMOS technology. But many other factors and constrains have not been
considered in those implementations, such as system complexity, synchronization, power consumption,
cost and regulations. Besides, they were much more focusing on ranging, rather than positioning.
The purpose of this thesis work is thus to design a positioning system based on IEEE 802.15.4a. The
system shall be able to calculate the position of a targeting object. The design work covers both
hardware and software, shall be conducted in an approach considering main issues and constrains for a
consumer product. Main components shall also be selected for commercial usage.
The rest of the report is organized as follows: Chapter 2 describes the fundamentals of the UWB
technology and the IEEE 802.15.4a standard. Chapter 3 shows the methodology and the approach the
design work is following. Chapter 4 focuses on the implementation of our positioning system. Both
hardware design and software implementation will be discussed in Chapter 4. The performance of the
system will be evaluated in Chapter 5. Chapter 6 will present our conclusion and future improvements
on our positioning system.
4
Chapter 2
Ultra-wideband Positioning System
In this chapter, fundamentals of the UWB technology will be discussed. The IEEE 802.15.4a standard,
as well as the regulations in different regions/countries, will be introduced. We will also discuss some
key ranging protocols based on time-of-flight (TOF) measurement.
2.1 Fundamentals of UWB
2.1.1 Definition of UWB Signal
An UWB signal is defined to be a signal with a fractional bandwidth of larger than 20% or an absolute
bandwidth of at least 500 MHz, regardless of the fractional bandwidth [17]. Designating the upper
frequency of the -10 dB emission point as and the lower frequency of the -10dB emission point as
. The absolute bandwidth B is calculated as the difference between them; i.e.
.
On the other hand, the fractional bandwidth
(2.1)
equals to
,
where
(2.2)
is the central frequency of the signal and it is given by
.
So the fractional bandwidth
(2.3)
in (2.2) can be expressed as
.
(2.4)
According to the definition by FCC [6], UWB systems with a central frequency larger than 2.5 GHz
must have a bandwidth of at least 500 MHz while UWB systems working at a central frequency
smaller than 2.5 GHz shall have a fractional bandwidth larger than 20%.
2.1.2 Impulse Radio UWB Signal
Impulse radio (IR) is a type of UWB system that transmits UWB pulses with a low duty cycle [18]. It
is a common way to achieve wide bandwidth, thus generating UWB signals. Figure 2.1 shows an
example of a UWB pulse and the second derivative of a UWB pulse is expressed as
5
,
(2.5)
where A>0 and are parameters that determine the energy and the width of the pulse, respectively [17].
Figure 2.1 A UWB Pulse with the Pulse Width of Around 1ns
There are many distinct advantages of using UWB pulses in a communication system for positioning
applications. First, a UWB signal is capable of penetrating through obstacles such as wall, wood, and
ceiling and so on. Besides, it is robust to multipath effect as a positioning system based on UWB
technology measures the first pulse it receives. All these features make UWB systems quite suitable
for indoor usage. Secondly, large bandwidth results in high time resolution, so it improves the
accuracy of ranging and positioning. Thirdly, a large bandwidth also allows a really high data rate. So
a UWB system is beneficial for high speed data communication. Fourthly, power consumption of a
UWB system can be really low to increase the battery life because the power is transmitted in a large
bandwidth. A low power density also minimizes the interference to other systems operating in the
same frequency band. Finally, since a UWB system can operate in the baseband, the hardware can be
simplified which makes low cost implementation to be possible.
2.1.3 International Regulations
As we have discussed before, many countries have specified their own regulations on the use of UWB
devices as UWB devices occupy a very large portion in the spectrum and they shall not cause
significant interference to other systems. UWB devices shall coexist with other systems operating
inside and outside the same frequency band.
According to the FCC regulations in the USA, maximum Equivalent Isotropically Radiated Power
(EIRP) in any direction shall not exceed the Part 15 limit of -41.3 dBm/MHz [6]. Additional, much
stricter rules have also been specified regarding various UWB systems depending on the specific
application area. For the communications and measurement systems, which are much more into our
concerns, the FCC has set slightly different limit (which is usually called spectrum mask) for indoor
6
and outdoor usage (as shown in Figure 2.2 and Figure 2.3 [17], respectively). Specifically, a 10 dB
reduction on the EIRP emission level of outdoor UWB systems in the frequency band between 1.610
GHz and 3.100 GHz shall be applied compared to that of indoor UWB systems.
Figure 2.2 FCC Regulations on EIRP Emission for Indoor UWB Devices
Figure 2.3 FCC Regulations on EIRP Emission for Outdoor UWB Devices
For indoor UWB systems, they are not allowed to be used outdoor, or to direct their radiation outside.
And only peer-to-peer communication is allowed.
7
As for outdoor UWB systems, they shall not operate on a fixed infrastructure and they shall only
communicate with their associated receivers.
In Europe, the European Commission (EC) has also regulated the use of UWB devices from 2007 [7]
[19] and the regulations are valid in all the member countries including Sweden. The spectrum mask
decided by EC is shown in Figure 2.4 for UWB devices that do not apply additional appropriate
mitigation techniques. Specifically, such UWB devices can transmit UWB signals at most -41.3
dBm/MHz from 6.0 GHz to 8.5 GHz. This value also applies for the 4.2 GHz - 4.8 GHz band until the
end of 2010. After that, the limit of EIRP has been changed to be -70 dBm/MHz for this band.
Figure 2.4 EC Regulations on EIRP Emission for UWB Devices without Mitigation Techniques
As for UWB devices that apply appropriate mitigation techniques, the EC regulation is shown in
Figure 2.5. Specifically, a maximum mean EIRP density of -41.3 dBm/MHz is allowed in the 3.1 GHz
– 4.8 GHz band when low duty cycle (LDC) mitigation is employed. The LDC mitigation shall fulfill
that the sum of all transmitted signals is less than 5% of the time each second and less than 0.5% of the
time each hour and each transmitted signal does not exceed 5 ms. On the other hand, when detect and
avoid (DAA) mitigation technique as described in Directive 1999/5/EC [20] is employed, a maximum
mean EIRP density of -41.3 dBm/MHz is allowed in the 3.1 GHz – 4.8 GHz and 8.5 GHz – 9.0 GHz
bands. Besides, limits for usage of UWB device in automotive and railway vehicles, as well as in
building material analysis, are also defined in [19]. Note that UWB signal is not allowed to be
transmitted from a device at a fixed installation or connected to a fixed outdoor antenna or in vehicles;
this also applies for regulations in Sweden [21].
8
Figure 2.5 EC Regulations on EIRP Emission for UWB Devices with Mitigation Techniques
2.2 The IEEE 802.15.4a Standard
The IEEE 802.15.4a Standard was first approved in 2007 by TG4a. It is the first international standard
that specifies a wireless PHY for precision ranging in low rate PWANs. Apart from ranging capability,
it also supports high data rate communication, extended range, low power operation, and improved
robustness against interference and high-speed motion.
2.2.1 Operating Frequency and Channel Allocations
The IEEE 802.15.4a has specified two alternate PHYs; one is based on IR UWB with the capability of
ranging while the other is based on chirp spread spectrum (CSS) which can only be used for
communication purpose. The UWB PHY can use frequency bands including 250 MHz – 750 MHz,
3244 MHz – 4742 MHz and 5944 MHz – 10234 MHz, while the CSS PHY can only use 2400 MHz –
2483.5 MHz band. There are 16 channels for the UWB PHY and 14 channels for the CSS one.
Operating frequency and channel allocations, as well as other related information are shown in Table
2.1. The 16 UWB channels are listed in Table 2.2.
PHY Option
Frequency Bands
NO. of Channels
Data Rate
Ranging Support
Range
Protocol
Table 2.1 The IEEE 802.15.4a Characteristics
UWB PHY
CSS PHY
250 MHz – 750 MHz (Sub-GHz)
3244 MHz – 4742 MHz (Low-Band)
2400 MHz – 2483.5 MHz
5944 MHz – 10234 MHz (High-Band)
16
14
110 kbps, 851 kbps (mandatory),
250 kbps, 1 Mbps (mandatory)
6.81 Mbps, 27.24 Mbps
Yes
No
10-100 meters
ALOHA, CSMA-CA
9
Channel NO.
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Table 2.2 UWB Channel Allocations for the IEEE 802.15.4a
Central Frequency (MHz) Bandwidth (MHz) UWB Band / Mandatory
499.2
499.2
Sub-GHz Mandatory
3494.4
499.2
Low Band
3993.6
499.2
Low Band
4492.8
499.2
Low Band Mandatory
3993.6
1331.2
Low Band
6489.6
499.2
High Band
6988.8
499.2
High Band
6489.6
1081.6
High Band
7488.0
499.2
High Band
7987.2
499.2
High Band Mandatory
8486.4
499.2
High Band
7987.2
1331.2
High Band
8985.6
499.2
High Band
9484.8
499.2
High Band
9984.0
499.2
High Band
9484.8
1354.97
High Band
2.2.2 PHY Specifications
The UWB PHY waveform is based on an IR signaling scheme using band-limited data pulses [9].
Figure 2.6 [9] shows the modular sequence of processing steps used to modulate and transmit a UWB
PHY packet; the procedure for receiving and demodulating the PHY packet is shown in the same
figure as well.
Figure 2.6 PHY Data Flow for the IEEE 802.15.4a Standard
Figure 2.7 [9] illustrates the format for a UWB frame composing three major components, the SHR
(Synchronization Header) preamble, the PHR (PHY Header) and the PSDU (PHY Service Data Unit)
with the SHR transmitted or received first in a UWB system, followed by the PHR and finally the
PSDU.
10
The SHR preamble consists of a SYNC preamble field and a start-of-frame delimiter (SFD). The
SYNC field, which is composed with specific preamble codes defined in the standard, is used for
automatic gain control (AGC) convergence, diversity selection, timing acquisition, and coarse
frequency acquisition. This field is important to the UWB receiver because it makes the receiver to
lock on the frame and configure the receiver for the incoming message. The SFD indicates the end of
the preamble and the beginning of the PHY header and it is used to establish the timing of a frame,
thus its detection is critical for accurate ranging counting.
The PHR is composed of the decoding information of the packet to the receiver. Decoding information
includes the data rate used to transmit the PSDU, the duration of the current frame’s preamble, and the
length of the frame payload. Six parity check bits are also encapsulated in the PHR to further protect
the PHR against channel errors.
Finally, the PSDU consists of the data sent to the receiver at the data rate indicated in the PHR. The
length of PSDU varies from 0 to 127 bytes.
Figure 2.7 also shows the encoding process of a UWB frame, corresponding to the PHY data flow
shown in figure 2.6.
Figure 2.7 Frame Format for the IEEE 802.15.4a Standard
11
The modulation scheme used in the standard is called BPM-BPSK (a combination of Burst Phase
Modulation and Binary Phase Shift Keying) to support both coherent and non-coherent receivers. The
combined BPM-BPSK is used to modulate the UWB symbols, with each symbol being composed of
an active burst of UWB pulses and carrying two bits of information. The positioning of the burst in
one symbol can be determined by a burst-hopping sequence, which helps in improving the robustness
of multi-user access interference.
2.3 Time-based Ranging Protocols
The IEEE 802.15.4a standard mainly focuses on the lower layers including PHY and MAC sub-layer.
It is from a technology point of view to employ this standard in a positioning system for better
accuracy. Apart from that, we also need to think in the systematic perspective. Ranging protocol is a
technique in a systematic level that determines how to use the UWB technology in the best way to
achieve the highest positioning accuracy.
In order to obtain the position of an object in a wireless system, the object needs to exchange signals
with a number of reference nodes. The position can be estimated from measuring the signals or certain
parameters extracted from the signals exchanging between the object and all nodes. There are several
signal measuring techniques that are commonly used nowadays, such as Received Signal Strength
(RSS), Angle of Arrival (AOA), Time of Arrival (TOA) and Time Difference of Arrival (TDOA) and
other hybrid solutions and so forth, which have been studied in [17]. The UWB technology is a
technique based on time-of-flight (TOF) measurement. With the help of TOF measurement, ranging
protocols can be developed upon it. The selection of a ranging protocol influences quite a lot on how
the system is designed and implemented. In this section, we will discuss some key time-based ranging
protocols, including two-way Time of Arrival (TW-TOA), symmetric double-sided two-way Time of
Arrival (SDSTW-TOA) and Time-Difference of Arrival (TDOA).
2.3.1 Two-way Time-of-Arrival
In a TW-TOA protocol, ranging is conducted by exchanging ranging frames (RFRAME) between two
ranging devices (RDEV) and marking their arrival times. Figure 2.8 illustrates the procedure of a TWTOA protocol in a two devices (RDEV A and B) scenario. RDEV A tracks the departure time of
RFRAME 1 and the arrival time of RFRAME 2 and
represents the round trip time at RDEV A.
RDEV B also marks the arrival time of RFRAME 1 and the departure time of RFRAME 2 and
donates the processing delay at RDEV B. Let represents the TOF between RDEV A and B and it
can be estimated as
.
If we consider the influence of clock drift and let
the estimated TOF is
and
(2.6)
to be the clock offsets of RDEV A and B,
.
The range
(2.7)
is calculated as
,
where c is the speed of light.
12
(2.8)
The position information of an object can be obtained by solving the following three geometrical
equations when at least three reference nodes exist in the system. This method is usually called
triangulation.
√
,
(2.9)
√
,
(2.10)
√
,
(2.11)
where , ,
are the ranges from the object being tracked to the three reference nodes and they
can be obtained by TW-TOA,
,
,
are the coordinates of the three reference
nodes, while
refers to the coordinate of the object being tracked.
RDEV A
RDEV B
𝑡𝑟
RFRAME 1
𝐵
𝑡𝑎𝑐𝑘
𝑡𝑟𝑜𝑢𝑛𝑑
RFRAME 2
𝑡𝑟
Figure 2.8 Illustration of Two-way Time-of-Arrival Ranging Protocol
2.3.2 Symmetric Double-Sided Two-Way Time-of-Arrival
The TW-TOA protocol has great ranging errors due to clock drift. In order to reduce the influence of
clock drift, a Symmetric Double-Sided Two-Way Time-of-Arrival (SDSTW-TOA) protocol can be
used. Figure 2.9 shows the procedure of a SDSTW-TOA protocol.
The following relationship can be obtained from Figure 2.9
.
(2.12)
The theoretical TOF regardless clock drift is
.
(2.13)
If we consider the clock drift, the TOF can be estimated as
.
13
(2.11)
Position information can be obtained by triangulation, the same procedure as shown in equations (2.9)
– (2.11).
The SDSTW-TOA protocol reduces the ranging error so that it is more tolerant to clock drift.
However, it doubles the communication traffic to be four RFRAMEs between two RDEVs when
comparing with TW-TOA.
RDEV A
RDEV B
𝑡𝑟
RFRAME 1
𝐴
𝑡𝑟𝑜𝑢𝑛𝑑
𝐵
𝑡𝑎𝑐𝑘
RFRAME 2
𝐵
𝑡𝑟𝑜𝑢𝑛𝑑
𝐴
𝑡𝑎𝑐𝑘
𝑡𝑟
RFRAME 3
RFRAME 4
𝑡𝑟
Figure 2.9 Illustration of Symmetric Double-Sided Two-Way Time-of-Arrival Ranging Protocol
2.3.3 Time Difference of Arrival
While both the TW-TOA and the SDSTW-TOA protocols result in heavy communication traffic in a
positioning system, Time Difference of Arrival is a protocol that helps reducing the traffic. Yet,
TDOA asks for synchronization between all the reference nodes while it is not necessary for the
former two.
In a positioning system where TDOA protocol is employed, at least three reference nodes are needed
together with an object that is being tracked, as illustrated in Figure 2.10. The object first broadcasts a
RFRAME with its identification number to all the reference nodes (RDEV A, B and C) with one of
them acting as a coordinator (RDEV A in this case). Each reference node that receives the RFRAME
records the time ( , and ) on receiving it. Both RDEV B and C transmit their time-stamp reports
to the coordinator RDEV A one after another. After collecting the time-stamps from all the reference
nodes, the coordinator RDEV A estimates the position of the object which is being tracked by solving
a non-linear algorithm
14
√
√
,
(2.12)
√
√
,
(2.13)
where c is the speed of light,
,
and
nodes, while
refers to the coordinate of the object.
are the coordinates of the three reference
By subtracting from (which is so-called TDOA) in equation (2.12), the difference between the
clock offset of the object and RDEV A and that of object and RDEV B can be eliminated. So it is with
subtracting from in equation (2.13).
With using TDOA protocol, communication traffic can be reduced so that channel occupancy can
maintain in a low level. Synchronization between all the reference nodes are essential and it can be
done by distributing reference clock (in wire) or wireless synchronization protocol.
Object
RDEV C
RDEV B
𝑡𝐵
RDEV A
𝑡𝐶
𝑡𝐴
𝑡𝐶𝐴
𝑡𝐵𝐴
Figure 2.10 Illustration of Time Difference of Arrival Ranging Protocol
15
Chapter 3
System Design
There are many issues that we have to consider about in the design of a positioning system, such as
system complexity, performance, power efficiency, synchronization requirements, channel occupancy,
compliance, and cost and so forth. The design work can be quite complicate if we do not follow any
design methodology. A top-down design methodology is employed to design the system.
To start with, the system requirements will be specified. After that, the ranging protocol will be
selected on the top level of system design, considering a few key criteria according to the system
requirements which are most important in this thesis. Then an overview of the system setup can be
illustrated according to the ranging protocol we have selected. The architecture of the system from
both hardware and software perspectives will be described in this chapter.
3.1 System Requirements
The goal of this thesis work is to design a positioning system based on the IEEE 802.15.4a standard
for indoor usage and preparing for outdoor usage as well. The system shall cover an area of 100 meters
×100 meters. The positioning accuracy shall try to get down to ±10 centimeters with an updating
frequency of at least 5 Hz. Positioning latency shall be less than 0.1 second. Besides, the system shall
be tolerant to a moving speed of the object which is being tracked (we will call it a tag in the rest of
the report) of 0.5 m/s. The power consumption of the tag shall be as low as possible. Performance of
the whole system is of the most importance while the system complexity shall be kept in a reasonable
level. Channel occupancy shall be maintained in a low level so that it is possible to extend the
positioning system to a wireless sensor network [22] which could contain thousands of tags. The
system shall at least comply with radio communication regulations in Sweden.
Here is a summary of all the requirements:











the system shall be based on the IEEE 802.15.4a standard,
design for indoor usage and prepare for outdoor usage as well,
operating area: 100 m × 100 m,
positioning accuracy: ±10 cm,
positioning updating frequency: ≥ 5 Hz,
positioning latency: ≤ 0.1s,
motion tolerance: 0.5m/s,
keep the power consumption of tag as low as possible,
low system complexity,
low channel occupancy,
scalability for wireless sensor network,
16

and comply with regulations in Sweden.
3.2 Protocol Selection
Three of the most commonly used ranging protocols have been introduced in section 2.3, which are
TW-TOA, SDSTW-TOA and TDOA. In this section, we will take a look into the performance of these
three protocols and also compare them from the traffic load and power efficiency perspectives.
As we have discussed before, the TW-TOA performs quite bad in the positioning accuracy because of
clock offset. SDSTW-TOA works much better from the perspective of accuracy. Table 3.1 and Table
3.2 [9] show the typical errors in TOF estimation by using TW-TOA and SDSTW-TOA, respectively,
where
,
, and
refer to the same terms described in section 2.3, respectively. 1 ns’ error
corresponds to about 30 cm inaccuracy. From these two tables, typical errors by using TW-TOA is
really high even if applying an high-quality crystal with tolerance of 2 ppm (parts per million), which
is not acceptable when considering performance and cost. On the contrary, SDSTW-TOA still gives
much better accuracy even if using low-quality crystals. As the targeting positioning accuracy has
been set to be ±10 cm, it will be difficult for a low cost implementation of TW-TOA protocol to meet
this requirement. So TW-TOA will not be employed in this design work.
Table 3.1 Typical Errors in Time-of-Flight Estimation Using TW-TOA
100 us
5 ms
2 ppm
20 ppm
40 ppm
80 ppm
0.1 ns
5 ns
1 ns
50 ns
2 ns
100 ns
4 ns
200 ns
Table 3.2 Typical Errors in Time-of-Flight Estimation Using SDSTW-TOA
2 ppm
20 ppm
40 ppm
80 ppm
1 us
10 us
100 us
5 ms
0.0005 ns
0.005 ns
0.05 ns
2.5 ns
0.005 ns
0.05 ns
0.5 ns
25 ns
0.01 ns
0.1 ns
1 ns
50 ns
0.02 ns
0.2 ns
2 ns
100 ns
As for channel occupancy, both TW-TOA and SDSTW-TOA result in a heavy load of communication
traffic. Assume in a simplest positioning system with three reference nodes and one tag, each node
needs to conduct two-way ranging with the tag. There will be at least 6 (2 × number of nodes) timestamp transmissions for TW-TOA while 12 (4 × number of nodes) transmissions for SDSTW-TOA
when the tag initiates the ranging and it maintains the positioning information. If the ranging is
initiated by a reference node, the numbers of transmissions turn to be 8 (2 × number of nodes + 2) for
TW-TOA and 14 (4 × number of nodes + 2) for SDSTW-TOA. As for TDOA protocol, only 3
transmissions are needed when synchronization is done via distributed cable from a reference clock. If
the synchronization is maintained wirelessly, number of transmissions turns to be 7. Therefore, the
communication traffic with TDOA is lighter than the other two protocols. With low channel
occupancy much more tags and reference nodes can coexist in one system, thus making the system
scalable for a wireless sensor network.
Apart from low channel occupancy, TDOA is also beneficial for keeping low power consumption on
tag from a power efficiency point of view. While the reference nodes may be charged by a charging
17
station, a tag is much preferable to be charged with battery as it is supposed to be a portable object. So
the power consumption of a tag shall be kept as low as possible. In order to achieve that, a tag shall
operate as little time as possible in high-power states and keep as much time as possible in low-power
states. For a UWB device, high-power states include transmission mode, receiving mode and idle
mode while low-power state usually refers to sleep mode or off mode. A TDOA protocol only require
a tag to transmit one time-stamp in transmission mode, while for the other time it may switch to sleep
mode. However, both TW-TOA and SDSTW-TOA will need a tag to work in either transmission
mode or receiving mode until the protocol is completed entirely. From this point, TDOA is much more
efficient on power consumption for tag.
Therefore, the TDOA protocol is employed in this design work for its good performance on accuracy,
low channel occupancy, and low power consumption on the tag side. TDOA also makes the system to
be scalable for future usage in large scale network such as wireless sensor network.
3.3 System Setup
As the TDOA protocol has been selected, the positioning system will then be designed according to
that. Design complexity, performance, power consumption, compliance and some other issues will be
stressed along with the system design.
A two-dimensional (2D) positioning system based on TDOA protocol requires at least one tag and
three reference nodes accommodating together. To be prepared for three-dimensional (3D) positioning,
the system in this thesis work will employ a tag in the simplest case and four reference nodes with
three to be called base stations and the left one to be the coordinator which coordinates the whole
system, collects the time-stamp information and computes the position information.
Synchronization is a crucial part of a positioning system based on TDOA. For indoor use, it can be
done wirelessly. As stated before, however, UWB signal is not allowed to be transmitted from a
device at a fixed installation in Sweden [21] and other EU countries [7]. So cables will be needed to
distribute a reference clock to the other base stations. While a base station cannot send time-stamp
reports to the coordinator by UWB signal, other means of transmission of those messages shall be
introduced which could be wired or wireless. As the coordinator may also need to send information to
a tag which is a portable object, a solution based on wireless communication is reasonable. So the way
of transmitting time-stamp reports from base stations to coordinator is wireless communication other
than UWB signal. To differentiate from UWB communication, the solution of wireless communication
is called wireless link in the rest of the report.
The proposed system setup is shown in Figure 3.1. The system consists of one coordinator, three base
stations, and one tag. Each device is composed of a UWB module and a wireless link module. The
UWB module, which is based on the IEEE 802.15.4a, is used for the UWB signal transmission and
reception. On the other hand, the wireless link is used for data transmission from the reference nodes
to any devices in the same system. All the reference nodes, including the coordinator and the three
base stations, share the same reference clock from the coordinator via distributed cables. Yet the clock
distribution is for the purpose of outdoor usage of the positioning system, so it may not be
implemented.
18
Base Station 1
Base Station 3
UWB
UWB
Wireless Link
Wireless Link
Tag
UWB
Wireless Link
Base Station 2
Coordinator
Distributed Cable
UWB
Wireless Link
UWB
Wireless Link
Figure 3.1 System Setup of Positioning System Based on the IEEE 802.15.4a
The system setup shown in Figure 3.1 only illustrates a simplified structure of a positioning system.
Detailed architectures of the tag and the reference node are shown in Figure 3.2 and Figure 3.3
respectively.
Battery
DC/DC
Regulator
UWB
Transceiver
Control
Interface
Microcontroller
Optional
Control
Interface
Wireless Link
Control Interface
Sensors
Optional
Figure 3.2 Architecture of a Tag in the Proposed Positioning System
A tag consists of one UWB transceiver, one microcontroller, one DC/DC regulator and one 3.6 V
(volts) coin-cell battery in the simplest case. The wireless link is optional because it is for outdoor
usage only. Yet, a control interface will be reserved for the extension of the wireless link. In order to
enhance the capability of a tag, a few sensors may be included in this design work. The tag is
considered for the simplest case because the board size of a tag is preferred to be as small as possible.
19
As for the reference node, a microcontroller, a UWB transceiver, a clock buffer, a wireless link, an
Ethernet PHY and a DC/DC regulator with external power supply will be included in this design work.
Besides, another UWB transceiver may be included on the same board for future enhancing of the
current system with better positioning accuracy by a hybrid protocol combining both TDOA and AOA.
An Ethernet PHY transceiver is going to be implemented in the reference node in order to connect the
positioning system to a positioning server that will be introduced in the future work. Besides, a few
sensors may be introduced in the reference node as well.
4-14 V
Power Supply
DC/DC
Regulator
On Board
Clock
Distributed
Clock
UWB
Transceiver
Control
Interface
Clock Buffer
Microcontroller
Control
Interface
Wireless Link
Control Interface
UWB
Transceiver
Sensors
Optional
Optional
Distributed Cable
Ethernet PHY
Optional
Figure 3.3 Architecture of a Reference Node in the Proposed Positioning System
3.4 Solutions Selection
In this section, solutions of all the key modules will be introduced.
3.4.1 UWB Transceiver
We have selected an UWB transceiver which is compliant with the IEEE 802.15.4a standard from a
manufacturer. It is a compact IC solution with 48-pin QFN (Quad-flat no-leads) package. It supports 6
frequency bands from 3.5 GHz to 6.5 GHz. Data rates of 110 kbps, 850 kbps and 6.8 Mbps are
supported. Maximal transmission power is -10 dBm and it is configurable to meet regulations in
different countries.
Apart from the capability of ranging, the transmission of data information in one packet is also
supported. Supply voltage can vary from 2.8 V to 3.6 V. The transceiver is facilitated with a highspeed SPI (Serial Peripheral Interface) interface so that it can be configured and controlled by an
external controller.
As we have discussed before, clock drift has significant influence on the performance of the ranging
accuracy. Therefore, a high-performance, low-drift temperature compensated crystal oscillator (TCXO)
is selected for the UWB transceiver; it is the IT3200C [29] from Rakon which has a maximal clock
20
drift of ± 1 ppm. Besides, a clock buffer with part number CDCLVC1102 [30] from Texas
Instruments is also selected for the clock distribution.
3.4.2 Wireless Link
The solutions for the wireless link are widely available in the market, so they are much more open to
be selected. There are many factors to be considered about when choosing a suitable solution, such as
operation frequency, power consumption, range, data rate, latency, learning curve, cost and so on.
Here are some requirements for the wireless link:





range ≥ 150 meters,
data rate ≥ 100 kbps,
packet error rate < 1%,
low power consumption,
and can be used all over the world.
Amount those factors operation frequency is usually the first one to be considered. According to
regulations of Swedish PTS, there are some non-specific unlicensed frequency bands for short range
general purpose devices (SRD) in Europe, as listed in Table 3.3 [21].
Table 3.3. Unlicensed Frequency Bands for Non-specific Short Range Devices in Sweden
Frequency band
ERP
Duty cycle Channel bandwidth Standards
13.553-13,567MHz
No limits(1) No limits
No limits
RFID
26.957-27,283MHz
10mW
No limits
No limits
40.66-40.7MHz
10mW
No limits
No limits
433.05-434.79MHz
15mW
No limits
No limits
863-865MHz
25mW
≤0.1%
No limits
865-868MHz
25mW
≤1%
No limits
RFID, Zigbee
868-868.6MHz
25mW
≤1%
No limits
868.7-869.2MHz
25mW
No limits
No limits
869.4-869.65MHz
500mW
No limits
<25kHz
869.7-870MHz
25mW
No limits
No limits
RFID, Zigbee,
(2)
2400-2483.5MHz
100mW
No limits
No limits
Bluetooth, WLAN
(2)
5.725-5.875GHz
25mW
No limits
No limits
24-24.25GHz
100mW(2)
No limits
No limits
61-61.5GHz
100mW(2)
No limits
No limits
122-123GHz
100mW(2)
No limits
No limits
(2)
244-246GHz
100mW
No limits
No limits
(1) High field strength: 42 dB u A / m at 10 m distance
(2) EIRP
The SRD bands cover from 6.765 MHz to 246.000 GHz. However, not all the bands give the targeting
performance because the lower bands have quite narrow bandwidths which cannot guarantee enough
data rate and higher bands have quite short range which is not sufficient for our application.
There is no direct formula between bandwidth and data rate since data rate depends on not only
bandwidth but also modulation scheme, noise and so on. A rough estimation can be done by Nyquist
theorem. According to Nyquist theorem, the bandwidth should be at least half of the data rate (noisefree channel, binary signals):
21
Bandwidth ≥ data rate / 2
(3.1)
To achieve 100kbps data rate, bandwidth of radio channel should be at least 50 kHz. The SRD bands
less than 100 MHz usually have bandwidth less than 50 kHz, so they are not in our concerning.
As for higher bands, the higher the frequency the shorter the range regardless the transmission power
and some other factors. Friis Formula gives a good estimation on the highest frequency we need to
investigate:
(3.2)
where and are the transmitted and received signal power respectively, and
are the antenna
gains of the transmitter and the receiver respectively, and λ is the wavelength. Assume link budget is
100 dB, d (range) = 150 meters, the shortest wavelength is 0.0188 meter which corresponds to 16 GHz
in frequency. The targeting operation frequency shall come down to bands much lower than 16 GHz,
however, since Friis Formula is a free space propagation model and link budget in near-ground nonline-of-sight environment is much smaller.
Therefore, only those bands at 433 MHz, 868 MHz and 2.4 GHz are in our concerning for this project.
The sub-1 GHz (including 433 MHz and 868 MHz) based solutions usually offers higher link budget
and they are quite robust to multipath effect. However, those bands are only allowed in Europe for
unlicensed usage. On the contrary, 2.4 GHz based solutions are allowed to be used all over the world
without a licensing requirement. High data rate can be easily obtained at 2.4 GHz. Besides, there are
many open source stacks available at this band designed for varies kinds of technologies, such as
WLAN, Bluetooth, RFID [23], and Zigbee [24] and so on. They are usually quite robust with low
packet error rate.
The range of solutions based on WLAN, Bluetooth and RFID technologies is difficult to get more than
100 meters, especially for the low power devices. RFID only has a short range up to tens of meters so
such kind of modules will be out of our consideration. As for Bluetooth, only those in power class 1
have a range of up to several hundred meters, but the power consumption may be a few hundred mW
which will be a big problem for energy starving applications. WLAN faces almost the same problem
as Bluetooth. Apart from that, the interference of WLAN signal is quite severe nowadays, making it
unstable from time to time. However, Zigbee solution is quite promising in our application when
considering ordinary Zigbee modules have a range up to several hundred meters and they are usually
designed for low power applications.
Therefore, a solution based on Zigbee standard will be employed in this work in order to provide a
robust wireless link in the positioning system. Amount all the solutions on Zigbee from varies
manufacturers, the CC2530 [25] from Texas Instruments is selected for its good performance, low
power, low cost. A robust communication stack called Z-Stack [26] is also available from Texas
Instruments and it is license-free for the development of Zigbee applications based on CC2530.
3.4.3 Microcontroller
A microcontroller is in needed for the control of the UWB transceiver and the wireless link. A more
powerful microcontroller is interested for the reference nodes (including coordinator and base stations),
while a higher power-efficient one is preferable for the tag. Requirements for the two microcontrollers
are listed in Table 3.4.
22
Parameter
Speed
Memory
Peripheral
OS
Table 3.4 Requirements for Microcontrollers
Reference Nodes
Tag
≥ 50 MHz
≥ 8 MHz
≥ 128 kB flash
≥ 64 kB flash
≥ 32 kB RAM
≥ 32 kB RAM
SPI, UART, I2C, CAN, Ethernet, SPI, UART, I2C, ADC
ADC
RTOS
RTOS
The speed requirement for the microcontroller on reference node is no less than 50 MHz for it needs to
be fast enough to handle tasks such as synchronization, position calculation and routing and related
issues in a wireless sensor network. On the other hand, the speed requirement for the microcontroller
on a tag is set to be no less than 8 MHz because it needs to keep the power consumption as low as
possible while maintaining a sufficient performance.
Amount all the microcontrollers from key manufacturers, the STM32F107VC [27] from
STMicroelectronics is selected for the reference node while the MSP430F5438A [28] from Texas
Instruments for the tag. The STM32F107VC provides high performance with up to 72 MHz system
frequency and varies kinds of peripherals including Ethernet. The MSP430F5438A has a 16-bit
architecture and is designed for low-power applications.
3.4.4 DC/DC Regulator
A DC/DC regulator that converts a higher level voltage to a lower level one is called buck (step down)
regulator while boost (step up) regulator for the one that converts a lower level voltage to a higher
level one.
The tag will be charged by a 3.6V coin-cell battery whose voltage will decrease gradually to be less
than 2V when using for a long time. Considering the supply voltage of its microcontroller and the
UWB transceiver is 3.3V, a buck-boost DC/DC regulator is needed for the tag system. The STBB1AXX [31] from STMicroelectronics is selected for its high converting efficiency. It supports input
voltage from 2 V to 5.5V while output voltage can be adjusted from 1.2V to 5.5V.
As for the reference nodes, they will be charged by AC/DC power adapters that convert the 220V AC
electricity to a stable 5V DC. Then the on board DC/DC regulator will convert the 5V voltage to 3.3V.
A buck regulator is needed in this case. The TPS62111 [32] from Texas Instruments is selected for its
high efficiency. It supports input voltage form 3.1V to 17V and has a fixed output voltage of 3.3V.
3.4.5 Connectivity
The Ethernet interface on a coordinator or a base station is designed for connecting the positioning
system to a positioning server when the system is extended for a wireless sensor network. The
Ethernet PHY transceiver should support at least 10/100BASE-TX. The DP83848K [33], a solution
from Texas Instruments, is selected in this design work for its high performance and low power
consumption.
Besides, a RS232 interface is implemented on the reference node for the system-level debugging
purpose. A RS232 transceiver ST3232C [34] from STMicroelectronics is chosen for that.
23
Finally, a JTAG interface is also included for programming and register-level debugging.
3.5 Software Architecture
To reduce the design complexity of software development and provide more flexibility on it, an
approach to design the software according to the open systems interconnection (OSI) model is widely
used in communication system. In such a model, a system is divided and implemented in multiple
layers with each layer interacting only with the layers beneath and above it. An OSI model isolates
each layer from the implementation details of the other layers in a system. A typical OSI model for a
communication device is shown in Figure 3.4.
APP
NET
LINK
MAC
PHY
APP Layer Protocol
APP
Layer-4 Protocol
NET
Layer-3 Protocol
LINK
Layer-2 Protocol
MAC
Layer-1 Protocol
Device
PHY
Device
Figure 3.4 Architecture of a Communication Device in OSI Model
A communication device usually contains a physical layer (PHY), a media access layer (MAC), a link
layer (LINK), a network layer (NET) and an application layer (APP). Each layer only interacts with
the layers next to it and specific protocol handles each layer, so that each layer is independent to each
other. Take the MAC layer for example, it only servers the upper LINK and responds to the PHY.
MAC Protocols such as CSMA and LEACH [22] take care how a number of devices access a shared
communication medium from the timing perspective. So a MAC protocol does not have to care about
how a PHY will act in this medium, which gives high flexibility on its implementation and porting
from one PHY to the other PHYs.
The software architecture of the proposed positioning system in this thesis work based on IEEE
802.15.4a will be designed according to the model shown in Figure 3.4. Some parts of the model will
be put much effort into so that sub-layers may be introduced while others may not be implemented.
The layers are divided in a way that they are independent to each other, which means one layer only
has to care about the how APIs (application programming interface) from the lower layer serves it
without knowing the details of their implementations.
24
3.5.1 Software Architecture of Tag
A tag has a microcontroller, a UWB transceiver and an optional wireless link and other sensors. The
software architecture of a tag is shown in Figure 3.5.
APP
MAC
Z-Stack
PHY
Device Driver
Device Driver
Device Driver
Interface
Interface
Interface
Wireless Link
UWB
Sensors
Microcontroller
Driver
Figure 3.5 Software Architecture of a Tag
The software architecture consists of five major parts, which are the microcontroller driver layer,
UWB transceiver layers, wireless link layers (optional), sensors layers (optional) and an APP layer.
The microcontroller driver provides all the peripherals needed for the UWB transceiver, the wireless
link and sensors. Basically this driver layer initializes the clock system, timers, interrupts and different
peripherals.
The UWB transceiver, the wireless link and the sensors have their own models with several layers. On
bottom of the UWB transceiver layers, there is an interface layer which handles the mechanism of the
communication interface (SPI in this case). Above of the interface layer, the device driver layer
manages the configuration of UWB transceiver on the register level. A PHY layer on top of the device
driver is an IEEE 802.15.4a compliant layer which sets all the communication parameters such as
channel, data rate, preamble length, synchronization and so on. Besides, a MAC layer is also included
which is also compliant to the IEEE 802.15.4a standard and the ALOHA protocol [35] based on timeslot allocation is implemented on this layer.
For the wireless link (which is optional in this thesis work), the interface layer and the driver layer
have the same features as those for the UWB transceiver. On top of them, a full communication stack,
Z-Stack from Texas Instruments, is employed for its high robustness.
25
The sensor layers are implemented for various kinds of sensors such as temperature sensor, humidity
sensor, and accelerometer and so forth.
Finally, the application layer takes care all the tasks involving with all the beneath branches. The
TDOA protocol is implemented on this layer.
3.5.2 Software Architecture of Reference Node
The software architecture of the reference node (including coordinator and base station) is much more
complicated than that of the tag.
APP
MAC
Z-Stack
PHY
TCP/IP
Device Driver
Device Driver
Device Driver
Device Driver
Interface
Interface
Interface
Interface
Wireless Link
UWB
Ethernet
Sensors
Microcontroller
Driver
Figure 3.6 Software Architecture of a Reference Node
Apart from the microcontroller driver layer, UWB transceiver layers, and wireless link layers, which
have the same features as those of a tag and have already been described in the previous section,
Ethernet layers are added for the reference node. The Ethernet contains an RMII (Reduced Media
Independent Interface) [36] interface layer, device driver layer specifically made for the selected
Ethernet PHY, and a widely practiced TCP/IP stack [37].
The sensors branch contains supports for temperature sensor, LEDs, and RS232 interface.
As for the APP layer, the TDOA protocol is also implemented here. But the implementation for
coordinator and base station will be different because the tasks in their own roles vary from each other.
26
Chapter 4
System Implementation
The system design has been introduced in chapter 3. Here in this chapter, the implementation of the
proposed positioning system will be described. The implementation covers both hardware and
software, according to the architecture shown in the previous chapter. For the hardware
implementation, a few challenges will be discussed in this chapter. As for software part, we will focus
on the application layer where some key issues including the TDOA protocol shall be addressed.
4.1 Hardware Design
The hardware of a reference node is divided into two parts to suit the targeting enclosure; one part
includes all the main components including a microcontroller, two UWB transceivers, a wireless link,
LEDs and a temperature sensor while the other contains a DC/DC regulator, an Ethernet PHY, and a
RS232 transceiver. These two parts are connecting together via a 2×10 flat cable. Here in the rest part
of the report, the former part of the reference node is called base board and connection board for the
latter one.
The base board is implemented on a 6-layer FR4 PCB while a 4-layer FR4 PCB for the connection
board. Stackups of these two boards are illustrated in Figure 4.1 and Figure 4.2, respectively. On the 6layer PCB, except the second layer and the fifth layer which are assigned as ground plane and power
plane, respectively, the other four are signal layers. As for the 4-layer PCB, the two layers in the
middle are also assigned as ground plane and power plane, respectively, with the other two for
signaling.
Figure 4.1 Stackup of a 6-layer FR4 PCB
27
Figure 4.2 Stackup of a 4-layer FR4 PCB
Apart from the base board and the connection board, a tag is also implemented with a 6-layer PCB
board, the same as shown in Figure 4.1.
4.1.1 Block Diagram
4.1.1.1 Base Board
The base board on a reference node consists of a STM32F107 microcontroller, two IEEE 802.15.4a
compliant UWB transceivers, a CC2530 wireless link, a temperature sensor, three LEDs, and a few
connectors for debug and connectivity purposes. A detailed block diagram of the base board is shown
in Figure 4.3 while figure 4.4 shows the two sides of the finished PCB. The two UWB transceivers are
connecting with the microcontroller via the same SPI interface with two different selection pins. The
two UWB transceivers share the same clock driven by a clock buffer CDCLVC1102 from a TCXO
IT3200C. Each of them connects a balun (which converts between a balanced signal and an
unbalanced signal) and a chip antenna. The CC2530 is connected with the microcontroller through a
UART (Universal Asynchronous Receiver/Transmitter) interface. The RF signal part that connects
PCB Antenna
Match
Balun
Chip Antenna
Balun
UWB
Transceiver
Crystal
CC2530
UART
TCXO
Clock Buffer
SPI
UWB
Transceiver
LEDs
ADC
Temperature
Sensor
STM32F107
Chip Antenna
Balun
GPIO
JTAG
Reset
RMII
UART
3.3V
Supply
Figure 4.3 Block Diagram of Base Board
with CC2530 contains a balun, an L-matching network and a PCB antenna. Additionally, there are
three LEDs which are connected via three GPIO ports, and a temperature sensor which is connected
through an ADC port from the microcontroller. Finally, a header for the JTAG interface is included on
28
the base board, as well as a RESET button, a 20-pin header for the RMII interface and UART interface
to the connection board. There is one pin of the 20-pin header which is assigned for the 3.3V power
supply and another pin assigned for the system ground.
PCB
Balun Antenna
Chip
Antenna
LEDs
Balun
RMII
Header
CC2530
STM32F107
UWB
Transceiver
UWB
Temperature
Transceiver
Sensor
Balun
TCXO Clock Buffer
Reset Button
JTAG
UART
Chip
Antenna
Figure 4.4 Finished PCB of a Base Board
4.1.1.2 Connection Board
The connection board contains a TPS62111 DC/DC regulator, a DP83848K Ethernet PHY transceiver,
a magnetic transformer, a RJ45 type connector and a ST3232C RS232 transceiver. Block diagram of
the connection board is shown in Figure 4.5. The DC/DC regulator (which supports input voltage from
3.1V to 17V) converts a 5V power source to 3.3V which will be used to power up both the connection
board and the base board. The DP83848K Ethernet transceiver is controlled via the RMII interface. On
the other side, a transformer with a RJ45 type terminal is connected to the Ethernet transceiver. Finally,
3.1-17V
DC Supply
RMII
UART
ST3232C
DP83848K
Debug Header
Transformer
RJ45
TPS62111
3.3V DC
Output
Figure 4.5 Block Diagram of Connection Board
29
3.3V DC
Output
a RS232 transceiver ST3232C is included on the board to provide a connection between the reference
node and a hyper terminal on a computer for debug purpose.
Figure 4.6 shows the finished connection board. The connection board is connected with the base
board through a 20-way flat cable, which is shown in Figure 4.7.
ST3232C
RJ45
UART
Header
Transformer
DP83848K
RMII
Header
TPS62111
Power
Jack
Figure 4.6 Finished PCB of a Connection Board
20-way
Flat Cable
Figure 4.7 Finished PCB of a Reference Node
4.1.1.3 Tag
The tag board shown in Figure 4.8 consists of a MSP430F5438A microcontroller, a UWB transceiver
with balun and chip antenna, and a STBB1-AXX DC/DC regulator with a 3.6V coin-cell battery. A
header with SPI interface is also included in the design to be prepared for the extension of a CC2530
wireless link. Figure 4.9 shows the finished PCB of a tag.
30
SPI for
Wireless Link
Chip Antenna
UWB
Transceiver
Balun
SPI
3.6V
Coin-Cell
Battery
MSP430F5438A
3.3V
Output
TCXO
STBB1-AXX
JTAG
Figure 4.8 Block Diagram of Tag
TCXO
UWB
Transceiver
Chip Antenna
MSP430F5438A
Balun
STBB1-AXX
Figure 4.9 Finished PCB of Tag
4.1.2 Microstripe Design
In high speed electronic systems, impedance discontinuity on a signal track will lead to signal
reflection, thus attenuating the signal to be received / transmitted. This effect highly influences the
performance of an RF device, specifically, the range and accuracy of the proposed positioning system.
Therefore, it is very important to avoid any impedance discontinuity and try to keep the same
impedance on a signal line.
The signal tracks from the RF output/input signals (positive and negative) of the UWB transceiver to
the chip antenna (with two capacitors and a balun in between) work at frequency range from 3.2 GHz
to about 7 GHz. The design of these tracks is critical to the positioning system. Layout of these signal
tracks are shown in Figure 4.10. RF_N and RF_P are output/input pin-outs of the UWB transceiver
and they have 100Ω differential load impedance which corresponds to 50Ω signal-ended impedance
on each pin. Therefore, the signal tracks routed from these two pins shall have 50Ω impedance in
order to avoid impedance discontinuity. This kind of signal track is called microstripe [39].
Figure 4.11 illustrates a simple structure of a microstripe. A microstripe is a signal line which is
separated by a dielectric from a reference ground plane. The impedance of a microstripe can be
calculated by various software tools such as AppCAD [38] from Agilent Technologies, or by
empirical formula (when 0.1<W/H<2.0 and 1< <15)
31
Ω.
√
(4.1)
AN
Balun
Capacitor
1&2
RF_P
RF_N
Figure 4.10 Layout of High Frequency Signal Tracks in Base Board
Signal Track
Ground Plane
Figure 4.11 Structure of a Microstripe
In this design work, RF_N and RF_P (which are on the top layer) take the GND plane (which is shown
in Figure 4.1) as the reference in order to keep width of the tracks as small as those two pads of the
transceiver. While the thickness (T) of the tracks is 38um and distance (H) between top layer and GND
plane is 123um, the calculated width (W) is 199.8um with AppCAD to achieve 50Ω impedance.
Figure 4.12(A) shows the calculation with this tool.
As for the track AN shown in Figure 4.10, its width shall be wider to match the size of the pad of
balun. Thus, it uses the Midlayer1 as the reference plane. As the GND plane is in between of top layer
and Midlayer1, the area on the GND plane beneath this track shall be kept open (which means no any
conductive track or region shall be placed). The calculated width (W) with AppCAD to achieve 50Ω
impedance is 983.5um when the distance (H) between top layer and Midlayer1 is 553um, as shown in
Figure 4.12(B). The corresponding layout from these calculations has already shown in Figure 4.10.
32
(A)
(B)
Figure 4.12 Microstrip Calculation with AppCAD
4.1.3 Grounding
The theory behind microstripe is actually called transmission line effect where a conductor must be
considered as a distributed series of inductors and capacitors [39]. Transmission line effect influences
a lot on the signal integrity in high speed electronic system. Generally, any circuit length of at least
1/10th of the wavelength of the conducting signal shall be considered as a transmission line. The
transmission line effect also has a huge impact on grounding in PCB design.
33
Ground plane/region is usually plated on several layers and covers a large board area in a multi-layer
PCB design. Figure 4.13 illustrates the influence of transmission line effect on grounding. Suppose
two ground regions plating on both layer 1 and layer 2 are connected with each other through a
through-hole via. Point A and B may have the same potential since they are quite near from each other.
However, the potentials at point C and D may be different when they are far apart which causing the
transmission line effect. This effect will make the whole system unstable and unreliable.
Layer 1
A
C
B
D
via
Layer 2
Figure 4.13 Illustration of the Impact of Transmission Line Effect on Grounding
It is fundamentally important to avoid this effect so as to keep the same electric potential on the whole
ground plane. One way to solve this problem is to place more through-hole vias in between ground
planes on different layers. Those vias shall be within a distance of 1/10th of the wavelength of the
conducting signal nearby. The distance (d) is calculated by the following formula
√
(4.2)
where c is the speed of light, is the dielectric parameter and is the central frequency. In this
design case = 4.6 and = 5.1 GHz, so the calculated distance is about 2.74mm. Therefore the
adjacent ground vias shall be placed less than 2.74mm from each other along the high frequency
regions. Figure 4.14 shows the placements of ground vias along the microstripe between the chip
antenna and the balun. All the adjacent ground vias are placed 1.5mm from each other.
Ground Vias
Figure 4.14 Placements of Ground Vias to Avoid Transmission Line Effect
34
4.3 Software Design
The software design in this work follows the software architecture as described in section 3.5. The
software is designed in a layered manner, covering all the devices including coordinator, base station
and tag. The design work mainly focuses on the lower layers such as microcontroller driver, interface
layer and device driver layer, as well as the highest application layer which is the most important part
in this thesis. As for the other layers in the middle such as TCP/IP and Z-Stack, we employ the work
from third-party suppliers. The PHY and MAC for the UWB transceiver are developed by B. S. Heyi
in [40] and A. Monge in [41]. This section will focus on describing the APP layer in the three types of
devices – one tag, three base stations and one coordinator. The APP in this thesis work is designed for
indoor usage only and is flexible to be changed for outdoor scenario.
The positioning system proposed in this thesis works in the following procedure:
1.
2.
3.
4.
Tag initiates the positioning by broadcasting a poll message with time-stamp. After that it
switches to sleep mode.
When a base station receives the poll message from the tag, it records the time on receiving that
message and send this time message together with its own ID to the coordinator via wireless link.
All the three base stations perform the same procedure one by one chronologically depending
upon the time sequence on receiving the poll message.
The coordinator receives the poll message from the tag and it records the time on receiving that
message. It waits for any time message from the three base stations through wireless link. After
collecting all the three time messages from them, the coordinator calculates the positioning of the
tag accordingly with the TDOA algorithm. After the calculation, it sends the position data, as well
as other information (such as the four time-stamps on receiving the poll message at base stations
and coordinator) to a positioning server via Ethernet.
Tag switches to active mode after some amount of time and sends a poll message again. Step 1-4
performs periodically in a frequency of 5 times per second.
Synchronization between base stations and coordinator is conducted in the MAC layer when the tag is
operating in sleep mode. So it shall not influence the positioning procedure. The synchronization
protocol has already been introduced in [41].
Z-Stack is used for the communication between base stations and coordinator because the wireless link
operates at 2.4 GHz which will not interfere with the UWB signal and Z-Stack has a dedicated MAC
and NET protocol which handles channel access, acknowledgement and re-transmission.
4.3.1 Flow Chart
4.3.1.1 Tag
The following flow chart in Figure 4.15(A) shows the procedure of the program on tag. The program
starts with the initiation of the microcontroller, including clock system and different peripherals such
as SPI and timer. The UWB transceiver will be initialized afterwards. Communication parameters
such as channel, preamble length, data rate, transmission power and so on will be set here. The UWB
transceiver will be switched to transmission mode after the initiation. Then a poll message with a timestamp encapsulated in the packet is broadcasting to other devices. The UWB transceiver will be
switched to sleep mode after the transmission and the microcontroller will change to sleep mode as
35
well. After sleeping for certain amount of time, both the microcontroller and UWB transceiver will
switch to active mode and a poll message will be sent again. The transmission of poll message is
conducted periodically at 5 times per second.
(A) Tag
(B) Base Station
(C) Coordinator
Figure 4.15 Program Flow Charts in Tag, Base Station and Coordinator
4.3.1.2 Base Station
The program in bases station starts with microcontroller initiation (setting clock system, different
peripherals and timers), UWB transceiver initiation (configuring UWB communication parameters),
and wireless link initiation (configuring communication parameters and setting wireless link network
with other base stations and coordinator). After the initiations, the UWB transceiver will be switched
to receiving mode and it waits for a poll message transmitted from a tag. It records the time at
receiving the poll message and switches to sleep mode after the reception is finished. Then a message
containing data such as the reception time and device ID is sent to the coordinator via wireless link.
36
Finally the UWB transceiver is switched to receiving mode again to wait for the next poll. The flow
chart in base station is shown in Figure 4.15(B).
4.3.1.3 Coordinator
Similar to the base station, the program in coordinator also starts with the initiation of microcontroller,
UWB transceiver and wireless link. The wireless link will work in receiving mode after the initiation.
The UWB transceiver is switched to receiving mode as well to wait for the poll message from a tag.
When a poll message comes, it records the reception time and switches the UWB transceiver to sleep
mode after the reception is finished. After that, the coordinator waits for the time messages from the
base stations and receives one by one once they come. When all the three time messages are collected,
the coordinator calculates the positioning of the tag by performing the TDOA algorithm calculation.
Finally, it sends the position information and other kinds of message to the positioning server via
Ethernet and it switches the UWB transceiver to receiving mode again for the next poll. Figure 4.15(C)
shows the whole process for a coordinator.
4.3.2 TDOA Algorithm
The fundamentals of TDOA calculation have already been introduced in section 2.3.3. There are many
ways on how resolving the equations (2.12) and (2.13). They can be categorized as noniterative
methods and iterative ones, which are described in [42]. Here in this thesis, a noniterative approach
which is a direct determination method is employed for two-dimensional positioning.
Assume that
is the coordinate of the coordinator (reference node 1), and
,
and
are the coordinates of the three base stations (reference node 2, 3, and 4), respectively, while
refers to the coordinate of the tag. The relationship between all these coordinates is
(4.3)
where c is the speed of light, is the reception time at reference node i and is the transmission time
at tag which is unknown because the clock at tag is not synchronized with the reference nodes.
Subtracting (4.3) at i = 1 from that at i = 2, 3, and 4 leads to
(
)
(4.4)
where
(4.5)
Defining the estimated TDOA between reference nodes i and j as
(4.6)
Then eliminating
from (4.4) produces
(4.7)
37
where
(4.8)
and
(4.9)
where
(4.10)
Combining (4.7) and (4.9) gets to
{
(4.11)
This method assumes that all the reference nodes are synchronized. It ignores the TOA estimation
errors and it only considers the two dimensional position. Yet this is the simplest approach to get the
position of a tag and it requires the minimal calculations. Three dimensional calculations and
iterative methods can be found in [42].
38
Chapter 5
Performance Evaluation
The proposed positioning system based on IEEE 802.15.4a has been designed and developed, covering
both hardware and software, as described in Chapter 4. Here in this chapter, the functionality of the
system will be verified and the performance of the whole system is going to be evaluated as well.
5.1 Functionality Verification
The functionality verification on the proposed system mainly focuses on the hardware, including the
base board, connection board and tag. Almost all the main components have been covered in the
verification. A few tests have been conducted to verify them. The test results show all the main
components on the finished boards work properly.
5.1.1 DC/DC Regulators
When powering up the connection board with a 5V AC/DC power adapter, the output voltage of
TPS62111 on connection board is constant 3.34V. When powered with a DC power supply changing
from 3.1V to 17.0V, the TPS62111 has an output voltage of 3.34V.
Similarly, the STBB1-AXX on the tag has an output voltage of 3.32V in both cases when supplied
with a 3.6V coin-cell battery and when powered with a DC power supply changing from 1.2V to 5.5V.
Both the DC/DC regulators work fine when powered with a proper supply voltage.
5.1.2 Microcontrollers
When powering up the base board via the connection board with a 5V AC/DC power adapter, the
external crystal of the STM32F107 microcontroller has a 200mV peak-to-peak sinuous clock output.
Properly configure the boot strap of this microcontroller and run a testing program which toggles the
three LEDs one by one on the base broad, the LEDs are toggled properly. High-level output voltage of
the corresponding GPIOs is 3.31V.
Power up the MSP430F5438A on the tag with a 3.6V coin-cell battery and run a testing program
which toggles three GPIOs one after another periodically (while the microcontroller is running based
on internal crystal), the corresponding GPIOs have rectangular pulse output which has a peak-to-peak
of 3.31V.
Both the microcontrollers on base board and tag work fine.
39
5.1.3 UWB Transceiver
Power up the base board properly and control the UWB transceiver by a computer with a USB-to-SPI
adapter. When running sample software from the manufacturer, the software recognizes the
transceiver properly. Configurations of the transceiver can be done with the software on the
transceiver.
Similarly, power up a tag properly and use a computer to control the UWB transceiver via a USB-toSPI adapter. The transceiver is recognized by the sample software from the manufacturer. Configure
the transceiver on the tag at receiving mode while transmission mode for the one on a base board with
the same communication parameters setting, the UWB transceiver receives the UWB packet from the
tag. This test has shown that the UWB transceivers on all the board work fine.
Figure 5.1 shows the UWB signals in time domain transmitted from the transceiver when measured
with a high frequency oscilloscope.
Figure 5.1 UWB Signals from the UWB Transceiver
40
5.1.4 Wireless Link
Connect the CC2530 wireless link on a base board with a SmartRF05 [43] evaluation board from
Texas Instruments to a computer, the SmartRF Studio [44] recognizes the device and configurations
on the device can be done with this software on the computer.
Connect another CC2530 on a base board to the computer in the same way and configure it at
receiving mode while the former CC2530 in transmission mode, the former one receives the packets
sent from the latter one.
This test has verified that the CC2530 device on the base board works fine.
5.1.5 Ethernet
Connect a base board to a connection board with a 20-way flat cable properly and connect the
connection board to a WLAN router via an Ethernet cable. Run sample program on HTTP application
in the STM32F107 microcontroller, the base board is registered as a client on the router and an IP is
assigned to the device. A website that runs in the microcontroller is accessible from a web browser on
a computer.
This test has verified that the DP83848K Ethernet transceiver on the connection board works properly.
5.2 Performance Evaluation
The previous section shows that all the main components on the boards designed in this thesis work
are functional which makes the software development workable based on the proposed platform. A
few tests and measurements have been done to give a comprehensive evaluation of the performance of
the whole positioning system. This part of work covers the RF measurements, wireless link range and
robustness measurements, and positioning range and accuracy measurements.
The RF performance highly influences the radio range and the positioning accuracy. The range of the
wireless link also affects the range of the positioning system. And the robustness of the wireless link is
important to the positioning latency. Finally, the range and the accuracy of the UWB Transceiver are
the most important parameters in this thesis work.
5.2.1 RF Performance
As we have discussed before, the performance of the radio part highly influence the range and the
accuracy of a positioning system. The S11 parameter measurement is usually conducted to evaluate
the performance of an RF device.
The S11 parameter is the ratio of the power reflected from a port / terminal to the power which is fed
to the same port. It is also known as reflection coefficient or return loss and it usually varies with
frequency. For a good radio design, the S11 shall be as low as possible so as to deliver as much power
as possible to a terminal with minimal reflection.
We have measured the S11 at the end point of the microstripe (which is discussed in 4.1.2) with a
spectrum analyzer. The signal track in our concerning for these measurements traces from the end
point of the microstrip (at antenna side) to the RF input ports of the UWB transceiver so the effects of
41
the balun and capacitors have already been considered. The results from the measurements are shown
in Figure 5.2 and Figure 5.3. As what is shown in Figure 5.2, the S11 parameter of the RF part on base
board is lower than -10dB in most part from 4GHz to 7.5GHz, which is good. The S11 from 3GHz to
4GHz is between -5dB to -10dB and that at around 5GHz is about -8dB, which are fine and will need
to be improved.
dB
Hz
Figure 5.2 S11 Measurements for the RF Part of Base Board
dB
Hz
Figure 5.3 S11 Measurements for the RF Part of Tag
42
As for the measurement on tag (shown in Figure 5.3), S11 at frequencies from 3GHz to 7GHz is
mostly less than -10dB except that from 3.2GHz to 4.4GHz which gives about -6dB to -10dB return
loss. This part shall also be improved in future work.
5.2.2 Wireless Link Performance
Z-Stack is implemented on the CC2530 wireless link with one working as a coordinator and the other
as a router. A coordinator and a router are Zigbee device types which are specified in [45]. The router
should send a bind request to the coordinator to join in the network and the coordinator should admit
that. After a successful binding, the coordinator transmits packets with the following configurations:








Transmitting power:
Transmitting packet:
Transmission rate:
Packet payload:
Data rate:
MAC acknowledgement:
APS acknowledgement:
Re-transmission (in case of failure):
4 dBm
5000 pkt
10 pkt/s (in other words, 10 Hz)
20 Bytes
250 kbps
Enable
Enable
3 times
The coordinator transmits 5000 packets at the transmission rate of 10 packets per second and at the
data rate of 250 kbps. There are 20 Bytes in the payload of each packet. The transmitting power is set
to be 4 dBm. MAC acknowledgement and APS (Application Support Layer) acknowledgement are
both enable to guarantee the reliability of communication.
The coordinator will count the number of packets it has sent and the APS acknowledgements it has
received. The router will also count the number of packets it has received. The packet error rate (PER)
is defined as the ratio of the number of APS acknowledgements received by the source to the number
of packets sent by the source:
(5.1)
A set of measurements has been conducted to evaluate the performance of the CC2530 wireless link
based on Z-Stack. Scenarios such as line-of-sight and non-line-of-sight are all covered. The router was
put from 10 to 100 meters way from the coordinator, with 10 meters difference at each point. The
router as a receiver was put at a height of 20cm, while the height of the coordinator could differ. For
line-of-sight the coordinator was put at 30cm height and for non-line-of-sight it was put at 60cm.
5.2.2.1 Line-of-Sight
For line-of-sight scenario, the measurement was taken at a large lawn in front of Angantyrvägen 10
(Djursholm, Stockholm, Sweden), as shown in the following figure (captured from maps.google.com).
43
N
120m
Figure 5.4 Line-of-Sight Tests on the Performance of Wireless Link
As mentioned before, the coordinator was placed at 30cm height and the router at 20cm height. If they
were put at a height less than 20cm, packet lost raised put sharply due to ground fading. In this set of
measurements, the router was placed at specific distances from the coordinator. The packet lost at
different distances is shown in Table 5.1. From that we can see PER was 0% at the range up to 50
meters. For range from 60 to 100 meters PER still maintained within less than 0.5%.
The longest range of the wireless link measured was around 200 meters due to limited space for even
longer line-of-sight measurement. Yet a range of 200m is enough to fulfill the requirement of this
thesis work.
Distance
10m
20m
30m
40m
50m
60m
70m
80m
90m
100m
Table 5.1 Packet lost rate for Line-of-Sight
Packets sent by
APS ACKs received
Packets received
coordinator
by coordinator
by router
5000
5000
5000
5000
5000
5000
5000
5000
5000
5000
5000
5000
5000
5000
5000
5000
4983
5000
5000
4993
4998
5000
4997
4999
5000
4998
5000
5000
4992
5000
44
Packet Error Rate
0%
0%
0%
0%
0%
0.34%
0.14%
0.06%
0.04%
0.16%
5.2.2.2 Non-Line-of-Sight
A set of measurements were conducted around the house at Falks väg 18 (Djursholm, Stockholm,
Sweden) for non-line-of-sight scenario, as shown in the following figure (captured from
maps.google.com).
N
150m
Figure 5.5 Non-Line-of-Sight Tests on the Performance of Wireless Link
This set of measurements was conducted with house, trees, hedge and even rocks between the
coordinator and the router to block them from line-of-sight. Multipath effect is severe in this scenario.
The coordinator was placed at 60cm height and the router at 20cm height. If the coordinator was
placed at less than 50cm height, the packet loss raised sharply at range larger than 50 meters. So it is
recommended for the source to be placed at more than 50cm height to get better performance.
The PER at different distances is shown in Table 5.2 and it kept less than 1% up to 100 meters. The
longest range was about 150 meters for this NLOS scenario.
Note that the performance of the wireless link may vary at different points of the same range due to
multipath effect and fading.
Distance
10m
20m
30m
40m
50m
60m
70m
Table 5.2 Packet lost rate for Non-Line-of-Sight
Packets sent by
APS ACKs received
Packets received
coordinator
by coordinator
by router
5000
4979
4982
5000
5000
5000
5000
4999
5000
5000
5000
5000
5000
4996
5000
5000
4999
4999
5000
4997
4997
45
Packet Error Rate
0.42%
0%
0.02%
0%
0.08%
0.02%
0.06%
80m
90m
100m
5000
5000
5000
4969
4963
4995
4969
4964
4996
0.62%
0.74%
0.10%
5.2.3 UWB Transceiver Performance
A few tests have also been conducted to test the range of the UWB transceiver at the same place as
shown in Figure 5.5. The line-of-sight range was about 120m from the measurements. So the
positioning system is able to cover an open area of 100m × 100m.
The ranging accuracy of the UWB transceiver was ±15cm in TOA measurements. By averaging 5
samples at the each measuring point, the ranging accuracy reached ±10cm. The positioning accuracy is
believed to be able to reach ±20cm or less with good synchronization among the reference nodes.
46
Chapter 6
Conclusion and Future Work
6.1 Summary
The goal of this thesis is to design an indoor positioning system based on IEEE 802.15.4a. In this
thesis, different positioning technologies have been studied. The fundamentals of UWB technology
and the IEEE 802.15.4a standard have been discussed. We have also investigated different time-based
ranging protocols which are suitable for the implementation of positioning system based on UWB
technology.
Requirements on the positioning system in this design work have been specified. The design of the
system was started with the selection of the ranging protocol with respect to the requirements we set.
We have selected TDOA as the protocol because it provides good performance on positioning
accuracy, low channel occupancy, low power consumption and other advantages. The system setup, as
well as the architectures of different devices (coordinator, base station and tag) in this system, was
illustrated according to the chosen protocol. After that, the main components were selected and the
software architecture was also introduced.
Three PCBs have been designed for the proposed positioning system, including a base board, a
connection board and a tag board. All the components have been selected and the final block diagrams
of the three PCBs were illustrated. We went through the manufacturing and assembly. For the
hardware design this thesis work emphasized the design on the RF parts, including the microstripe and
grounding, to avoid transmission line effect. Besides, the thesis work has also covered parts of the
software design, focusing on the higher APP layer and the lower layers such as drivers and interfaces.
Finally, we verified the functionalities of different components on the three PCBs. The verification
results have shown they are all functional. We have also measured the performance of a few critical
parts in the proposed positioning system; they are the RF performance of the microstripe on the PCBs,
the range and packet error rate of the wireless link, and the range and ranging accuracy of the UWB
transceiver.
6.2 Conclusion
A positioning system based on IEEE 802.15.4a standard has been designed in this thesis work,
covering both hardware and software. With the ranging accuracy of ±15cm for the UWB transceiver,
the positioning accuracy is believed to be able to achieve ±20cm or less with good synchronization
among the reference nodes. It is able to cover a range of 100m × 100m for indoor positioning
applications.
47
TDOA is a good choice for the ranging protocol as it provides good performance on positioning
accuracy, low channel occupancy and low power consumption. By using this protocol, the system is
scalable to be used in large network such as wireless sensor network.
The design of RF parts, such as microstrip and grounding, highly influences the range and accuracy of
the positioning system. So it is important to optimize those parts to achieve good performance at the
targeting frequency bands.
The design of the software architecture eases a lot on the software development. It is easy to carry on
the development with a layered architecture.
However, the PHY and MAC layers of the UWB transceiver do not fully fulfill the requirements for
the APP layer, especially the synchronization part which is extremely important to the proposed
system. Therefore, measurements such as positioning accuracy, positioning latency, power
consumption, and motion tolerance cannot be performed.
6.3 Future Work
In order to get better performance of the proposed positioning system, there is some more work need
to be done in the future.
The RF part on the boards should be improved to get better performance. The PHY and MAC layers
of the UWB transceiver shall be refined to meet the targeting performance. A CSMA media access
protocol is preferred to avoid channel collisions. Synchronization protocol needs to be well designed.
With the help of all these, more statistic work shall be done in order to estimate the positioning
accuracy. Positioning latency, power consumption and motion tolerance can also be measured
afterwards. Then more RF measurements can be performed in order to check if it fulfills the
regulations or not.
The APP layer for all the reference nodes and tag shall be improved in order to enable all of them to be
powered with battery for a few years battery life time. With minor modifications to the system, it is
good to extend the usage of this system from indoor to outdoor as well.
More importantly, the positioning system in the future shall be able to track more than tens of
thousands of tags. A NET layer shall be introduced to the system to enhance the system to operate in a
wireless sensor network.
48
Reference
[1] E. D. Kaplan and C. Hegarty, Understanding GPS: Principles and Applications, Second Edition,
Artech House, 2006.
[2] F. V. Diggelen, A-GPS: Assisted GPS, GNSS, and SBAS, Artech House, 2009.
[3] H. Liu, H. Darabi, P. Banerjee and J. Liu, “Survey of Wireless Indoor Positioning Techniques and
Systems”, IEEE Transactions on Systems, Man, and Cybernetics - Part C: Applications, and Reviews,
vol. 37, NO. 6, Nov. 2007.
[4] P. A. Zandbergen and S. J. Barbeau, “Positional Accuracy of Assisted GPS Data from HighSensitivity GPS-enabled Mobile Phones”, The Journal of Navigation, vol. 64, pp. 381–399, July 2011.
[5] R. J. Fontana, “A brief history of UWB communications”, [Online], Available:
http://www.multispectral.com
[6] Federal Communications Commission, “First Report and Order”, 02-48, Feb. 2002, [Online],
Available: http://www.fcc.gov
[7] The Commission of the European Communities, “Commission Decision of 21 February 2007 on
Allowing the Use of the Radio Spectrum for Equipment Using Ultra-wideband Technology in a
Harmonised Manner in the Community”, Official Journal of the European Union, 2007/131/EC, Feb.
23, 2007, [Online], Available: http://eurlex.europa.eu/LexUriServ/site/en/oj/2007/l_055/l_05520070223en00330036.pdf
[8] R. Kohno, “Interpretation and Future Modification of Japanese Regulation for UWB”, IEEE
P802.15-06/261r0, May 16, 2006.
[9] IEEE Standard 802.15.4a-2007, “IEEE Standard for Information technology—Telecommunications and information exchange between systems—Local and metropolitan area networks—Specific
requirements Part 15.4: Wireless Medium Access Control (MAC) and Physical Layer (PHY)
Specifications for Low-Rate Wireless Personal Area Networks (WPANs) Amendment 1: Add
Alternate PHYs,” 2007, [Online], Available: http://book.itep.ru/depository/zigbee/802_15_4a2007.pdf
[10] M. Fischer, A. Ayse, A. L. Scholtz, and C. F. Mecklenbräuker, "Architecture of a Modular IEEE
802.15. 4a Ultra-Wideband Transmitter", In Microelectronics Conference ME2008, Vienna, Austria,
pp. 15-16. 2008.
[11] M. S. Kim, H. Kim, and J. T. Kim, "High-level modeling of UWB PHY for IEEE802.15.4a",
International Conference on Convergence and Hybrid Information Technology, 2008. ICHIT'08, pp.
351-354, IEEE, 2008.
[12] Y. Zheng, M. A. Arasu, K. W. Wong, Y. H. The, A. P. H. Suan, D. D. Tran, W. G. Yeoh, and D.
L. Kwong, "A 0.18μm CMOS 802.15.4a UWB Transceiver for Communication and Localization”, In
49
IEEE International Solid-State Circuits Conference, 2008, ISSCC 2008, Digest of Technical Papers,
pp. 118-600, Feb. 2008.
[13] J. Y. Oh, J. H. Park, M. K. Oh, J. Y. Kim, and H. S. Lee, "Baseband Transceiver Chip for IRUWB System", In International SoC Design Conference (ISOCC), 2009, pp. 564-568. IEEE, Nov.
2009.
[14] M. K. Oh, J. H. Park, and J. Y. Kim, "IR-UWB Packet-based Precise Ranging System for uHome Networks", IEEE Transactions on Consumer Electronics, vol. 55, pp. 119-125, Feb. 2009.
[15] A. Adalan, T. Gigl, M. Fischer, and K. Witrisal, "A Modular Impulse Radio Ultra-wideband
Research & Development Platform for IEEE802.15.4a", In Proceedings of the 2nd European Wireless
Technology Conference, 2009, pp. 116-119, Sep. 2009.
[16] D. Martynenko, G. Fischer, and O. Klymenko, "UWB Transmitter for Communication and
Localization for IEEE 802.15.4a Standard", In IEEE International Conference on Circuits and Systems,
2012, pp. 32-37, Oct. 2012.
[17] Z. Sahinoglu, S. Gezici, and I. Guvenc, Ultra-wideband Positioning Systems: Theoretical Limits,
Ranging Algorithms, and Protocols, Cambridge University Press, 2008.
[18] M. Z. Win and R. A. Scholtz, "Impulse Radio: How It Works", IEEE Communications Letters,
vol. 2, no. 1, pp. 10-12, Jan. 1998.
[19] The Commission of the European Communities, Commission Decision of 21 April 2009 on
Amending Decision 2007/131/EC on Allowing the Use of the Radio Spectrum for Equipment Using
Ultra-wideband Technology in a Harmonised Manner in the Community, Official Journal of the
European Union, 2009/343/EC, Apr. 21, 2009, [Online], Available: http://eurlex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:L:2009:105:0009:0013:EN:PDF
[20] The Commission of the European Communities, "Directive 1999/5/EC of the European
Parliament and of the Council of 9 March 1999 on Radio Equipment and Telecommunications
Terminal Equipment and the Mutual Recognition of Their Conformity", Official Journal of the
European Union, L 91/10-L 91/28, Apr. 07, 1999, [Online], Available: http://eurlex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:L:1999:091:0010:0010:EN:PDF
[21] The Swedish Post and Telecom Authority, “Post- och telestyrelsens författningssamling: Postoch telestyrelsens föreskrifter om undantag från tillståndsplikt för vissa radiosändare”, PTSFS 2010:8,
Dec. 20, 2010. [Online], Available: http://www.pts.se/upload/Foreskrifter/Radio/foreskrifterundantag_101209.pdf
[22] H. Karl and A. Willig, Protocols and Architectures for Wireless Sensor Networks, John Wiley &
Sons, Jul. 2007.
[23] S. Shepard, RFID: Radio Frequency Identification, McGraw-Hill, 2005.
[24] S. Farahani, ZigBee Wireless Networks and Transceivers, Newnes, 2008.
[25] CC2530, http://www.ti.com/product/cc2530, Texas Instruments.
[26] Z-Stack, http://www.ti.com/tool/z-stack, Texas Instruments.
50
[27] STM32F107VC, http://www.st.com/internet/mcu/product/221020.jsp, STMicroelectronics.
[28] MSP430F5438A, http://www.ti.com/product/msp430f5438A, Texas Instruments.
[29] IT3200C, http://www.rakon.com/products/families/tcxo#cat_2, Rakon.
[30] CDCLVC1102, http://www.ti.com/product/cdclvc1102, Texas Instruments.
[31] STBB1-AXX, http://www.st.com/web/catalog/sense_power/FM142/CL1456/SC1612/PF252511,
STMicroelectronics.
[32] TPS62111, http://www.ti.com/product/tps62111, Texas Instruments.
[33] DP83848K, http://www.ti.com/product/dp83848k, Texas Instruments.
[34] ST3232C, http://www.st.com/web/catalog/sense_power/FM1968/CL1179/SC392/PF66114,
STMicroelectronics.
[35] N. Abramson, “Development of the ALOHANET”, IEEE Transaction on Information Theory, vol.
31, NO. 2, pp. 119-123, Mar. 1985.
[36] AMD, Broadcom, National Semiconductor, and Texas Instruments, “RMII Specifications”, Mar.
1998, [Online], Available: http://sva.ti.com/assets/en/other/rmii_1_2.pdf
[37] E. Insam, TCP/IP Embedded Internet Applications, Newnes, 2003.
[38] AppCAD, http://www.hp.woodshot.com/, Agilent Technologies.
[39] S. H. Hall, G. W. Hall, and J. A. McCall, High-Speed Digital System Design: A Handbook of
Interconnect Theory and Design Practices, Wiley, 2000.
[40] B. S. Heyi, “Implementation of Indoor Positioning using IEEE 802.15.4a”, Master Thesis at KTH
the Royal Institute of Technology, Jan. 2013.
[41] A. Monge, “Synchronized MAC layer for Ultra-wideband Wireless Sensor Network – Design,
Implementation, Analysis and Evaluation”, Master Thesis at KTH the Royal Institute of Technology,
Mar. 2013.
[42] K. Yu, I. Sharp, and Y. J. Guo, Ground-based Wireless Positioning, Wiley, 2009.
[43] SmartRF05, http://www.ti.com/lit/ug/swru210a/swru210a.pdf, Texas Instruments.
[44] SmartRF Studio, http://www.ti.com/tool/smartrftm-studio, Texas Instruments.
[45] Zigbee Alliance, “Zigbee Specification”, Standard, Jan. 2008, [Online], Available:
http://www.zigbee.org
51
52
53
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertisement