Basic Research on UWB Antenna and Indoor Positioning Systems
ZHU, Hui September 2008
Basic Research on UWB Antenna and Indoor Positioning Systems
Ｚ Ｈ Ｕ ， Ｈ ｕ ｉ
G r a d u a t e S c h o o l o f I n f o r m a t i o n , P r o d u c t i o n a n d S y s t e m s Wa s e d a U n i v e r s i t y
S e p t e m b e r 2 0 0 8
Abstract
In recent years the ultra-wideband (UWB) technology has experienced significant developments. The origin of UWB technology stems from work in time-domain electromagnetics begun in 1962. UWB technology differs from conventional narrow band radio frequency technologies, such as Bluetooth and IEEE 802.11. It uses an extremely wide frequency band (3.1 - 10.6 GHz) to transmit data. It is able to transmit more data in a short time than the other narrow band technologies. In USA, according to the regulations released by Federal Communication Commission (FCC), UWB systems for indoor communications have a limited transmit power (-41.3dBm/MHz). Although the UWB technology has experienced many significant developments in recent years, there are still many exiting challengers in utilizing this novel technology to variable applications. Currently, the UWB technology has been induced into wireless communications and positioning systems.
Firstly, A UWB antenna should be capable of operating over an ultra-wide bandwidth with satisfactory radiation properties. The interference from the nearby wireless systems should be notched out. Secondly, the maintenance of pulse characteristics requires low distortion transmitter and receiver antennas over huge bandwidths. At last, extremely narrow pulses imply the difficulty in the processing of transmitted data. Pulse generation at transmitter and the demodulation at receiver may require a high performance circuit. One of the valuable aspects of UWB technology is the ability to determine the time that a radio signal transmits between the transmitter and receiver at various
frequencies. UWB technology provides an effective approach for the indoor positioning systems. For instance, the most famous positioning system is Global Positioning Systems (GPS). Signals from GPS satellites get attenuated and reflected by various objects and can not overcome position estimation errors due to these indoor effects. Therefore, a new positioning system is necessary for the indoor environments. Two basic technologies in the indoor positioning systems should be developed. One is the hardware for high accuracy distance measurement. The other technology is the position estimate algorithms based on the measured distance. This thesis makes the basic exploration on position estimate algorithms. Several effective algorithms are presented to achieve the coordinates of target. Considering above exiting challengers in UWB systems, the basic research on the UWB antenna and the algorithms for indoor positioning system are developed. This dissertation will consist of 5 chapters which are as following:
Chapter 1 [Introduction] This chapter briefly discusses the UWB technique and the requirement in the design of UWB positioning systems. The problems faced in the use of UWB technology are presented. The objective and scope of this project will be indicated in this chapter.
Chapter 2 [Ultra-wideband (UWB) Radio Antennas] A planar structure of the slot antenna with flat group delay for UWB radios is proposed in our research. By inserting the slot into the disc element, the band notched characteristic was also generated. The slot in the radiation part of the antenna is introduced to avoid interference between UWB and nearby communication systems. The length and width of the slot in the radiation part of the developed
antenna determines the notch frequency. The simulated results suggest a way to tune the band notched characteristic of the antenna. The notch frequency moves significantly to a higher frequency with decreasing smaller slot length. The notch frequency moves slightly to a lower frequency with increasing larger slot length. In this way, the notch frequency can be adjusted accurately to the desired value.
The measured notch frequency was 4.3 GHz, which is in good agreement with the simulated result. The proposed antenna covers the spectrum from 3.4 to 12 GHz and eliminates interference from nearby wireless systems. According to the simulated radiation patterns, nearly omnidirectional radiation patterns are observed in the H-plane. The proposed antenna also shows a high radiation efficiency of more than 92% from 3 to 11 GHz. The measured variation of group delay is 0.3 ns, which is smaller than that in the previous work. The variation of antenna gain is less than 3 dB for 3 to 10 GHz, excepting the notch frequency. In addition to the design flexibility, the simple configuration and easy fabrication make the proposed antenna suitable for integration into RF embedded systems.
Chapter 3 [UWB low noise amplifier] The design of UWB receivers has gained substantial significance due to the explosion of wireless applications.
Several low noise amplifiers (LNA) for UWB radio have been proposed in recent years. The resistive shunt-feedback based amplifiers have the limited input match at higher frequencies due to the parasitic input capacitance. The distributed amplifiers normally tend to consume large dc current due to the distribution of multiple amplifying stages. A wideband amplifier adopts a
bandpass filter at the input stage to achieve wideband input matching. The adoption of the filter leads to a larger chip area and noise figure degradation.
One objective in this thesis is to develop a CMOS LNA design technique suitable for UWB radio. This work aims to explore a circuit topology capable of low-voltage low-power operation. A new output load topology is proposed to improve the power gain while not decreasing the noise performance, which is quite different from the conventional shunt peak technique. Based on the new output load topology, the gain flat and good noise performance can be achieved while consuming small current. From the simulated results, the amplifier is operated in 1.2V power supply, a -10dB input match from 3 to 5 GHz, with 8.5 dB power gain while only consuming 4 mA current from power supply.
Chapter 4 [Position Estimation algorithms] This chapter discusses the algorithms that convert measured distances to the coordinates of target. Genetic Algorithm (GA) is the one of the most widely known paradigm in Evolutionary Algorithms. The structure of the GA is usually using three operators: selection, crossover and mutation. For the indoor positioning systems, the computation time of the algorithms is one the most important parameters. Particle Swarm Optimization (PSO) has been successful in solving a wide range of optimization problems. The PSO algorithm is easy to implement, has few parameters, and converge faster than the traditional algorithms. An improved PSO (PSO-RTVIWAC) is applied to solve optimization problems in the indoor positioning systems. A new parameter automation strategy, the time variable random inertia weight and acceleration coefficients are proposed in the PSO-RTVIWAC algorithm. The proposed strategy is adaptive to the indoor
positioning systems. The computation time could be reduced using the PSO-RTVIWAC algorithm. From the simulated results, it is indicated that, to achieve the same convergence, the particle number is reduced by 25% in the PSO-RTVIWAC algorithm. The computation time could be reduced by 40%, compared with conventional PSO algorithms. The optimal solutions can be quickly and effectively reached using PSO-RTVIWAC algorithm. The simulated results of positioning error as the function of the measured distance errors are demonstrated. In this chapter, the PSO algorithm has been introduced to the indoor positioning systems for the first time. Moreover, the PSO-RTVIWAC concept is proposed to reduce the computation time. Form above results, it is indicated that the PSO-RTVIWAC concept is appropriate for the hardware implement and it is helpful in the future development of the locator and tag.
Chapter 5 [Conclusion] In this research, firstly, a new strategy to tune the band notched characteristic of the UWB antenna is demonstrated. Secondly, a new output load topology is proposed to improve the power gain flatness and at the same time the noise figure is decreased, which is different from the conventional shunt peak technique. Finally, a new concept of PSO, which has the random time-vary inertia weight and acceleration coefficients, is introduced to solve the nonlinear optimization for the indoor positioning systems. These proposed methods should be considered as favorable contributions to the indoor positioning systems.
Table of Contents
Abstract...i
Table of Contents...vi
List of Figures...ix
List of Tables ... xiii
Chapter 1 Introduction...1
1.1 Background and motivation...1
1.1.1 Market Overview ...1
1.1.2 Typical Wireless technologies...4
1.1.3 Ultra-Wideband for wireless networking...7
1.2 Objectives and outline of the thesis ... 11
Chapter 2 Ultra-wideband (UWB) Radio Antennas...14
2.1 Antenna technologies...14
2.2 UWB antenna technologies...17
2.2.1 Microstrip line fed disc monopole ...20
2.2.2 Coplanar waveguide fed disc monopole ...21
2.3 The proposed antennas in this work...24
2.3.1 Introduction of the proposed antenna ...25
2.3.2 Characteristics in frequency domain...29
2.3.3 Radiation patterns and radiation efficiency...32
2.3.4 Antenna transfer function...35
2.4 Conclusion ...39
Chapter 3 UWB Low-noise Amplifier ...41
3.1 Introduction...41
3.2 Review of wideband amplifiers ...42
3.3 Two-port networks ...46
3.4 CMOS LNA fundamentals...49
3.5 Improved UWB LNA design ...52
3.6 Conclusion ...63
Chapter 4 Position Estimation algorithm ...65
4.1 Overview of existing algorithms...65
4.2 The PSO algorithms...71
4.3 The proposed PSO-RTVIWAC algorithm...75
4.3.1 Previous Work on the PSO Algorithm...80
4.3.2 The Proposed PSO-RTVIWAC Algorithm...87
4.3.3 Position Estimate using PSO-RTVIWAC...91
4.3.4 The Process of Position Estimate...94
4.3.5 Experimental Settings ...97
4.3.6 Experimental Results ...101
4.3.7 Exploration of Hardware Implement ...103
4.3.8 Summary...108
4.4 An artificial neural network based model...109
4.5 Conclusion ... 116
Chapter 5 Conclusion and future work... 119
5.1 Summary of results ... 119
5.2 Future work...122
Acknowledgments ...123
References ...124
Publications ...137
Academic journals ...137
International conferences...138
List of Figures
Figure 1.1: The digital environment: (a) Computers and peripherals (b)
Mobile CE and communications (c) Consumer electronics ...2
Figure 1.2: Indoor Positioning Systems ...4
Figure 1.3: Many different types of wireless technologies ...5
Figure 1.4: FCC regulation of UWB spectrum ...8
Figure 1.5: Band plan for OFDM UWB system...10
Figure 2.1: The microstrip line fed circular disc antenna...18
Figure 2.2: The simulated return loss of microstrip line fed circular disc antenna...19
Figure 2.3: The CPW fed circular disc antenna...21
Figure 2.4: Photograph of CPW fed circular disc antenna...23
Figure 2.5: The simulated and measured results of CPW fed circular disc antenna...23
Figure 2.6: UWB radio overlay nearby wireless operation frequency band ...25
Figure 2.7: Configuration of proposed antenna...27
Figure 2.8: Simulated return loss for (a) D3 of 8.0 mm, 8.4 mm, an/d 8.8 mm with a fixed value of D2 = 9.4 mm and (b) various D2 of 9.0 mm, 9.4 mm, and 9.8 mm with a fixed value of D3 = 8.8 mm ...29
Figure 2.9: Photograph of the proposed antennas ...30
Figure 2.10: Measured and simulated return loss of the proposed antenna (D1=14 mm, D2=9.4 mm, D3=8.8 mm, Di=0.6 mm, G1=0.2 mm, L1=21.7 mm, L=36 mm, W=26 mm, W1=1.6 mm, and W2=0.3 mm) compared with the antenna without a slot ...31
Figure 2.11: Simulated radiation pattern for various frequencies: (a) X-Y plane (b) Z-Y plane and (c) X-Z plane ...34
Figure 2.12: Simulated radiation efficiency ...35
Figure 2.13: Measured group delay when a pair of antennas are set face to face at the distance of 50 mm ...36
Figure 3.1: The common source amplifier with resistive feedback ...43
Figure 3.2: The common source amplifier with multiple LC sections...44
Figure 3.3: The distributed amplifier...46
Figure 3.4: The two port network...46
Figure 3.5: (a) Noisy two port (b) input referred noise model ...48
Figure 3.6: Small signal model of MOSFET...49
Figure 3.7: Induced gate noise model...51
Figure 3.8: (a) the simple common gate amplifier with shunt peak load (b) output load equivalent small signal model (c) the common gate amplifier with proposed load topology and its (d) output load equivalent small signal model ...53
Figure 3.9: power gain of shunt peak design for various Rd...54
Figure 3.10: Noise figure of shunt peak design for various Rd...55
Figure 3.11: (a) Simplified schematic of proposed UWB LNA (b) simplified small signal model...56
Figure 3.12: Power gain of proposed design for various Rd...58
Figure 3.13: Noise figure of proposed design for various Rd ...59
Figure 3.14: The simplified model of on-chip inductor ...61
Figure 4.1: The positioning systems using wireless signals...66
Figure 4.2: Genetic Algorithm flowchart ...69
Figure 4.3: (a) The general positioning under the idea environment and (b) positioning under the noisy environment ...79
Figure 4.4: (a) The inertia weight and (b) the acceleration coefficients in the PSO-TVIW ...82
Figure 4.5: (a) The inertia weight and (b) the acceleration coefficients in the PSO-TVAC ...83
Figure 4.6: (a) The inertia weight and (b) the acceleration coefficients in the PSO-RANDIW ...84
Figure 4.7: (a) The inertia weight and (b) the acceleration coefficients in the PSO-RTVIWAC...88
Figure 4.8: General idea of position estimated in PSO-RTVIWAC algorithm: (a) distance measured in TOA technique, (b) position estimated in PSO-RTVIWAC and (c) estimated results and error ...93 Figure 4.9: Estimation process in PSO-RTVIWAC: (a) system initialization,
(b) estimation process in PSO-RTVIWAC and (c) estimated results and error...96 Figure 4.10: The architecture of PSO-RTVIWAC concept ...105 Figure 4.11: The proposed three layer neural network... 110 Figure 4.12: Average estimation error Ep,average as the function of locator
number... 112 Figure 4.13: Average estimation error Ep,average as the function of measured
error ... 113 Figure 4.14: Average estimation error Ep,average as the function of side
length ... 114
List of Tables
Table 1.1: Proposed UWB band in the worldエラー! ブックマークが定 義されていません。
Table 1.1: Proposed UWB band in the world...7 Table 4.1: The average positioning error Ep,average produced with 20
iterations in a dimension of 50m × 50 m...99 Table 4.2: The average positioning error Ep,average produced with 50
iterations in a dimension of 50 m × 50 m...100
Table 1.1: Proposed UWB band in the world...7 Table 4.1: The average positioning error Ep,average produced with 20
iterations in a dimension of 50m × 50 m...99 Table 4.2: The average positioning error Ep,average produced with 50
iterations in a dimension of 50 m × 50 m...100
Chapter 1 Introduction
1.1 Background and motivation
1.1.1 Market Overview
Computers and peripherals
As demonstrated in Fig.1.1 (a), it is common for an office to have a computer, printer, scanner and other peripherals, in addition to several laptops and external hard drives. As the number of devices in an office grows, so does the number of wires connecting these peripherals. It is always fond that a user is confused by a large number of cables surrounding his office desk.
Mobile consumer electronics
Wireless connectivity enables a novel mobile lifestyle filled with conveniences for mobile users. As illustrated in Fig. 1.1 (b), consumers demand the same conveniences throughout their digital environments, such as connecting their digital cameras, personal digital assistants and cell phones to each other in a wireless personal area network. But the exiting wireless technologies cannot meet the needs of connectivity since some of mobile consumer electronic devices require high bandwidth.
Figure 1.1: The digital environment: (a) Computers and peripherals (b) Mobile CE and communications (c) Consumer electronics
Consumer electronics in the house
The digital house environment is made up of personal computing devices, different consumer electronic devices and mobile devices. The consumer electronics devices, in the home, are driving a need for robust and high speed wireless networks. There is a strong case for short range wireless networks, which operate over a few feet and provide data rates in the hundreds of Mega Bits per Second.
An entertainment room in a home will consists of a cable receiver, a digital video recorder and a DVD player. As shown in Fig.1.1 (c). All of these need to be connected to a digital television and to an audio/video receiver via cables.
Devices such as digital cameras and portable music players frequently require huge data transfer. It is discouraging to see a user’s entertainment center cluttered with a large number of cables.
Indoor Positioning Systems
The Global Positioning System (GPS) systems help us navigate while driving, hiking and flying. GPS uses several satellites to locate a target and obtain the timing. It is only able to triangulate the location of a target if it is outdoors. If the target is indoors, the GPS signals will not be able to pass through the wall. Therefore, the accuracy of the GPS systems cannot satisfied the requirement of indoor positioning. As demonstrated in Fig. 1.2, the short range higher speed wireless technique is a good candidate for the high accuracy indoor positioning systems.
From the above discussion, it is very easy to see how a short range high speed wireless network could ease our life. Seamless connectivity is a compelling proposition to the typical consumer. Devices, which are not only capable of positioning, discovering and communicating with each other automatically, but also capable of printing or playing under command without user’s intervention as well, provide value in their simplicities.
Figure 1.2: Indoor Positioning Systems
1.1.2 Typical Wireless technologies
There are many different types of wireless technologies for communication and positioning, as illustrated in Fig. 1.3. Bluetooth consume small amount of power and is viewed as a better solution fit in Consumer electronics. As time progresses, new systems will come up that have faster transferring rates and handle larger data capacity to meet the increasing demand for bandwidth. The choice of these techniques depends on the range, transferring speed, power consumption and implementation cost. For example, IEEE 802.11 has a good communication range and can be used in most wireless networking situation.
But it consumes a significant amount of power.
Figure 1.3: Many different types of wireless technologies
Bluetooth
Bluetooth wireless technology is a short-range communications technology intended to replace the cables connecting portable and fixed electronic devices [1]. The key features of Bluetooth technology are robustness, low power, and low cost. It was designed to allow low bandwidth wireless connections seamlessly integrated into our daily life.
Bluetooth technology provides a way to exchange information between devices such as cell phones, computers and video player over a secure, short range radio frequency. It operates in the vicinity of 2.4 GHz and gains the support of many famous companies. Bluetooth offers many attractive
characteristics such as support up to 8 slave devices in a piconet and non line-of-sight transmission. It primarily is designed for low cost and low power consumption, with a short range of 10 m when having a data rate of several Mbps.
IEEE 802.11
There are two available spectrum bands for IEEE 802.11 standards.
802.11b and 802.11g use the 2.4 GHz ISM band, operating in the United States under Part 15 of the US Federal Communications Commission Rules and Regulations [2]. Because of this choice of frequency band, 802.11b and g equipment may occasionally suffer interference from other systems such as microwave ovens. 802.11a is much faster than 802.11b at 54Mbps but has somewhat less range because of operation in the higher frequency 5GHz band.
These communication protocols were designed for relatively long range communication and are very power hungry.
Ultra-wideband (UWB)
UWB is a relatively new technology. A few companies have been developing products with UWB technology. UWB signals have a relative large spectral bandwidth. Along with other unique abilities such as able to penetrate through walls and have high security, UWB is going to find its way in the indoor environments for communication and positioning. The characteristic of UWB signals will be discussed in the following sections in order to have a more comprehensive view on this exiting technology.
Table 1.1: Proposed UWB band in the world
Region UWB band
United States 3.1-10.6 GHz
Europe
Low band: 4.2-4.8 GHz High band: 6-8.5 GHz
Japan
Low band: 3.4-4.8 GHz High band: 7.25-!0.25 GHz
1.1.3 Ultra-Wideband for wireless networking
UWB differs from conventional narrow band radio frequency technologies.
UWB uses an extremely wide frequency band to transmit data. It is able to transmit more data in a short time than other narrow band technologies. The requirement of a UWB system is that the UWB system needs to occupy at least 500 MHz bandwidth. The utilization of UWB technique should not interfere with the existing nearby wireless systems. UWB technology has been around since the 1960’s, when it was known as Time-Domain Electromagnetics. In most UWB applications the system operates by sending short pulses. This provides an effective approach in positioning. Now it is proposed to utilize this technique in the wireless communications and indoor positioning systems. Since the transmit power is very low, these systems are inherently adaptive to low power, short range and high speed applications.
Figure 1.4: FCC regulation of UWB spectrum
Application of UWB techniques
UWB technology offers a solution for the bandwidth, power consumption, and physical size requirements of next generation consumer electronic devices.
UWB technique enables wireless connectivity with high data rates across electronic devices within the digital environment. This technology provides the high bandwidth that required by the multiple digital video and audio streams.
With the support of industry workgroups, UWB technology promises to make it easy to create high speed communication and positioning that can connect devices throughout the home and office. UWB technology can enable wide range of applications. It can replace cables between multimedia devices, such as camcorders, digital cameras, and portable MP3 players, with wireless
connectivity. The short range high speed wireless USB connectivity including printers, scanners, and external storage devices is available if utilizing UWB technique.
As shown in Table 1.1, In USA, according to the regulations released by Federal Communication Commission (FCC), UWB systems for indoor communications have been allocated to the spectrum from 3.1 to 10.6 GHz at a limited transmit power (-41.3dBm/MHz) [3]. The European Commission only uses part of the spectrum that was approved in the USA. The UWB systems have been allocated to the spectrum from 6.0 to 8.5 GHz at a limited transmit power (-41.3dBm/MHz). It is also applied provisionally in the spectrum from 4.2 to 4.8 GHz until the end of 2010 [4]. In Japan, lower frequency band from 3.4 to 4.8 GHz and higher frequency band from 7.25 to 10.25 GHz are suggested to allocated UWB systems at a limited transmit power (-41.3dBm/MHz). In the spectrum from 4.8 to 7.25 GHz signals are transmitted at limited power (-70dBm/MHz) [5].
Regulation of UWB systems
As indicated in Fig. 1.4, part 15 of the FCC rules provides the definitions for UWB operation. The UWB bandwidth is the frequency band bounded by the points that are 10 dB below the highest radiated emission, as based on the complete transmission system including the antenna. UWB has a fractional bandwidth equal to or greater than 0.20 or has a bandwidth equal to or greater than 500 MHz.
Figure 1.5: Band plan for OFDM UWB system
OFDM Approach to UWB
UWB Orthogonal Frequency Division Multiplexing (OFDM) systems have been proposed as an emerging solution to wireless communication applications requiring high data rates over short distances [6]. In the Multi-Band Orthogonal Frequency Division Multiplexing (MB-OFDM) version, a 528 MHz wide OFDM signal is created from 128 subcarriers, with a subcarrier spacing of 4.125 MHz, as shown in Fig. 1.5. Because the UWB emission regulations limit the power per megahertz, a larger utilized bandwidth will result in a larger total emitted average power. This OFDM approach uses 528-MHz channels, but uses at least three at a time in a time-frequency hopping manner, thus emitting the total power allowed in 1,584 MHz. Frequency hopping is used in this OFDM
approach as a way of having this system occupy a large total bandwidth for the purpose of increasing the total radiated power.
DS-UWB Approach to UWB
Direct Sequence Spread Spectrum (DS-SS) has roots in secure and military communications systems [7]. Users of the spectrum could simultaneously occupy a large common channel and be separated by digital codes rather than by the wavelength of frequency. This is precisely the premise behind the DS-UWB approach to a UWB system intended for the UWB spectrum. It is found that direct-sequence approach designed to occupy at least 1.5 GHz bandwidth in the 3.1 to 5.15 GHz range and 3.7 GHz bandwidth in the 5.8 to 10.6-GHz range.
Unlike a traditional carrier-based DS-SS system, this UWB approach uses nonsinusoidal wavelets tailored to occupy the desired spectrum in an efficient manner.
1.2 Objectives and outline of the thesis
The thesis will concentrate on the introduction and description of our researches on UWB technology. In recent years the UWB technology has experienced significant developments. But there are still many challengers in utilizing this novel technology to variable applications. One particular challenge is the UWB antenna. There is a growing demand for small and low cost UWB antennas that can provide satisfactory performances in both frequency domain and time domain.
In chapter 2 the circular disc monopole antenna is investigated in detail in
order to understand its transfer function that leads to the UWB characteristic.
Some quantitative guidelines for designing of this type of antenna are also obtained. This chapter will illustrate the concepts of the band notched antenna with flat group delay characteristic. A planar structure of the slot antenna with flat group delay for UWB radios was proposed. By inserting the slot into the disc element, the band notched characteristic was also generated. The notch frequency can be easily and accurately determined by adjusting the slot length and slot space. According to simulated radiation patterns, nearly omni-directional radiation patterns are observed in the H-plane. The proposed antenna also shows the high radiation efficiency. The measured variation of group delay and antenna gain is very small in lower and higher frequency band, which is quite improved from previous work. In addition to the design flexibility, the simple configuration and easy fabrication make the proposed antenna to be integrated into RF embedded systems.
The design of UWB system has gained substantial significance due to the explosion of wireless applications. The first active block in most wireless receivers is the low-noise amplifier (LNA). The LNA needs to amplify the signal without adding a large amount of noise and distortion while consuming minimal power. The objective in chapter 3 is to develop a UWB CMOS LNA design technique suitable for the prevalent inductively degenerated LNA architecture. This work aims to explore the tradeoffs between LNA performance and power consumption and find a circuit topology capable of low-voltage low-power operation. The proposed design demonstrates a high performance design with on-chip low-Q passive components in a deep submicron technology.
Particle Swarm Optimization (PSO) has been successful in solving a wide range of optimization problems. The PSO algorithm is easy to implement, has few parameters, and has been shown to converge faster than traditional techniques for a wide variety of benchmark optimization problems. In past several years, PSO has been successfully applied in many research and application areas. In chapter 4 a novel PSO algorithm is applied to solve nonlinear optimization in the indoor positioning systems. As a new parameter automation strategy for the PSO concept, the random time-varying inertia weight and acceleration coefficients (PSO-RTVIWAC) are proposed. The proposed strategy is adaptive to the indoor positioning systems. The computation time is reduced using the PSO-RTVIWAC algorithm. The general idea of position estimated in PSO-RTVIWAC is demonstrated and the process of utilizing PSO-RTVIWAC to resolve nonlinear optimization in TOA technique is explained in detail. PSO-RTVIWAC algorithm has merits of easy to implement, few parameters. The computation time and complexity is smaller, compared with the conventional PSO algorithms. Simulated results are discussed and compared with other positioning algorithms.
Finally, chapter 5 will present the conclusions on our researches and some of the future work will be demonstrated in this chapter.
Chapter 2 Ultra-wideband (UWB) Radio Antennas
2.1 Antenna technologies
The growing use of wireless devices has forced the need of larger bandwidths. As the result using large frequency bands have been proposed in recent years. The UWB technique has several attracted characteristic since it can realize a bit rate up to Gbps. The Antenna is a challenging part of UWB technology. It needs to be specifically designed and optimized for the wideband applications.
For conventional narrow band systems, antennas are designed only to radiate over relatively narrow frequency band, which is no more than about 25
% of the center frequency. If an impulse is fed to such narrow band antennas, it tends to ring and distort the transmitted signals and spread them out in time.
A variety of UWB antennas exist for many years. The log periodic antenna and spiral antennas are typical examples. These antennas are unsuitable for short pulse applications. These antennas transmit different frequency components from different parts of the antenna. This will distort and stretch out the radiated waveform. In addition, for UWB applications, there is a great need for flat group delay and antenna gain, omni directional, high radiation efficient, low cost, easy manufacture and integration with RF circuits.
The objective of this chapter is to design a CPW-fed UWB antenna that is
capable of operating in the 3.1 GHz to 10.6 GHz bandwidth. As mentioned above, the purpose of this research is to simulate, fabricate and verify the simulation results of a CPW-fed antenna for UWB applications. In this project, the commercially available ADS simulation tool was used. This tool is capable of plotting return loss values in the desired operating frequency range as well as generating radiation patterns for selected frequencies. In addition, radiation patterns would also be simulated. The design was fabricated and tested to compare the computer simulated results with the results obtained in a laboratory environment to conclude on its suitability for UWB applications.
Definition of antenna
The antennas are a critical part of any wireless systems. As the definition in the IEEE Standard, an antenna is a means for radiating or receiving radio waves.
A transmit antenna is a device that receive the signals from a transmission line, converts them into electromagnetic waves and then broadcasts them into free space. In receiver stage, the antenna collects the incident electromagnetic waves and converts them back into signals.
In the current wireless system, an antenna is usually required to optimize for the radiation energy in some frequencies and suppress it in others certain frequencies. Thus the antenna may also serve as a directional in addition to a transition device. In order to meet the particular requirement, it must take various forms. A good design of the antenna can relax system requirements and improve overall system performance.
Return Loss
The return loss is a logarithmic ratio measured in dB that compares the
power reflected by the antenna to the power that is fed into the antenna from the transmission line. The basic requirement of this antenna is that it should exhibit a return loss of less than –10 dB in the operating frequency range, which indicates that at least 90 percent of the input power is transmitted through the proposed antenna. This value corresponds to a Voltage to Standing Wave Ratio (VSWR) of approximately 2:1.
Bandwidth
The bandwidth of the antenna refers to the range of the frequencies over which the antenna can operate quickly. Mathematically, the bandwidth of the antenna is defined as:
⎟⎟⎠
⎜⎜ ⎞
⎝
⎛ −
×
=
c L H
f f
BW 100 f (2.1)
where fH is the highest frequency, fL is the lowest frequency and fc is the centre frequency.
Directivity and Gain
The directivity of the antenna is the ability of an antenna to focus energy in a particular direction during transmitting, or to receive energy better from a particular direction when receiving. The gain of an antenna is defined as the amount of energy radiated in a particular direction compared to the amount of energy directed in the same direction of an isotropic antenna.
Radiation Pattern
The radiation or antenna pattern describes the relative strength of the radiated field in various directions from the antenna, at a fixed distance.
Measured radiation patterns are slices of the three dimensional radiation pattern taken in the horizontal and vertical planes.
2.2 UWB antenna technologies
Requirement of UWB antenna
An antenna plays a crucial role in UWB systems, which is same in conventional wireless communication systems. There are more new challenges in designing a UWB antenna than a narrow band one.
Firstly, what a UWB antenna is significantly different from other antennas is its ultra wide frequency bandwidth. According to the FCC's definition, a UWB antenna should be able to yield a bandwidth no less than 500MHz or a fractional bandwidth of at least 20%. A variety of resonant antenna technologies have been presented. Parameters such as directivity, gain, and polarization often received more attention than the bandwidth extension. A fractional bandwidth of a few percents the impedance-bandwidth performance of antennas is enough in the narrow band systems. However, with today’s UWB requirement the antenna bandwidth can be a bottle neck of a wireless communication system.
Secondly, the performance of a UWB antenna is required to be consistent over the entire operational band. The antenna radiation patterns, gains and impedance matching should be stable across the entire band. It is also demanded that the UWB antenna provides the band notch characteristic to coexist with nearby narrow band systems occupying the same operational band.
Figure 2.1: The microstrip line fed circular disc antenna
A UWB antenna is required to achieve good time domain characteristics.
For the narrow band case, it is approximated that an antenna has same performance over the entire bandwidth and the basic parameters, such as return loss, group delay and antenna gain, have little variation across the operational band. In contrast, UWB systems often employ extremely short pulses for data transmission. The antenna imposes more significant impacts on the input signal.
As a result, a good time domain performance, i.e. minimum pulse distortion in the received waveform, is a primary concern of a suitable UWB antenna because the signal is the carrier of useful information.
Figure 2.2: The simulated return loss of microstrip line fed circular disc antenna
Thirdly, directional or omni-directional radiation properties are needed depending on the practical applications. For radar systems where high gain is desired, directional radiation characteristics are preferred. Omni-directional patterns are normally desirable in mobile systems.
Fourthly, a good design of UWB antenna should be optimal for the performance of overall system. For example, the antenna should be designed such that the overall device (antenna and RF front end) complies with the mandatory power emission mask given by the FCC or other regulatory bodies. A suitable antenna needs to be small enough to be integrated into mobile and
portable devices. It is also highly desirable that the antenna feature low profile and compatibility for integration with printed circuit board (PCB).
2.2.1 Microstrip line fed disc monopole
As illustrated in Fig. 2.1, a planar circular disc monopole can also be realized by using microstrip feed line [20]. L and W denote the length and the width of the substrate. The circular disc monopole with a radius of R and a 50 ohm microstrip feed line are printed on the same side of the FR4 substrate (thickness of 1.6 mm and a relative permittivity of 4.4). The width of the microstrip feed line g is fixed at 2 mm to achieve 50 ohm impedance. h is the feed gap between the feed point and the ground plane. The conducting ground plane with a length of H=16 mm only covers the section of the microstrip feed line, which is printed on the other side of the substrate. In the Fig. 2.1, the microstrip line fed disc monopole has R=8 mm, h=1.0 mm, W=16 mm and L=33 mm. The simulated return loss is plotted in Fig. 2.2. It is observed in Fig.
2.2 that the simulated -10dB bandwidth ranges from 3 GHz to 12 GHz.
Figure 2.3: The CPW fed circular disc antenna
2.2.2 Coplanar waveguide fed disc monopole
As depicted in Fig. 2.3, the CPW fed disc monopole antenna has a single layer-metallic structure. W and L denote the width and the length of the ground plane, respectively. The first important parameter is the feed gap h. The performances of the CPW fed disc monopole are quite sensitive to h [19]. It is found that the return loss curves have similar shape for the different feed gaps, but the -10dB bandwidth of the circuit disc antenna varies significantly with h.
If h becomes bigger, the -10dB bandwidth is getting narrower due to the fact that the impedance matching of the antenna is getting worse. It is also found that a bigger gap doesn't affect the resonant frequency very much. The optimal feed gap is found to be at h=0.3mm, which is close to the CPW line gap. It makes perfect sense that the optimal feed gap should have a smooth transition to the CPW feed line. Another important parameter influencing the antenna operation is the width of the ground plane W [19]. It can be seen that the variation of the ground plane width shifts the resonant frequency. The -10 dB bandwidth is reduced when the width of the ground is narrow. When the ground plane width is either reduced, the current flow on the edge of the ground plane would be reduced. This corresponds to a decrease the inductance of the antenna if it is treated as a resonating circuit, which causes the first resonant frequency moves to up end. Therefore, the change of the ground plane width makes some resonances become not so closely spaced across the spectrum and reduces the overlapping between them. Thus, the impedance matching becomes worse in these frequency ranges. Figure 2.4 shows the photograph of CPW fed circular disc antenna, and Fig. 2.5 shows the simulated and measured results. A circular disc antenna with a 50 ohm CPW is printed on the same side of a dielectric substrate. Wf is the width of the metal strip and g is the gap of distance between the strip and the ground plane. h is the gap between the disc and the ground plane. In this study, a dielectric substrate (a thickness of 0.8 mm and a relative permittivity of 4.4) is chosen. Wf and g are fixed at 1.6 mm and 0.3 mm to achieve 50 ohm input impedance. The CPW-fed disc monopole has R=7 mm, h=0.3 mm, W=26 mm, H=21.7 mm and L=36 mm.
Figure 2.4: Photograph of CPW fed circular disc antenna
-30 -20 -10 0
2 4 6 8 10 12
frequency (GHz)
return loss (dB)
measured resul t si mul ated resul t
Figure 2.5: The simulated and measured results of CPW fed circular disc antenna
2.3 The proposed antennas in this work
In our research, a novel band notched antenna with flat group delay is proposed for UWB radios. The proposed antenna covers the spectrum from 3.4 to 12 GHz with the notch frequency at 4.3 GHz. The simulated result is in very good agreement with the measured data of antenna return loss, particularly notch frequency, as a result of tuning the slot for band rejection. The interference between UWB and nearby wireless systems will be alleviated.
According to simulated radiation patterns, nearly omnidirectional radiation patterns are observed in the H-plane. The proposed antenna also shows a high radiation efficiency of more than 92% from 3 to 11 GHz. Furthermore, by recovering transmitted pulses properly for DS-UWB systems, the variation of the group delay and antenna gain owing to the antenna transfer function are designed to be minimized and flat.
UWB systems may cause electromagnetic interference on nearby wireless communication systems [8]. An example is the wireless local-area network (WLAN: frequency band adopted by IEEE 802.11a is from 5.15 to 5.825 GHz).
As shown in Fig. 2.6, interference can occur between UWB regulated by the FCC and nearby wireless systems such as WLAN. A UWB antenna with a band notched characteristic is suitable for solving this problem. We introduced a slot in the radiation part of the antenna to avoid interference between UWB and nearby communication systems. The length of the slot in the radiation part of the developed antenna determines the notch frequency. The notch frequency can be easily tuned to the desirable frequency band by adjusting the parameters of the slot.
frequency (GHz)
PSD (d B m /M H z)
-41.3
3 3.1
5 - 6 GHz (WLAN)
10.6 11 4.2-4.8 GHz
(EC commision untill 2010)
3.1-10.6 GHz (UWB)
Figure 2.6: UWB radio overlay nearby wireless operation frequency band
2.3.1 Introduction of the proposed antenna
Various UWB antennas and wideband antennas have been introduced [9]-[18]. The circular disc CPW-fed antenna can yield an ultrawide 10 dB return loss bandwidth with satisfactory radiation patterns [19]. This antenna has been proved to be highly suitable for UWB applications. However, the variation of the group delay and antenna gain, which represents the degree of pulse distortion, was not discussed in [19]. The nearby wireless signals also should be
eliminated to decrease the interference to UWB radios. In this work, small variation of the group delay and antenna gain is achieved, which satisfies the requirement of UWB communication systems.
The requirements imposed by UWB systems place stringent demands on UWB antenna design. The impedance-matching property is the primary requirement. A UWB antenna should be capable of operating over an ultrawide bandwidth with satisfactory radiation properties. The undesired frequencies should be notched out. For DS-UWB systems, the variation of group delay and antenna gain should be minimized and flat for achieving a good impulse response with minimal distortion.
Compared with the microstrip antenna [20], the CPW-fed antenna has many attractive features, such as low radiation loss, less dispersion at high frequencies, easy integration with RF integrated circuits and a simplified configuration with a single metallic layer. Etching on only one side of the substrate eliminates the alignment problem that arises in etching on both sides of the substrate in a microstrip antenna, and makes the CPW-fed antenna easy to fabricate.
Figure 2.7 shows the geometry of the proposed CPW antenna. It consists of a disc radiation element and a slot in the disc. The slot length (the bold line shown in Fig.2.7) which is close to perimeter of the smallest circle, determines the notch frequency. By adjusting the diameter of the smallest circle D3, the slot between two circles is adjusted and the notch frequency will be moved to the desired frequency.
ground ground Radiation part
x y
L
W
L1 D1 D2
D3
W1 G1
W2
Radiationpart
slot
Slot length
z Di
Figure 2.7: Configuration of proposed antenna
The proposed antenna has a total volume of 36*26*0.8 mm^3. The dimensions of the designed antenna are substrate length L=36 mm, substrate width W=26 mm, ground-plane length L1=21.7 mm, distance between ground plane and radiation part G1=0.2 mm, D1=14 mm, and slot space Di=D2-D3.
-60 -50 -40 -30 -20 -10 0
3 5 7 9
frequency (GHz)
re
11
turn l oass (dB )
D3=8.8mm Di=0.6mm D3=8.4mm Di=1.0mm D3=8.0mm Di=1.4mm
4.4GHz 4.9GHz 4.7GHz resonant
frequency f1
resonant frequency f2 notch frequency
(a)
-50 -40 -30 -20 -10 0
3 5 7 9
frequency (GHz)
re turn lo ass
11
(dB )
D2=9.0mm Di=0.2mm D2=9.4mm Di=0.6mm D2=9.8mm Di=1.0mm resonant
frequency f1
resonant frequency f3 notch frequency
4.3GHz 4.4GHz
4.5GHz
(b)
Figure 2.8: Simulated return loss for (a) D3 of 8.0 mm, 8.4 mm, an/d 8.8 mm with a fixed value of D2 = 9.4 mm and (b) various D2 of 9.0 mm, 9.4 mm,
and 9.8 mm with a fixed value of D3 = 8.8 mm
The proposed antenna is printed on a dielectric substrate FR4 (thickness 0.8 mm, relative permittivity 4.7 and loss tangent 0.02). The width of the center strip of the CPW-fed line W1 is 1.6 mm, and the gap between the center strip of the CPW-fed line and the ground plane W2 is 0.3 mm. With these dimensions, the characteristic impedance of the CPW-fed line is 50 ohm.
2.3.2 Characteristics in frequency domain
The proposed antenna is simulated by the momentum in ADS (Advanced Design Systems, Agilent Tech. Inc) [21], the simulator based on the method of moments (MoM) technology that is particularly efficient for analyzing planar conductor and resistor geometries.
Figures 2.8 (a) and (b) illustrate the simulated return loss for both D3 lengths (8.0 mm to 8.8 mm) with D2 fixed and for D2 lengths of 9.0 mm to 9.8 mm with D3 fixed. All the other dimensions are the same as in the previous definition, except slot space Di. In Fig. 2.8 (a) it is obvious that notch frequency shifts significantly from 4.4 to 4.9 GHz with decreasing D3. Di increases from 0.6 to 1.4 mm. Resonant frequencies f1 and f2 in Fig. 8 (a) also move slightly towards higher frequency. In Fig. 2.8 (b) the notch frequency shifts slightly from 4.5 to 4.3 GHz with increasing D2. Di increases from 0.2 to 1.0 mm in this case. Resonant frequencies f1 and f3 shift to a lower frequency in Fig. 2.8 (b).
Figure 2.9: Photograph of the proposed antennas
The above simulated results suggest a way to tune the band notched characteristic of the antenna. The notch frequency is dominantly affected by D3.
The notch frequency moves significantly to a higher frequency with decreasing D3 (slot space Di increases) since the slot length is related to D3. D2 has some influence on the notch frequency. The notch frequency moves slightly to a lower frequency with increasing D2 (slot space Di also increases in this case). In this way, the balanced combination of D3 and D2 helps to adjust the notch frequency to the desired value.
-40 -30 -20 -10 0
2 4 6 8 10
frequency (GHz)
return loss (dB)
12
antenna with slot (measured) antenna with slot (simulated) antenna without slot (measured) 4.3GHz
4.4GHz 3.5GHz
3.7GHz
7.9GHz
8.9GHz
Figure 2.10: Measured and simulated return loss of the proposed antenna (D1=14 mm, D2=9.4 mm, D3=8.8 mm, Di=0.6 mm, G1=0.2 mm, L1=21.7 mm, L=36 mm, W=26 mm, W1=1.6 mm, and W2=0.3 mm) compared with
the antenna without a slot
The photograph of proposed antenna is shown in Fig. 2.9. The measured return loss illustrated in Fig. 2.10 shows the effect of the band notch and impedance matching over the UWB bandwidth. Compared with antenna without a slot, the lower frequency limit of the proposed antenna is extended to 3.4
GHz.
The bandwidth enhancement of 2.7 GHz is achieved. Compared with the disc antenna, it is concluded that the proposed antenna achieved the band notched characteristic and desired bandwidth with -10 dB return loss. The limit of low frequency is also extended. By inserting the slot, more surface current can be distributed along the edge of radiation part than in the reference antenna.
As shown in Fig. 2.10, the measured slot antenna shows the 3.4-12 GHz bandwidth with -10 dB return loss. The measured notch frequency (4.3 GHz) shows good agreement with the simulated result (4.4 GHz). The small difference of 0.1 GHz is due to the fabrication error of D3 and D2. Although it is proved that the characteristic curve notches out the spectrum from 4.2 to 4.8 GHz, a further design for tuning the slot length and slot space is expected to enable particular application.
2.3.3 Radiation patterns and radiation efficiency
The simulated radiation patterns of the proposed antenna in the X-Y (E-plane), Z-Y (E-plane) and X-Z (H-plane) plane are plotted in Figs. 2.11 (a)-(c), respectively. Dipole-like radiation patterns are shown in the X-Y and Y-Z planes. The radiation pattern demonstrates that the power gain of the antenna at the notch frequency is lower than the power gain at other frequencies.
Nearly omnidirectional radiation patterns are observed in the X-Z plane.
Particularly, it is noted that the proposed antenna can be used to transmit in the X-Z plane. For practical UWB application, the antenna may be placed on the X-Z plane to avoid null point.
(a) X-Y plane
(b) Z-Y plane
(c) X-Z plane
Figure 2.11: Simulated radiation pattern for various frequencies: (a) X-Y plane (b) Z-Y plane and (c) X-Z plane
The simulated radiation efficiency of the proposed antenna is demonstrated in Fig. 2.12. The radiation efficiency of an antenna is the ratio of the total power radiated by the antenna to the net power accepted by the antenna at its terminals during the radiation process [22]. The radiation efficiency achieved in the previous design [23] was more than 90%. The proposed antenna shows that the radiation efficiency is more than 92% from 3 GHz to 11GHz. In particular, the radiation efficiency at the notch band does not drop regardless of the transmission loss, i.e., impedance mismatch. It is significant to note that the radiation efficiency of the proposed antenna can provide sufficient radiation energy from 3.1 to 10.6 GHz.
0. 7 0. 75 0. 8 0. 85 0. 9 0. 95 1
3 5 7 9 11
f requency (GHz)
ra di at io n ef fi ci en cy
notch f requency 97%-99%
92%-97%
Figure 2.12: Simulated radiation efficiency
2.3.4 Antenna transfer function
In the DS-UWB technique, the information is transformed into impulse signals in extremely short time duration. This signal has rapidly time-varying performance. The pulse distortion due to group delay variation can cause serious system performance degradation.
-2.00 -1.75 -1.50 -1.25 -1.00 -0.75 -0.50 -0.250.000.250.500.751.001.251.501.752.00
3 5 7 9 11
frequency (GHz)
group delay (ns)
0.3ns
notch frequency
0.3ns
Figure 2.13: Measured group delay when a pair of antennas are set face to face at the distance of 50 mm
For DS-UWB communications, flat group delay is desired to minimize the distortion of receiving signals. Figure 2.13 shows the measured group delay between a pair of proposed antennas set face to face at a distance of 50 mm.
Considering the measured values in the range of UWB communication from 3.1 to 10.6 GHz, the variation of group delay is 0.3 ns in the lower frequency band from 3 to 4 GHz. In the higher frequency band from 5 to 11 GHz, the group delay variation is also 0.3 ns. The group delay is changed in the vicinity of the notch frequency. Compared with [20], the variation of group delay is significantly improved by 0.7 ns in the proposed antenna. Although it is still
insufficient to satisfy the requirement of DS-UWB systems, further improvement will be carried out in the future.
The slot notches out some signals to avoid interference from nearby wireless communication systems. According to S21 (measured transfer function between pair of proposed antennas), the signal at the notch frequency band is about 10 dB lower than signals at other frequencies. Thus each proposed antenna has a 5 dB dip. This result helps to understand the effect of the band notched characteristic in UWB antenna design. According to Friis' transmission formula [24], antenna gain is given by
2
4 ⎟
⎠
⎜ ⎞
⎝
= ⎛ G R P G
P
t r t
r π
λ (2.2)
if Gt=Gr, then
S21
P P
t
r = (2.3)
⎟⎠
⎜ ⎞
⎝
= ⎛
λ πR P G P
t t r
4 (2.4)
where Pr is the power received by the antenna, Pt is the power input to the transmitting antenna, S21 can be regarded as Pr/Pt in terms of the S-parameter, is the wavelength, R is the separation between transmitter and receiver antennas, and Gt and Gr are the antenna gains of the transmitting and receiving antennas, respectively. In fact, the antenna gain Gt (or Gr) shown in Fig. 2.14 includes radiation loss and impedance mismatch. It is noted that the antenna gain shown in Fig. 2.16 is not its maximum value.
-20 -15 -10 -5 0 5
3 4 5 6 7 8 9 10 11
frequency (GHz)
antenna gain (dB)
2.5dB 2.5dB
notch frequency
Fig. 2.14: Measured antenna gain when a pair of antennas are set face to face at the distance of 50 mm
A flat antenna gain is desirable for decreasing the pulse distortion and ringing, since pulse distortion can cause serious system performance degradation. Figure 2.16 shows the measured antenna gain when a pair of antennas is set face to face at the distance of 50 mm. It is found that antenna gain drops in the notch band. The variation of antenna gain is 2.5 dB in lower frequency band from 3 to 4 GHz. In the higher frequency band from 5 to 10 GHz, the antenna gain variation is 2.5 dB. Both small variation of group delay and flatness of antenna gain are assumed to play an important role in
minimizing the distortion of pulses in DS-UWB.
2.4 Conclusion
Antennas are a challenging part of UWB technology. Conventional antennas are designed only to radiate over relatively narrow range of frequencies used in conventional narrow band systems. If an impulse is fed to such an antenna it tends to ring, severely distorting the pulse and spreading it out in time. Normal wideband antennas, including the log periodic antenna and spiral antennas, will not transmit short impulse since both these antennas transmit different frequency components from different parts of the antenna that distorts and stretches out the radiated waveform. The circular disc monopole antenna has attracted considerable research interest due to its simple structure and UWB characteristics with nearly omnidirectional radiation patterns.
However, it is still not clear the group delay and antenna gain of this kind antenna. For a UWB antenna minimizing both frequency and spatial dispersion should be considerate carefully.
A planar structure of the slot antenna with flat group delay for UWB radios is proposed in our research. By inserting the slot into the disc element, the band notched characteristic was also generated. The slot in the radiation part of the antenna is introduced to avoid interference between UWB and nearby communication systems. The length and width of the slot in the radiation part of the developed antenna determines the notch frequency. The simulated results suggest a way to tune the band notched characteristic of the antenna. The notch
slot length. The notch frequency moves slightly to a lower frequency with increasing larger slot length. In this way, the notch frequency can be adjusted accurately to the desired value.
The measured notch frequency was 4.3 GHz, which is in good agreement with the simulated result. The notch frequency can be easily and accurately determined by adjusting the slot length and slot space. The proposed antenna covers the spectrum from 3.4 to 12 GHz and eliminates interference from nearby wireless systems. According to the simulated radiation patterns, nearly omnidirectional radiation patterns are observed in the H-plane. The proposed antenna also shows a high radiation efficiency of more than 92% from 3 to 11 GHz. The measured variation of group delay is 0.3 ns in the lower and higher frequency bands, which is better than that in the previous work. The variation of antenna gain is less than 3 dB for 3 to 10 GHz, excepting the notch frequency.
In addition to the design flexibility, the simple configuration and easy fabrication make the proposed antenna suitable for integration into RF embedded systems.
Chapter 3 UWB Low-noise Amplifier
3.1 Introduction
Most communication protocols in existence today are narrow-band. For example, Code Division Multiple Axis (CDMA) has a bandwidth of 1.2 MHz.
Wireless broadband access protocols such as 802.11 occupy 20 MHz of frequency spectrum. Most of receivers existing today are towards narrow band operation. For example, a heterodyne receiver may use a quadrature Local Oscillator (LO) for image rejection. It could be hard to be realized in UWB systems. Standard direct conversion receivers have a time varying DC offset, which can be detrimental to the performance of a UWB system. These standard receiver architectures need to be examined very carefully to verify their viability in a UWB system.
To meet the objective of a widely deployed high-speed wireless network, an UWB system needs to be cheap, low power and reliable. Meeting all these objectives at the same time is not an easy task. For example, a designer could increase linearity in the receiver to make the design robust. But this increases the power consumption of the system. To reduce the power consumption in the front-end, a designer may add extra degeneration inductance in the low noise amplifier. But this would increase die area and, therefore, the cost of the system.
A very careful trade-off needs to be executed for an optimum design.
3.2 Review of wideband amplifiers
The most popular technique for designing a Low Noise Amplifier (LNA) is inductively degenerate the LNA. This technique uses a series inductor for input impedance matching and provides a LC resonant tank for the output load and output impedance matching.
This technique is effective for the narrowband LNA. However, it is not suitable for the UWB LNA. Input matching with a series inductor makes the LNA highly frequency selective. Inductive degeneration would make the gain of the amplifier frequency dependent. The new design approaches are necessary for the challenges in UWB systems.
The most popular amplifiers that can simultaneously provide power gain and 50 ohm input and output impedance match over a wide bandwidth are the common source amplifier resistor terminated, shunt-feedback amplifier and common gate amplifier. Almost all the published wideband LNA are based on one of these configurations [25]-[30].
In contrast to the narrowband LNA design that has converged to the inductor-degenerated common-source topology, how to design a low-power wideband amplifier with up to 10 GHz of bandwidth using existing CMOS technologies is still a question mark. The low supply voltage constraint imposed on the design also exacerbates the issue. This is why there have been many UWB LNA publications with good performance using BJT technology [31]-[36], but rarely in CMOS. UWB CMOS LNAs are either too power hungry [37], with limited bandwidth [38], or has inferior gain/noise performance [39]-[40].
Figure 3.1: The common source amplifier with resistive feedback
Broadband LNA with resistive input and output match
As shown in Fig. 3.1 Broadband resistive LNA provide an input match through RC feedback and output loading with a resistor. Resistive LNA occupy little die area because they only consist of a few transistors, passive resistors and capacitors. The resistive feedback amplifiers have limited input match at higher frequencies due to the parasitic input capacitance, and tend to suffer from poor noise figure and limited gain flat. They do have the following drawbacks: firstly, Broadband resistive LNA has poor high frequency performance. Although broadband matching is good in a resistive LNA, it is difficult to provide good high frequency response. The gain will therefore be severely degraded near high frequency. Techniques such as multiple resistive feedbacks have been employed.
However, such techniques come at the expensive of higher power consumption and poor noise figure.
Figure 3.2: The common source amplifier with multiple LC sections
Secondly, the noise figure and linearity is poor in Broadband resistive LNA.
A resistive input and output matched LNA is the best solution in terms of die area, but this will result in a significantly increased current consumption and high noise figure. To achieve the linearity requirements, the current has to be increased significantly.
Broadband LNA with multiple LC sections
As shown in Fig. 3.2, a new topology of a wideband amplifier for UWB system, which adopts a bandpass filter at the input of the cascode low noise amplifier for wideband input matching was proposed. By embedding the input network of the amplifying device in a multisection reactive network so that the overall input reactance is resonated over a wider bandwidth. The bandpass filter-based topology incorporates the input impedance of the cascode amplifier
as a part of the filter, and shows wide band impedance match and good gain flat while dissipating small amounts of power. The adoption of the filter at the input mandates a number of reactive elements, which lead to a larger chip area and noise figure degradation in the case of on-chip implementation, or the additional external components.
Multiple LC section LNA can provide very good broadband matching, high gain and low Noise Figure. However, multiple LC sections result in a very large die area. Multiple LC section LNA uses a ladder matching network with three on-chip inductors, which results in a large die area.
Multiple LC section LNA uses multiple LC sections for input impedance match and a shunt peaking load, which results in five on-chip inductors. A multiple LC match would require high Q inductors, which might require a thick top metal, adding additional cost to the chip.
Distributed LNA
As shown in Fig. 3.3, the distributed amplifiers normally provide wide bandwidth characteristics with good noise figure but tend to consume large dc current due to the distribution of multiple amplifying stages, which makes them unsuitable for low power application. Distributed amplifiers can provide very good wideband gain and linearity. Power consumption of distributed amplifiers is high and they occupy a relatively large die area. Their applicability as a low noise amplifier for a low-cost low-power system is therefore very limited.
Figure 3.3: The distributed amplifier
Figure 3.4: The two port network
3.3 Two-port networks
It is useful to abstract circuit blocks into two-port networks to simplify the design of analog circuits. As shown in Fig. 3.4, there are usually two quantities:
voltage and current associated to each port.
Linearity is a key requirement in the design of an LNA. The LNA must be able to maintain linear operation when the input is driven by a large signal.
Large-signal linearity characterized by the input-referred 1dB voltage compression point (P1dB) is important. It determines the maximum input level
that can be amplified linearly, which is the upper bound of the dynamic range of the LNA.
In order to design a circuit for low noise, it is helpful to determine that under which condition the noise can be minimized. For the analysis of noise in two-port systems, consider a noisy two port network driven by a noisy source as shown in Fig. 3.5 (a). The noise factor of a two-port network is expressed as:
source n
output n out s in
P P SNR F SNR
,
= ,
=
(3.1)
where Pn,output is the noise power outputted by the two-port and Pn,source is the noise presented at the input of the two-port. An ideal noiseless two port network contributes no noise. Then the noise factor is equal to one. Noise figure NF, which is noise factor expressed in decibel, often used to specify noise performance and has an ideally value of 0 dB.
To simplify analysis, the noise of a two-port network can be modeled as a noise voltage and a noise current at the input as shown in Fig. 3.5 (b). The signal source is represented by a current source is in parallel with and an admittance Ys to simplify derivation. The noise factor can then be expressed in terms of Fmin and the source admittances by:
( ) ( )
[
, 2]
2 ,
min s sopr s sopr
s
s n G G B B
G F R
F = + − + − (3.2)
Figure 3.5: (a) Noisy two port (b) input referred noise model
The above analysis shows that the source impedance optimized for a minimum noise factor exists, but this source impedance is often not the same as the impedance that achieves maximum power transfer. The ratio Rn/Gs appears as a multiplier in front of the second term. For a fixed source conductance, Rn
represents the sensitivity of the noise factor as Gs and Bs departs from their optimal values. A large Rn implies a high sensitivity, which obligates the design to stay close to optimal noise matching. As discussed subsequently, operation at low bias currents is associated with large Rn, mainly due to small device transconductance gm. This is an example of the difficulty in achieving high performance at low power consumption.
