30-08-2014, 01:27 PM
EMERGING COMMUNICATION TECHNOLOGIES
EMERGING COMMUNICATION.pdf (Size: 688.88 KB / Downloads: 25)
INTRODUCTION
Ultra Wideband (UWB) wireless technology is the prime candidate for becoming the next
step in the evolution of wireless technology. It is potentially well suited for use wherever
high-speed data rates (to at least several hundred Mb/s) are desired over ranges up to
several hundred meters in locations prone to fading due to multi-path propagation. This
emerging wireless technology uses short duration pulses known as monocycles to
propagate signals over physical distances instead of the sinusoidal carriers used by legacy
wireless systems.1
Two major UWB wireless technology application areas exist today, addressing
communications and radar needs, respectively. This report largely focuses on UWB
communication applications since UWB radar applications will likely not see widespread
use within the communication networks of future Spaceport and Range.2
Whether occupied all in one band, or sub-banded into 5 to 15 sub-bands, fundamental
UWB communication concepts in use today all derive from simpler pulse-based systems
first used in radar systems. The modulation waveforms currently used in UWB systems
today have not changed significantly since their first use over 30 years ago in radar
systems. As a result, UWB wireless systems often retain many traditional radar
capabilities, even when intended solely for communication purposes.
This characteristic capability of UWB technology is expressed by stating that UWB
systems are position-aware; that is, receiving UWB modulated signals requires an
inherent, automatic assessment of relative distances among the transmitters and receivers
within a UWB wireless network. Coupling communications with position-aware features
simultaneously enables wireless systems based on UWB to provide capabilities that were
never previously possible in traditional wireless communication systems.
In spite of occupying very large bandwidths, UWB is often found to be extremely benign
to existing wireless systems and services. The use of ultra wide bandwidths also has
advantages relative to narrower bandwidths. Since the correlation bandwidth of the
dense urban and dense structure propagation channel is typically less than 10 MHz over
3.1 GHz to 10.6 GHz, the use of extremely short-duration bursts achieves ultra wideband
occupancy over much greater than the correlation bandwidth of the channel and this
completely mitigates the effects of destructive interference (i.e., fading) in multi-path
signals.3
Because of this advantage, a high fidelity UWB replacement for FM tactical
radios would completely avoid much of the fading so commonly heard when operating in
1
Moe Z. Win and Robert A. Scholtz, "Ultra-wide Bandwidth Time-Hopping Spread-Spectrum Impulse
Radio for Wireless Multiple-Access Communications", IEEE Trans. Comm. Vol. 48, No. 4, April 2000. 2
UWB radar functions will still likely play a critical role in enhancing security around future Spaceports
and Ranges; they just will not play any significant role within the communication networks. 3
Correlation bandwidth refers to the bandwidth over which a spectral null is typically correlated and all
frequencies fade simultaneously. It is the bandwidth over which a fade exists in, for example, an urban
channel. Any signal within this bandwidth is simultaneously lost during fading events, and the fade is said
to be ‘correlated’ over this range of frequencies.
UWB REGULATORY AND TECHNOLOGY OVERVIEW
1.1.1 Regulatory Overview
Current UWB applications typically use one of two fundamental types of modulations:
Time-Hopping (TH) Pulse Position Modulation (PPM) or Bi-phase Pulse Modulation. By
current FCC Part 15 rules adopted February 14, 2002, a total of 7500 MHz of unlicensed
spectrum is available for UWB communication over 3.1 to 10.6 GHz.4
The present UWB
communication rules specify neither the exact modulation or waveform shapes that must
be used; instead, only the maximum effective isotropic radiated power (EIRP) levels
(-41.3 dBm/MHz), the maximum permitted frequency spectrum allocation (3.1 GHz to
10.6 GHz, for emissions above a maximum spectral mask limit of 10 dB down from the
peak radiated emission of the complete system, including the antenna), his laissez faire approach
sets the minimum characteristics necessary to encourage the peaceful co-existence of and additional
usage specifications (indoors, ac power only) are established.
Technology Overview
UWB communication systems use very low power (Part 15 levels are 5 mW or less),
unlicensed, very short duration (< 2 ns, typically 10 to 1000 ps) UWB pulses at repetition
rates from 10 to 40 MHz. Centered at a typical center frequency of 2 GHzgeneration UWB typical system occupied 1.4 GHz. To avoid interfering with GPS
signals and other low-power signals below 2 GHz, newer UWB systems, in compliance
with current Part 15 UWB requirements, now occupy 3.1 to 10.6 GHz, either in one band,
or within several sub-bands.
Because the pulses are pseudo-randomly (PN) shifted in time, transmitted signals
resemble white noise to narrowband, conventional receivers. Because of their wideband,
low-power characteristic, UWB systems typically co-exist with existing narrowband
communication systems, without causing significant interference. Likewise, because of
their high processing gains of 30 dB or better due to occupying wide bandwidths, noise
rejection performance of UWB systems is superior to that seen in narrowband systems.
UWB DESCRIPTION AND VENDORS
1.2.1 Description
The history of UWB dates to the earliest days of radio, and to even before radio was
called radio, back to when radio was first called wireless.
10 Recent advances in digital
processing have made it possible to re-think the fundamental trades long used for
implementing radios, allowing improvements over the trades when analog circuits were
the sole means by which to fashion communication system building blocks. With a fresh
re-thinking of communication system implementations arising with UWB technology, it
becomes possible to gain significant advantages over previous communication systems
implementations, while simultaneously reducing implementation complexity, physical
volume, and power consumption.
How is this re-thinking of implementation details, long established by practice, possible?
It is possible because UWB communication is simply traditional radio or wireless
technology with a different choice of ranked importance of the variables than what has
traditionally been chosen. Specifically, UWB communication systems trade pulse
shortness, thereby gaining high peak powers, in exchange for two other variables:
1.) Bandwidth (the needs of which are increased in UWB due to the short
duration of the pulses), and
2.) Signal to noise ratios of individual pulses (which are decreased in UWB,
thereby requiring correlation to combine coherent pulse energies
coherently, thereby gaining an advantage over noise powers that only can
combine non-coherently, being uncorrelated.)
Some refer to UWB communication as impulse radio. Others see it as simply being
traditional radar modulation used for communication purposes. Both viewpoints are
technically correct.
With a re-thinking of the rules that have governed radio design for so long, UWB
technology enables new communication systems to be created with higher performance
levels than have ever before been possible.
BASIC UWB THEORY
The following introduces UWB theory starting with the simplest monocycle
representation that incorporates all the fundamentals necessary for understanding basic
UWB communication principles. Then, additional levels of detail are added as necessary
for building on these principles for introducing more esoteric UWB concepts. The
general approach chosen is to start with the representation of a monocycle seen at the
output of a receive antenna, and to base all the correlation calculations on this most
commonly used representation of a received monocycle. As UWB theory is expanded,
different correlation templates are derived.
The preliminary introduction, in turn, is followed by a discussion of higher levels of
complexity in the monocycle waveform itself, through examining the monocycle
waveform (1) as it is produced as a Gaussian current pulse, (2) as it is transmitted from
the transmitter antenna, (3) as it is received through the receive antenna, and (4) as it
becomes a current pulse that is processed by the receiver.
Simplified Monocycle Introduction
Traditional wireless radio transmissions have utilized sinusoidal waveforms since the
1920’s for a variety of reasons. Perhaps the most compelling reason has been that
sinusoidal waveforms are very amenable to mathematical modeling. Another reason is
that, because of this ease of analysis, sinusoidal waveforms also make the analytical task
easier for reducing occupied communication system bandwidths to near the minimum
Nyquist-limit bandwidths required for such transmissions, thereby increasing spectral
occupancy efficiency and permitting more transmitters to occupy the airwaves without
causing one another harmful interference.
1.3.6 Finding a UWB Correlation Template through Cross Correlation
The first correlation template function used previously is but one such template function
that can be used successfully to detect UWB monocycles. It is also possible to generate a
successful correlation template function through first taking a derivative of a monocycle
time function and then applying a cross correlation technique.
1.3.10 Pulse Position Modulating UWB Monocycles
With the theory developed thus far, it becomes possible to model monocycles as well as
the Time Hopped Pulse Position Modulation (TH PPM) of the same monocycles. Recall
that Time Hopping/Hopped (TH) PPM is one of the two major modulations applied to
UWB communication systems.35
In TH PPM UWB modulation, multiple monocycles are used to transmit each bit in each
symbol. The monocycles that comprise each bit, which in turn comprise each bit symbol,
must individually be detected, and soft bit decisions must be made, prior to making a
final bit symbol decision. Consider a UWB TH PPM modulation scheme whereby a
monocycle occurring at exactly the repetition period is defined to be a “ZERO” and a
delayed monocycle, delayed by δ (the value of which is chosen to minimize ISI, that is,
Inter-Symbol Interference), is defined to be a “ONE”. Assume a period of 3, i.e., T = 3,
for the following example.
A clandestine UWB monocycle is definitely present, and has been detected equally well
as if by a UWB receiver having had the good fortune of knowing in advance all of the
parameters of the transmitted clandestine UWB monocycle.
Without assuming any particular parametric values for the clandestine UWB monocycle,
it is therefore possible to detect the presence of a monocycle through using CWT
techniques, by transforming a sampled signal with a set of scaled wavelets spanning
multiple octaves. This is important, for with conventional Fourier analysis spectrum
analyzer techniques, it is difficult to detect the presence of a UWB monocycle even when
knowing that it is present. [Based on measurements in the lab, a maximum detection
distance of only a few inches was the limit at which a spectrum analyzer with an external
Low Noise Amplifier and antenna could detect a UWB transmission from a UWB
transmitter antenna.] With a Continuous Wavelet Transform technique, it is thus possible
to increase the detection distance to a value approaching the maximum communication
range of the UWB link designed a priori with complete knowledge of the monocycle
waveform chosen.
A standard figure of merit for clandestine transmitters can be defined in terms of the ratio
of maximum communication range to the maximum detection range, assuming no details
are known about the signal. For example, if a specific UWB transmitter has a range of
150 feet when communicating with a receiver designed based on known UWB
monocycle parameters, and the maximum detection distance is only 1/4 foot for the
uncooperative case, then the Figure of Merit for Covertness would be 150/(0.25), or 600.
With a CWT technique, it would be possible to reduce this Figure of Merit of Covertness
to perhaps 1.5, for an improvement of over 52 dB versus a prior-art, non-CWT technique.
With this much improvement in detection, a CWT-detection technique has considerable
value for ferreting out clandestine UWB transmitters. In all likelihood, the detection
bubble would likely be 2/3 or more of the communication range bubble, greatly reducing
the volume (and area) over which a security sweep would need to be conducted to
guarantee detection of a covert UWB transmitter with a high probability of intercept
(POI). The only significant impediment with this technique is that the CWT hardware
would need to employ a wide-range of wavelet widths, and the maximum sample rate
would need to be at least twice the maximum frequency occupied by the UWB signal to
guarantee finding the unknown UWB monocycle transmission. The detection equipment
would also be slightly more complex than the typical UWB receiver designed with a
priori knowledge of the monocycle particulars. Still, this would not be unreasonable,
since it is likely that only a small number of such detection units would be required for
sweeping areas securely, while guaranteeing that covert UWB transmitters were not in
operation.
1.3.11 Maximizing Disorder for a Fixed Standard Deviation
As introduced earlier, the presence of noise can greatly diminish the robustness of a nonoptimal communication link. Likewise, the modeling of noise can greatly influence the
apparent robustness of a communication link, especially if the noise that is modeled is not
representative of the actual noise properties that will be encountered. The need for
finding the ‘noisiest noise’ is therefore a real one, if the modeling is to reflect the actual
performance likely to ensue with particular correlator architectures.
Finding the probability density function that maximizes the uncertainty or measure of
disorder (i.e., the entropy) for a one-dimensional probability distribution function (PDF),
P(x), with a fixed standard deviation, σ, is important for selecting the noisiest noise
model for simulating random variables. (The following derivation largely follows the
same method used by Claude Shannon in his original analysis of entropy for
communication systems.39)
1.3.12 Using Normal Distributions to Model and Explore Noise and Path Loss
There are two key advantages to using a Normal Distribution for modeling natural
processes. First, as derived previously, Normal Distributions model natural processes
well since they maximize entropy, which natural processes tend to maximize. Second,
Normal Distributions are easy to use, since it is possible to define probability density
functions (PDFs) through defining just two parameters, namely the mean, μ, and the
standard deviation, σ, or equivalently the entropy, H, in place of the standard deviation.
Knowing just two values, therefore, anyone can reproduce an entire PDF, with all of its
properties.
1.3.13 Detecting Monocycles in a Noisy, Variable Attenuation Channel
This same technique can be applied to the problem of detecting monocycles in a noisy,
variable attenuation channel. For this, assume that a Standard Deviation for Gaussian
Noise of 0.25 is used; that is, σ = 0.25. Likewise, assume gain constants of Vg = 0.3 and
Mg = 0.62 (both chosen through iteratively trying various values in the following
equations until the resulting noise matched experimental expectations of real noise.)
Next, randomize the mean of the noise to implement a highly variable channel
attenuation, even within a monocycle, when simulated by double sideband (DSB)
modulating a monocycle waveform, p(t), with noise. (This variability just approximates
the spectral correlation nulls that occur over a few MHz of bandwidth at various points
within the ultra wide bandwidths occupied by the UWB monocycle.)
1.3.16 Back to the Future: Damped Sinusoidal Pulses vs. Monocycles
During the earliest days of radio, all transmitters were broadband emitters. Prior to about
1918, the only limit to the bandwidth of a transmitter was often the operating bandwidth
of the transmitting antenna (i.e., the aerial) itself.48 Broadband sparkgap transmitters
based on this earliest sparkgap technology ruled the airwaves prior to 1918, and more
advanced methods to restrict transmitter emissions (transmissions) to a narrower range of
frequencies did not come into widespread use until CW (Continuous Wave) transmitters
became dominant around 1920-1921. From about 1918 until 1921, during the last years
of “King Spark” as the modulation was often called, limiting the emissions of sparkgap
transmitters through using damped sinusoids (also known as resonant sparkgap
transmissions) was tried in an attempt to save sparkgap equipment producing companies.
These “narrowband” sparkgap transmissions consisted of keyed pulses of decaying
sinusoids used in place of keyed pulses of equal amplitude sinusoids. A keyed damped
sinusoid consisted of a waveform, kds(t), of the form
A clandestine UWB monocycle is definitely present, and has been detected equally well
as if by a UWB receiver having had the good fortune of knowing in advance all of the
parameters of the transmitted clandestine UWB monocycle.
Without assuming any particular parametric values for the clandestine UWB monocycle,
it is therefore possible to detect the presence of a monocycle through using CWT
techniques, by transforming a sampled signal with a set of scaled wavelets spanning
multiple octaves. This is important, for with conventional Fourier analysis spectrum
analyzer techniques, it is difficult to detect the presence of a UWB monocycle even when
knowing that it is present. [Based on measurements in the lab, a maximum detection
distance of only a few inches was the limit at which a spectrum analyzer with an external
Low Noise Amplifier and antenna could detect a UWB transmission from a UWB
transmitter antenna.] With a Continuous Wavelet Transform technique, it is thus possible
to increase the detection distance to a value approaching the maximum communication
range of the UWB link designed a priori with complete knowledge of the monocycle
waveform chosen.
A standard figure of merit for clandestine transmitters can be defined in terms of the ratio
of maximum communication range to the maximum detection range, assuming no details
are known about the signal. For example, if a specific UWB transmitter has a range of
150 feet when communicating with a receiver designed based on known UWB
monocycle parameters, and the maximum detection distance is only 1/4 foot for the
uncooperative case, then the Figure of Merit for Covertness would be 150/(0.25), or 600.
With a CWT technique, it would be possible to reduce this Figure of Merit of Covertness
to perhaps 1.5, for an improvement of over 52 dB versus a prior-art, non-CWT technique.
With this much improvement in detection, a CWT-detection technique has considerable
value for ferreting out clandestine UWB transmitters. In all likelihood, the detection
bubble would likely be 2/3 or more of the communication range bubble, greatly reducing
the volume (and area) over which a security sweep would need to be conducted to
guarantee detection of a covert UWB transmitter with a high probability of intercept
(POI). The only significant impediment with this technique is that the CWT hardware
would need to employ a wide-range of wavelet widths, and the maximum sample rate
would need to be at least twice the maximum frequency occupied by the UWB signal to
guarantee finding the unknown UWB monocycle transmission. The detection equipment
would also be slightly more complex than the typical UWB receiver designed with a
priori knowledge of the monocycle particulars. Still, this would not be unreasonable,
since it is likely that only a small number of such detection units would be required for
sweeping areas securely, while guaranteeing that covert UWB transmitters were not in
operation.
1.3.18 Third-Order Introduction to Monocycles
Classic electromagnetics theory historically has always been applied to designing
antennas while tacitly assuming steady-state responses. The reason for making this
simplifying assumption is that it greatly simplifies the application of Maxwell's Equations
for designing antennas and simultaneously permits using simplified Electromagnetics
Theory, thereby avoiding the discontinuous anomalies that exist during the startup of the
waveform (i.e., antenna capacitance charging effects.)
For UWB antennas, there is no shortcut that can be used to avoid the discontinuous
events at the start of the pulse waveform, for this is all there is. Steady-state electrical
conditions are never reached in UWB antennas.
1.4 TESTING DESCRIPTION
Volume II of this Final Report previously described the detailed testing description for
the tests conducted on the UWB hardware investigated on this project. For expediency,
these descriptions are not repeated here.
1.5 TEST OBJECTIVES
The fundamental objectives for the UWB hardware testing conducted on this project were
the interference profiles both from and to the tested UWB hardware relative to
conventional wireless (radio) equipment. In order to understand these results, the
theoretical results documented earlier in this, Volume III of this Final Report, were
derived to enable establishing a firm theoretical understanding of the limits of UWB
technology.
1.6 TEST SETUP
Volume II of this Final Report previously described the detailed testing setups used for
the tests conducted on the UWB hardware investigated on this project. For expediency,
these descriptions are not repeated here.
1.7 TEST EQUIPMENT AND EVALUATION KIT
1.7.1 Test Equipment
Volume II of this Final Report previously described the test equipment used during
testing UWB hardware on this project. For expediency, these descriptions are not
repeated here.
1.7.2 Evaluation Kit (EVK)
Taking advantage of one of the first commercial products available, an early pair of
Evaluation Kit (EVK) TM-UWB radios was purchased and received early in January
2003 from Time-Domain Corporation, of Huntsville, AL. This EVK, consisting of a pair
of UWB transmitter/receiver radios with Ethernet link interfaces, along with controlling
software for use on a laptop, was the UWB exemplar tested on this project
PROPOSED UWB FOLLOW-ON RESEARCH ACTIVITIES
The focus of this project has been on UWB communication and UWB communication
theory for application within future communication networks on Spaceports and Ranges.
This work culminated in a New Technology Report describing a method to detect both
cooperative and non-cooperative UWB transmitters, based on the CWT theory described
earlier in this report.
The next step in understanding and benefiting from UWB technology is to extend this
UWB research into through-wall and ground penetration radar applications, and
especially into non-destructive inspection of non-metallic composites, to investigate the
theoretical limits of UWB technology as applied to these allied areas.
RESEARCH CONTRIBUTORS
3.1 BIOGRAPHICAL THUMBNAIL SKETCHES
Gary L. Bastin, Ph.D.
Dr. Gary Bastin, Engineer Scientist, and Task Order Lead, was responsible for the
detailed day-to-day execution of this project, including the preparation of the written final
report, as well as of the choice and execution of the UWB evaluation hardware and test
methods selected. He also derived and wrote the UWB theory sections of this document,
and prepared the New Technology Report resulting from this research project.
Robert Chiodini
Mr. Robert Chiodini, Telecommunications Engineer, was responsible for the detailed
testing of the UWB Evaluation Kit hardware. He wrote and executed the UWB test plan
and test procedure, and collected and analyzed the test data.
PoTien Huang
Mr. PoTien Huang, as Principal Investigator, was responsible for monitoring the weekly
progress of the overall project. Additionally, he supported the theoretical UWB
investigations.
David A. Kruhm
Mr. David A. Kruhm was responsible for managing the overall UWB project