Cognitive Radio for WPANs

 
Cognitive Radio for WPANs
In this and the following sections, we will discuss the applications of cognitive radio technology in
different wireless networks. In this section, we are going to discuss the issues on the possible application
of cognitive radio technology in a WPAN environment, in which a UWB (IEEE 802.15.3a)4
network operates, although the ideas presented here are equally applicable to other WPANs, such as
Bluetooth (IEEE 802.15.1) [461], ZigBee (IEEE 802.15.4), and so on.
In December 2004 the FCC officially began the effort to extricate the concept from academia by
issuing a NPRM [795] calling for input on how cognitive radio could be realized. In the response
stage, the NPRM underscores the unprecedented willingness of the commission in recent years to
explore innovative ways to open up new spectra to commercial unlicensed use. One of the examples
includes the release of new spectra in the 5 GHz U-NII band in 2003, as well as the opening up
of 7.5 GHz of bandwidth for UWB signaling in the region between 3.1 and 10.6 GHz. Though the
power levels allowed for UWB were extremely low, with its roof being 41 dBm, the move marked
the first time the FCC had allowed unlicensed use across otherwise licensed bands.
The UWB technology, as an important part of the WPANs, was one of the first beneficiary parties
of the FCC’s NPRM. In fact, UWB transmits signals that are already below the noise floor. With the
help of cognitive radio technology, a UWB terminal can also operate by jumping out really fast from
one channel to another if it detects incumbent users, and yet it can transmit at a very reasonably low
power level.
As mentioned in Section 7.6, UWB technology has generated a great deal of interest recently as an
attractive means to provide high-speed short-range communications. However, it has also generated
a lot of controversy. As UWB signals are flat over a broad range of spectrum at a power level close
to the noise floor, some people are concerned that UWB will artificially raise up the noise floor
and degrade the performance of the existing primary users located at the same spectrum. After a
lot of public debates, the FCC approved the first Report and Order on February 14, 2002, in which
the FCC not only gave the definition of the UWB signals, but also defined a spectrum mask that specifies the amount of power that can be sent out by any UWB system working in the band. The
spectrum mask is shown in Figure 7.19. The FCC’s definition is explained in the following text.
According to the FCC ruling that permitted the operation of UWB systems, an UWB signal is
defined as a signal that must satisfy the criteria mentioned below.
Assume that fH is the upper 10 dB cutoff frequency and fL the lower 10 dB cutoff frequency.
Also identify that the center frequency is fc = fH +fL
2 and the fractional bandwidth is Fp = s fH−fL
fH+fL
.
Then, it is a UWB signal if the condition that either FP > 0.2 or fH + fL > 500 MHz holds good.
Two types of interference problems
The FCC’s spectral mask for UWB signals can help to control the interference to other users of the
spectrum, such as GPS, radar and satellite systems, and so on. However, this mask also requires
UWB to avoid sending signals in the given band even if it is not in use by any other devices within
the spectrum. On the other hand, according to the power emission requirements as specified in IEEE
802.15.3a, the UWB signal is generally sent out at or close to the thermal noise floor. Therefore, it is
essentially imperceptible in most cases by an incumbent user more than a few tens of meters away.
In addition, most UWB devices will normally work indoors, and thus the possible interference to
those incumbent users (i.e., GPS, radar and satellite systems, and so on.) is very unlikely. However,
the outdoor operation of UWB systems will generate the problem of the interference to primary users
operating in the same band.
A more serious concern of interference is due to the fact that IEEE 802.11 WLANs and IEEE
802.15.1 Bluetooth devices are also operating in the same 2.4 GHz ISM bands. Therefore, there are
basically two major interference issues that we need to investigate in the WPAN working environment:
one being the UWB’s interference to the licensed users allocated in the same band (such as GPS,
radar and satellite systems, and so on.), and the other being the mutual interference among unlicensed
users (such as WLANs and Bluetooth devices, and so on.). The cognitive radio technology is well
suited for the UWB applications to overcome the above two interference problems.
To overcome the first kind of interference problem in UWB applications, it is impossible to
mandate a coordination mechanism for spectrum sharing, as non-UWB applications are primary users
of the spectrum. Therefore, there will be no collaboration in this scheme and the use of the cognitive
radio in the UWB devices will try to avoid using the band if it detects it is in use by some incumbent
users, in order not to interfere with their normal operations. Clearly, in this case, the cognitive radio
is used by UWB devices on a noncollaborative basis, and only part of cognitive radio functionalities
can be used.
As an example to using cognitive radio in a UWB system to avoid interfering with other incumbent
users, let us consider a multiband OFDM UWB system, whose signal can be shaped in the frequency
domain to avoid generating interference in some particular parts of the spectrum. This particularly
shaped signal spectrum can be based on a priori decision from the spectrum scanning, which is an
important part of the cognitive radio’s functionalities. Furthermore, the tone weighting coefficients
can also be derived from the spectrum scanning process, which in fact measures the interference
temperature of the working environment. In this way, a multiband OFDM UWB signal permits great
flexibility in this kind of “spectrum sculpture,” which is hardly possible with other UWB schemes,
such as pulsed or direct-sequence UWB signals.
On the other hand, we can also consider a WPAN environment, in which other nonlicensed
devices, such as Bluetooth and Wi-Fi networks, may exist. In order to achieve the optimal performance
for all nonlicensed devices, cognitive radio technology can be applied to all the terminals which form
a cognitive radio network, and can work jointly on a collaborative basis. Only in this case, can
the benefit of a cognitive radio network be fully realized. Basically, there are four parameters that
a cognitive radio network can try to optimize: (1) transmission data throughput, (2) error rate, (3)
quality of service, and (4) cost of connection. The optimization of the aforementioned four parameters
can be achieved by leveraging the following seven approaches: (1) power level control, (2) antenna beam steering, (3) carrier frequency, (4) channel-coding adaptation, (5) transmission time slot, (6)
MAC protocol adaptation, and (7) using CDMA to manage interference.
Obviously, the set of MAC protocols is generally fixed to some extent, but the other parameters
can normally be chosen independently from a fairly wide range of values. There is probably some
fruitful research into methods for including protocol as part of the optimization; for example, a WLAN
user can often choose to modify packet size and data rate, and so does a Bluetooth network. Optimal
choices of these parameters in a WPAN using cognitive radio technology are an active research area.
PulseLINK cognitive radio UWB
Before ending this section, we would like to talk about one already-in-market UWB cognitive radio
chipset developed by PulseLINK, Inc., which conducted private showings of the PulseLINK chip
along with various demonstrations of its UWB technology at the ITU Global Conference on Ultra
Wideband held on from June 9–18, 2004 in Boston, Massachusetts. PulseLINK, Inc. is a private
Delaware Corporation founded in June 2000 and headquartered in Carlsbad, California. PulseLINK
has over 190 issued and pending patents pertaining to UWB wired and wireless communications
technology.
The company unveils an architecture that supports PulseLINK’s UWB wireless, UWB power line
(i.e., UWB across electric power lines), and UWB cable communications technologies (i.e., UWB
for Cable Television Networks) simultaneously on the same chipset.
Combining these networking technologies on a single chipset allows consumer electronics and
computing devices around the home to be seamlessly networked together wirelessly, through existing
home electrical wiring or across CATV cabling as conditions and usage warrants. For Cable Television
customers, those home networks can then be networked to the rest of the world through a massive
new two-way UWB data pipe enabled through the existing CATV infrastructure, all on the same
chipset.
PulseLINK’s target data rates for its chipset are up to 1 Gbps for UWB wireless connectivity,
up to 200 Mbps for UWB power line communications (electrical wiring in homes and small offices),
and up to 1 Gbps of new downstream bandwidth across existing cable television networks in addition
to hundreds of megabits of new bandwidth per node upstream.
The fundamental architecture of the chipset, implemented on Jazz Semiconductor’s SiGe 120
process, is that of a Software-Defined Cognitive Radio. The hardware platform used in the chipset
reduces the complexity and implementation of wireless and wired PHY to a software abstraction.
Thus, future evolutions can be anticipated based on this same chip to be able to support and deliver
narrowband carrier signals such as Wi-Fi, WiMAX, or IEEE 802.15.3a wireless UWB standard as
well. Because it is driven by software, such evolution will not require any modifications to the chipset
hardware architecture and will be purely a matter of software/firmware upgrades.