Cognitive Radio for WLANs

 
Cognitive Radio for WLANs
It is widely believed that the technical foundations established by WLANs provide a launching pad for
cognitive radios. WLANs already incorporate essential cognitive radio features such as DFS and TPC.
Also, while the RF front ends may require wideband receivers and transmitters, the hardware exists
now, and the software only involves the generation of the software engineering to make functions
like filtering, band selection, and interference mitigation available as plug-in software modules for
the radios.
In the following text, we will discuss several study cases for WLAN implementation using
cognitive radio technology.  Cognitive radio in IEEE 802.11h
The first case we would like to discuss here is the IEEE 802.11h standard, which was a modified
version of IEEE 802.11a standard, as an effort to reduce possible interference to some existing users in
the same RF band, such as radar applications, and so on. The two important elements, DFS and TPC,
used in IEEE 802.11h standard, bear the important characteristics of a cognitive radio. Therefore, it
is justified to say that the IEEE 802.11h standard incorporates some functions only used in cognitive
radios.
As we have discussed in Section 4.2, the IEEE 802.11a wireless networks operate in the 5-GHz RF
band and support as many as 24 nonoverlapping channels, which are less susceptible to interference
than their IEEE 802.11b or IEEE 802.11g counterparts. However, regulatory requirements governing
the use of the 5-GHz band vary from country to country, hampering 802.11a’s rapid deployment in
different regions of the world.
To overcome the problem, the ITU recommended a harmonized set of rules for IEEE 802.11a
WLANs to share the 5-GHz spectrum with primary user devices such as military radar systems, and
so on. Issued in 14 October 2003, the IEEE 802.11h standard [806] defines mechanisms that 802.11a
WLAN devices can use to comply with the ITU recommendations. These mechanisms are DFS and
TPC, which are two very important functionalities that every cognitive radio must provide. WLAN
products supporting 802.11h are already available in the market.
The DFS detects other devices working in the same RF channel, and it switches WLAN operation
to another channel whenever necessary. DFS is responsible for avoiding interference with other
devices, such as radar systems and other WLAN segments, and for uniform utilization of channels.
An access point (AP) specifies that it uses DFS in the frames WLAN stations use to find APs.
When a WLAN station associates or reassociates with an AP, the station reports a list of channels that
it can support. When it is necessary to switch to a new channel, the AP uses this data to determine
the best channel.
The AP initiates a channel switch by sending a frame to all stations associated with the AP
that identifies the new channel number, the length of time until the channel switch takes effect, and
whether or not transmission is allowed before the channel switch. Stations that receive the channel
switch information from the AP change to the new channel after the elapsed time.
An AP measures channel activity to determine if there is other radio traffic in the channels being
used for a WLAN or other applications, such as radar systems. The AP sends a measurement request
to a station or group of stations identifying the channel where activity is to be measured, the start
time of the measurement and the duration of the measurement. The station performs the requested
measurement of channel activity and generates a report to the AP.
On the other hand, the TPC is intended to reduce interference from WLANs to satellite services
by reducing the radio transmit power that WLAN devices use. The TPC can also be used to manage
the power consumption of wireless devices and the range between APs and wireless devices.
An AP specifies telephony control protocol (TCP) support in the frames it generates to WLAN
stations. These frames also specify the maximum transmit power allowed in the WLAN and the
transmit power the AP is currently using. The transmit power used by stations associated with an AP
cannot exceed the maximum limit that the AP sets.
When a WLAN station associates with an AP, the station indicates its transmit power capability.
The AP uses this data about the stations associated with it to determine the maximum power for
the WLAN segment. This means the radio power in a WLAN segment can be adjusted to reduce
interference with other devices while still maintaining a sufficient link margin for the operation of
the wireless network.
Frames are also sent between the AP and the stations to monitor the signal strength of the wireless
network. The AP can dynamically adjust the radio signal strength, if necessary, to maintain wireless
communications. The original motivation for the DFS and TPC mechanisms defined in 802.11h ensure a standard
method of operation under the regulatory requirements governing the 5-GHz band, which will spur
the deployment of 802.11a wireless networks around the world, especially those places where a strict
regulation is imposed. Along with meeting regulatory requirements, the DFS and TPC can be used
to improve the management, deployment, and operation of WLANs.
Figure 9.15 illustrates the five-step working procedure specified in IEEE 802.11h protocol based
on DFS and TPC etiquettes.
Cognitive radio to enhance WLANs security
Another application of cognitive radio to helpWLANs is the security enhancement of existing WLANs
standards, such as IEEE 802.11b/g/a, the explosive growth of which has created managerial challenges
for those who monitor them.With wireless intrusion threatening security and RF interference impeding
performance, WLAN managers need an up-to-date, detailed understanding of the RF environment to
make informed decisions about how to solve these problems. Cognitive radio technology enables a
WLAN device and its antenna to sense its RF environment and adapt its spectrum use as needed to
avoid interference. Integrated software and silicon solutions enable cognitive radio to be built into
enterprise-class WLAN APs to boost security and optimize performance. APs with this feature are
expected to soon be available in the market.
Wireless intrusion occurs when unauthorized WLAN users gain access to a secured network.
The causes of intrusion include hackers creating ad hoc networks with WLAN clients or a rogue AP
connected to a wired network without proper security levels. In either case, the keys to prevention are
quick identification, containment, and defensive action. The first step in preventing wireless intrusion
is to identify the intrusion point. Because WLANs are fixed to a specific WLAN channel during
operation, they cannot simultaneously detect intrusion points on other WLAN channels. Relying on a
single-radio AP to provide access and security is far from sufficient. Network managers must be able
to monitor the full WLAN frequency range to be able to reliably detect and identify intrusion points.
Cognitive radio for WLANs provides a means to observe the RF environment in both 2.4-GHz
and 5-GHz frequency bands, on which IEEE 802.11 WLANs operate without disrupting normal
wireless voice over IP (VoIP) and data traffic. Continuous scanning of both bands for IEEE 802.11
and non-802.11 devices allows for the timely detection of intruders. With detailed information from
multiple cognitive radios within a WLAN, administrators can take preventive action. Cognitive radio can detect non-802.11 devices. This is important because interference, regardless
of sources, lowers effective data throughput and overall network performance. IEEE 802.11b/g/abased
WLANs operate within the unlicensed radio spectra around 2.4 GHz (802.11b and 802.11g)
and 5 GHz (802.11a). Other wireless connectivity standards, such as Bluetooth and HomeRF, operate
in the same unlicensed radio spectra. Microwave ovens, cordless phones and industrial equipments
can also generate noise in these bands.
This technology allows a network to detect, identify, and avoid these noise sources. Proactive
behavior allows a network to be established on clearer channels during initial deployment. Ongoing
monitoring allows network administrators to take rapid corrective action against RF noise. This affords
the optimal network performance that WLAN users expect. Integrated software and silicon solutions
enable APs to be developed to provide simultaneous WLAN access in the 2.4-GHz and 5-GHz bands
while providing integrated cognitive radio functionality that ensures security and performance for
enterprising wireless networks.
Figure 9.16 shows how a cognitive radio AP works to identify interference, rogue APs, and other
wireless intruders to enhance the security level of WLANs.
More works on WLANs
IEEE is very active in promoting the applications of cognitive radios in all its already very popular
IEEE 802.11 standards, as we have discussed in the previous text on IEEE 802.11h standard. In
fact, there are at least two other IEEE 802.11 standards that have either been issued or are in their
standard-making process, including IEEE 802.11e for the support of Quality of Service (QoS), and
IEEE 802.11k for new types of radio resource measurements, which are two well suited candidate
technologies for future cognitive radios.
Cognitive radios coordinate the usage of radio spectra without the involvement of restrictive radio
regulation. They operate in spectra when it is not used by licensed radio systems, and therefore share
spectra with the radio systems that have priority access. This is referred to as vertical sharing. An
unused radio spectrum is called a spectrum opportunity. In the vertical sharing scenario, cognitive
radios adapt their transmission schemes in such a way that they fit into the identified spectrum usage
patterns of the incumbent radio systems. Note that cognitive radio does not only refer to a novel radio technology. It also requires a
revolutionary change in how our spectrum will be regulated. However, with this change and the
new cognitive radio approach for open spectrum sharing, it will be difficult to achieve complete
fairness and efficient spectrum sharing. This is particularly challenging in ad hoc mesh networks,
where cognitive radio systems rely on their own capabilities not only to maintain connectivity, but
also to increase spectrum efficiency. This horizontal sharing problem in mesh networks is a challenge
for cognitive radios, and a lot of work needs to be done in designing new spectrum etiquette rules,
which are voluntary rules based on the mechanisms like DFS, TPC, adaptive duty cycles, or carrier
sensing.
In addition to the design of spectrum etiquettes, we also need to provide insights on cognitive
algorithms and reasoning based on machine-understandable languages and logics. The support
of the quality of service in spectrum sharing scenarios is another challenging problem. Decentralized
cognitive algorithms on the basis of spectrum observation for mutual coordination are
required.
Here we will focus on a cognitive radio operating in the unlicensed radio band. Particularly, we
focus on the challenges faced by IEEE 802 WLANs operating in the newly allocated 5.470–5.725 GHz
band and the 3.650–3.700 GHz band currently under consideration.
In order to utilize these bands a cognitive radio must sense the presence of the incumbent
spectrum owner and vacate within a very short period of time with a minimum number of transmissions.
It must also perform communications with a minimum amount of transmit power. In the
5.470–5.725 GHz band the incumbent users are radio-location, radio navigation, and meteorological
radars. In the 3.650–3.700 GHz band the incumbent user is the C-band fixed space-to-earth satellite
services.
The transmission characteristics can vary significantly even within the same band. According to
ITU-R M.1461 [807] pulse repetition rates for radars operating in the 5 GHz band can vary from 20
to 100,000 pulses per second and the 3-dB bandwidth of the transmitter varies from approximately
500 kHz to approximately 150 MHz.
Detection must be successful irrespective of the varied transmission characteristics of the incumbent
users and irrespective of the instantaneous RF propagating conditions. In order to ensure that
minimum power is transmitted, the device must determine the minimum power that will be necessary
to maintain communications in a dynamic RF environment. Furthermore, in order to avoid
unnecessarily declaring a channel of the network as occupied, the designer must trade off the stringent
requirements of the probability of detection against an unnecessarily high probability of false
alarm.
Usually 802.11 system designers perform this trade-off by combining an energy detect threshold
that can be triggered by noise with a correlator which detects packet preambles. The initial energy
detect can be used to adapt the gain in the front end of the receiver; while the correlator can be used
to establish if a valid packet is truly present. Unfortunately, this is not applicable to radar detection
due to the varied characteristics of the signals that the system is trying to detect.
Furthermore, using a static energy detect threshold is particularly problematic as the duration and
bandwidth of the incumbent signals is variable. The system must have a spectrum analysis capability
in order to effectively contend with this situation.
Another challenging issue in the implementation of cognitive radio for WLANs is that of moving
the network and resuming communications with minimum disruption to the network. In the
European Telecommunications Standards Institute (ETSI) regulatory domain a WLAN must vacate
a channel in 10 s after the first radar pulse is detected with a channel closing time, that is, maximum
transmission time during a channel move, of 260 ms. Furthermore, before a channel can be
utilized it must be sensed for a minimum of 60 s. This presents some very significant challenges.
One of the most important issues is how to detect the interferer while maintaining communications
and simultaneously being prepared to move to a new channel without ceasing operations
within 60 s.