Niche Wi Fi has been primarily used as a networking technology for implementing wireless LAN in enterprise and residential domains, as well as connecting personal mobile devices, such as mobile phones, tablets, laptops, etc. to the Internet in homes, cafes, airports, and university campuses. These mainstream WiFi predominantly used the ISM bands 2.4GHz and 5GHz, with the new versions aiming to use the 6GHz band. In addition to these mainstream WiFi, IEEE has also released several 802.11 amendments that target some niche applications. These niche WiFi standards operate outside the mainstream bands, both at the very low end of the spectrum, i.e., below 1GHz, as well as at the very high end, i.e., 60GHz (see Figure 6.1). For example, 802.11af is targeting the exploitation of 700MHz spectrum recently vacated by TV stations due to their digitization, 802.11ah using 900MHz to connect emerging Internet of Things operating at low power, and 802.11ad/ay at 60GHz to support multi-gigabit applications at short range. In this chapter, we shall examine the features and techniques used by these niche WiFi standards.
TVWS, Long-distance, Rural
IoT, low energy 700MHz 900MHz
Mainstream WiFi: personal devices, wireless local area networks
Copyright By PowCoder代写 加微信 powcoder
Cable replacement, Data center, backhaul
2.4GHz 5GHz 6GHz 60GHz
Figure 6.1 Mainstream and niche WiFi
6.1 802.11af (a.k.a. White-Fi)
When TV transmissions switched from analog to digital, they vacated a lot of spectrum in the licensed TV bands. The vacated TV spectrum is called the White Space. IEEE 802.11af [802- 11af], which is also referred to as White-Fi (or Super-Fi), was designed to effectively exploit the white space for data communications.
6.1.1 Over-the-Air Television Channels
When TV was invented, it was using analog signals for transmitting and distributing programs over the air. Analog television channels used the spectrum between 30 MHz to 30 GHz. The channels are called High Frequency (HF), Very High Frequency (VHF), Ultra High
b/g/n/ax/be
a/n/ac/ax/be be
Frequency (UHF), and so on as shown in Figure 6.2. VHF is basically a meter band as the wave length is between 1 to 10 meters. UHF could be called decimeter band and so on.
Each channel uses 6 MHz in USA, 8 MHz in Europe, and 7 MHz at some other parts of the world. The numbering system for the VHF and UHF TV channels in USA is shown in Figure 6.3. Channel 37 is used for radio astronomy, hence excluded from TV transmissions. Also, some channels between 88 and 174 are reserved for FM radio.
At least one channel is skipped between two analog stations in neighboring areas to avoid interference. For example, if a small town has channel 2, then it will not have channel 3. Basically, all channels cannot be allocated to all cities and towns.
Figure 6.2 TV spectrum and channels
Figure 6.3 TV VHF and UHF channel numbers in USA. Frequencies are in MHz.
6.1.2 Digital TV
Analog TV broadcast has been discontinued recently in most parts of the world. The world has switched to digital broadcast due to many advantages. The main mantra for digital TV is that all pictures are represented as pixels and each pixel is represented by some bits. Once the pictures are converted to bits, it becomes like computer communications. Encryption, multiplexing, mixing with different services and types of data, etc. all become very efficient, just like computer communication networks.
Another main advantages of going digital is that we no longer need to provision for significant guard bands between occupied frequencies because interference from adjacent frequencies can be managed by sophisticated frame and error control techniques. Digital transmission also uses compression at the transmitter and decompression at the receiver, which further reduces spectrum usage for digital TV. Consequently, multiple digital channels can be transmitted within 6 MHz, which was previously used to transmit only one analog TV program. This bandwidth efficiency has freed up a lot of TV spectrum, which is dubbed as Digital Dividend.
There was a particular demand for this “new” spectrum in 700 MHz band for Cellular, Emergency Services, and ISM. Consequently, governments were able to raise significant revenue by auctioning part of this spectrum to cellular companies while reserving the rest for unlicensed use. Similar practices happened in other countries.
Figure 6.4 illustrates the basic differences between 700 MHz and higher frequency. The wavelength in 700 MHz is much longer and hence it can travel far and penetrate many obstacles, such as buildings.
700 MHz has lower attenuation (1/7th to 1/9th of 1800/1900/2100 MHz), which means it requires lower transmission power and can provide longer mobile battery life for mobile devices. It can have larger cell radius, which means smaller number of towers. Such long- distance propagation is good for rural areas. It means providing cellular and wireless broadband services to rural areas become more cost effective and affordable. Because of these reasons, availability of new spectrum in 700 MHz is considered a very good opportunity for wireless networking.
Figure 6.4 Differences between 700 MHz and higher frequency. A wave cycle in 700 MHz can travel much further than that in 2.4 GHz.
6.1.3 Spectral White Spaces
A lot of spectrum is allocated to certain services. However, the spectrum is not fully used at all locations and times. In general, white space is defined as any spectrum at a given area at a given time available for use on a non-interfering basis. The white space may be due to unallocated spectrum, allocated but under-utilized, channels not used to avoid interferences in adjacent cells, or spectrum available in the TV band due to digital dividend.
Figure 6.5 shows that allocation does not mean it is always used. It is called “white” because when spectrum usage is plotted in blue, the white gaps are the spectrum not used. Figure 6.6 shows a measurement conducted in Ottawa, Canada, for the UHF spectrum and we can see that most of it is white!
Well, it is clear that if we can use the white space, which appears to be in abundance, we can address some of the high demands for data. However, we must acknowledge that the white space that belongs to licensed spectrum poses interesting legal and policy issues, because the spectrum was already licensed for massive amount of fee to certain companies. Under previous ruling, those companies actually had the right to say no to the use of their spectrum whether they use it or not. Effective use of white space therefore requires new rules to be in place first.
Figure 6.5 Concept of spectral white space. Allocation does not mean it is used.
Figure 6.6 Spectral usage example from real measurements [TVWS-Ottawa].
6.1.4 TVWS Databases
It was agreed that the TVWS databases, a.k.a. geolocation database (GDB), which would hold information about which channel is free and when, would be operated by third parties. These databases do the following four things:
1. Get info from FCC database
2. Register fixed TVWS devices and wireless microphones
3. Synchronize databases with other companies
4. Provide channel availability lists to TVWS devices
Google was one of the third parties that acquired license to operate such databases, but it does not provide this service anymore. Figure 6.7 shows an example of what was available in the city of St. Louis (zip code 63130) using the Google database. We can see that there were 17 channels available at the time of accessing the WS database.
Figure 6.7 White space available near University of Washington in St. Louis.
6.1.5 802.11af Database Operation
Recall that in whitespace networking, the APs do not have a fixed set of channels, as the available spectrum is not known in advance, but rather must be found out dynamically. Therefore, to implement 802.11af, which uses white space, protocols and mechanism must be developed for the LAN to obtain the available channels from the white space databases, i.e., GDBs, maintained by the third parties and distribute such channels within the 802.11af network.
To achieve these objectives, a local cache or database called Registered Location Secure Server (RLSS) is maintained which stores the channel availability information learned from the public GDBs. This provides faster access to channel information. The idea is that all large companies and ISPs will have their own RLSSs, just like DNS cache or local DNS server.
To facilitate communication with GDBs and RLSSs, two new protocols are defined. One is called PAWS [PAWS2015], defined by IETF, which is used by the GDBs and the APs. PAWS is a general protocol that can be used by 802.11, or any other networks to query the spectrum in GDB. The other protocol is called Registered Location Query Protocol (RLQP), defined by IEEE, to be used locally between the AP and the stations. The use of these two protocols in accessing the white space databases is shown in Figure 6.8. The APs are called Geolocation Database Dependent (GDD) enabling, as they can interface directly with GDB, while the stations are called GDD dependent, because they cannot talk to GDB directly.
Figure 6.8 White space database access protocols.
6.1.6 Registered Location Query Protocol (RLQP)
RLQP is a protocol for exchange of white space map (WSM), a.k.a., Channel Schedule Management (CSM), among RLSS, APs, and stations. An example of message exchanges for RLQP is shown in Figure 6.9.
As we can see, AP uses CSM request and response to obtain the available channels first before these channels can be allocated internally with 802.11af network. Stations can be disassociated by the APs if necessary, such as if the channel becomes unavailable. Table 6.1 explains the meaning of all the other messages.
CSM Request CSM Response CVS
Figure 6.9 Message exchange in RLQP.
Table 6.1 Description of RLQP messages
Description
APs asks other APs or RLSS about white space map
White space map is provided
APs supply white space map to their stations and confirm that stations are still associated
Stations ask AP if they do not receive the map within a
timeout interval
NCC Request Sent by stations to AP requesting use of a channel. AP may
forward to RLSS
NCC Response Permission to transmit on requested channel
6.1.7 Protocol to Access White-Space (PAWS)
There can be many different technologies, such as 802.11, 802.22, etc. that may work on white space and need access to WS databases. For a WS database provider it would then be necessary to design interfaces for all these different networking technologies. Instead, IETF has decided to come up with a single protocol, called PAWS, which is independent of any network technologies as well as the underlying spectrums. All WS networks will have to implement PAWS to access the WS databases and all WS database providers will have to implement PAWS to support WS networking.
PAWS has the concept of master and slave. Master device is one that can directly interface with the GDB using PAWS. A slave device is a WS networking device that cannot talk to a GDB directly, i.e., does not implement PAWS. Instead, the slave devices will need to communicate to a master device to find out spectrum availability. A device can act as both master as well as a slave. In Figure 6.10, the RLSS is a master device. The AP and BS are acting as both masters, as they can talk directly with the GDB, as well as slaves because they can get spectrum information from RLSS. Some 802.11af clients, not shown in the figure, that must get channel information from the AP are slaves only.
Figure 6.10 The master-slave concept in PAWS
How does the WS networking devices find out the addresses or URLs of the GDBs in the first place? There are several ways this could be implemented. One method could be to preconfigure devices with certificates to talk to some known database authority. Another could be using a listing server to list all national database servers.
Query is a pull-based method. In pull-based, the master sends a query to the database each time it needs to know the availability of white space. For a master device, it may also be possible to receive push notifications from the GDB. The master can register with the GDB for such push notifications using its certificate and the database can push channel availability information whenever some new spectrum is available or availability of an old channel changes. Finally, to ensure security, all PAWS messages are encrypted.
Some sample PAWS messages and how they are exchanged are shown in Figure 6.11. As we can see, after the exchange of initialization messages, the registration messages are exchanged. It is only after the registration that the master device can send a query message to the database server and get a response. The master device can also send batch query to include requests for a set of slave devices located in different locations with different antenna heights etc. and get a batch response.
Figure 6.11 Message exchange in PAWS
6.1.8 802.11af channels and data rates
In 802.11af, a Basic Channel Unit (BCU), a.k.a. W, is one TV Channel. In the USA, this means that W = 6 MHz. While the use of single channel is default, channel bonding is optional. Two kinds of channel bonding are allowed. For contiguous channel bonding, 2W or 4W are allowed, i.e., 2 or 4 contiguous channels can be bonded together. This means that it is possible to have 12 MHz or 24 MHz contiguous spectrum as a single (bonded) channel.
802.11af uses a maximum of 256-QAM and 5/6 coding. It uses OFDM similar to 40 MHz in 802.11n, but down clocked by 7.5x. This gives a total of 144 subcarriers, of which 108 are data subcarriers. The down clocking increases the GI from 0.4μs to 3μs, and the data interval from 3.2μs to 24μs. As a result, the total symbol interval becomes 27μs, which yields a maximum data rate of 26.67Mbps for a single stream and single channel link. Note that 802.11af supports MIMO with up to 4 streams, which can further boost the data rate. Table 6.2 shows the various data rates supported by 802.11af for a 6MHz channel.
Example 6.1
What is the maximum possible data rate achievable with 802.11af?
Data rate with single stream and single 6MHz channel = 26.67Mbps
Data rate with 4 streams and 4 bonded channels = 4x4x426.7 = 426.7Mbps.
Table 6.2 802.11af data rates in Mbps: Single Stream, single unbonded (6MHz) channel
MCS Modulation Coding
0 BPSK 1/2
1 QPSK 1/2
2 QPSK 3/4
3 16-QAM 1/2
4 16-QAM 3/4
5 64-QAM 2/3
6 64-QAM 3/4
7 64-QAM 5/6
8 256-QAM 3/4
9 256-QAM 5/6
6.2. 802.11ah (a.k.a. HaLow)
Data Rate 2
12 16 18 20 24 26.7
IEEE 802.11ah [802-11ah] is also known as HaLow. The most interesting and historical change in 802.11ah compared to all previous versions is that this is the first time 802.11 is considering wide area networking, while all other versions were in the space of local area networking. Its ability to support long range is therefore the key difference.
To achieve the long range, spectrum is shifted from high frequency (above GHz, e.g. 2.4GHz and 5GHz) to sub-GHz. With the lower frequency, comes several key advantages for IOT. At sub-GHz, signals can travel longer distances with low power and penetrate buildings, roads, and other infrastructure, which will hide many future IOT devices. Also, there is less congestion at sub-GHz as all other WiFi devices work in either 2.4GHz or 5GHz. Also, number of devices that use sub-GHz are not many.
With lower frequency band, the achievable bit rate is low, but this is not an issue for IOT, because the sensors do not need to stream high definition video, but only short messages. With low data rate, we can also reduce MAC overhead, which is important for short messages. In fact, with MIMO, the data rate can be from 150 kbps to 78 Mbps per spatial stream (up to 4 streams are allowed), which is sufficient for all types of IOT devices. Finally, the low data rate allows APs to connect 4 times more devices than existing WiFi, which is very important for densely deployed IOT.
The spectrum allocation for HaLow is shown in Figure 6.12. We can see that different countries have allocated slightly different spectrum, but they are close to 900 MHz, just below the GHz mark.
Figure 6.12 spectrum allocation for HaLow [802-11ah-bands]
6.2.1 Sample Applications
As the Figure 6.13 shows, the main application of 802.11ah is the neighborhood area network (NAN). The NAN is used to read various meters from houses as well as some municipality owned devices for monitoring smart cities, such as monitoring manholes, underground pipes, cables etc. The 802.11ah APs, which could be deployed on the streetlight poles, are then connected via wire to the cloud, where all the data ends up for processing by the data analytic.
Figure 6.13 802.11ah sopports neighbor hood area network (NAN)
6.2.2 802.11ah PHY
802.11ah PHY is actually built on top of 802.11ac PHY, but down clocked by 10x. This means each clock tick is now 10 times longer than 802.11ac, which will have a 10x effect on many aspects of the protocol.
First of all, 802.11ah will have 2/4/8/16 MHz channels in place of 20/40/80/160 MHz in 802.11ac. That is the channels are 10x smaller in MHz. However, the
20 MHz 11ac and 2 MHz 11ah both have 52 data subcarriers plus 4 pilots, which means 1/10th inter-carrier spacing in 11ah. The shorter spacing may mean higher inter-carrier interference, but as we shall see shortly, this is well
compensated by longer symbol lengths.
subcarriers for 802.11ah channels are the same as the corresponding higher channel
number of data
bandwidths in 802.11ac. For example,
802.11ah has 10x longer symbols, which allows 10x delay spread. Therefore, longer multipath can be accommodated within the symbol, which avoid inter-symbol interference even in long distance communication (longer distance means longer multipath).
In 802.11ah, all type of times, such as SIFS, are 10x longer. 802.11ah defines a new 1 MHz PHY with 24 data subcarriers. However, channel bonding is defined for two 1 MHz channels to form a single 2 MHz channel. All stations have to support both 1 MHz and 2 MHz channels.
With 1 MHz channel, 802.1ah defines a new modulation and coding scheme, MCS10, which is basically the previous MCS0, but after MCS0, every bit is transmitted twice. This allows 802.11ah to achieve long range as it can now sustain more errors. The rest of the MCS indices, i.e., the modulation and coding combinations remain the same, but the data rates are 10x lower from the corresponding MCS in 11ac. For example,
802.11ah supports 4 spatial streams instead of 8 in 802.11ac. 802.11ah supports beamforming to create sectors, which can be sued by a single AP to read meters from houses in different sectors more efficiently.
Example 6.2
If we reduce the clock speed of 802.11ac by a factor of 10, what would be the new symbol rate (symbols/s)?
802.11ac has a symbol duration of 3.6 μs (for 400 ns GI). New symbol duration with a 10x slower clock = 36 μs New symbol rate = 1/(36 x 10-6) = 27,777 sym/s
Example 6.3
In USA, 902-928 MHz has been allocated for 802.11ah. How many different channels can be used if 16 MHz channel option is used?
902-928 MHz has a total bandwidth of 26 MHz. There is only one (non-overlapping) 16 MHz channel possible out of 26 MHz.
6.2.3 IEEE 802.11ah MAC
802.11ah MAC faces some new challenges due to IOT requirements and hence new features are introduced to address these challenges.
coding), data rates for 11ac and 11ah, respectively are 6.5Mbps and 0.65Mbps for single
stream 2/20MHz channels using the longer GI option.
程序代写 CS代考 加微信: powcoder QQ: 1823890830 Email: powcoder@163.com