Mainstream WiFi Standards
Since the introduction of the basic 802.11 wireless LAN in 1997, there have been many amendments and advancements to date to address the requirements of new applications and demands. While some of these amendments sought to improve the capacity and efficiency of the mainstream WiFi used by billions of people to access the Internet, others targeted some niche applications to further enhance the utility of the technology. In this chapter, we will focus on the mainstream WiFi standards while the niche standards will be examined in the following chapter.
5.1 802.11 Amendments and WiFi Evolution
Since its first appearance in 1997, WiFi has gone through significant evolutions increasing its efficiency and data rates to meet the growing demand for wireless connectivity. Table 5.1 provides a chronological list of the major amendments along with the key enhancements and the maximum data rates they can support. As we can see, WiFi data rate has increased from a mere 2 Mbps in 1997 to a whopping 9.6Gbps in 2020, approximately a 5-fold increase in 23 years. Interestingly, the next version of WiFi is striving to achieve another 3-fold increase to 30 Gbps in just 4 years, possibly beating the speed of wired wireless LANs for the first time in history.
Copyright By PowCoder代写 加微信 powcoder
While it is easy to see the benefits of higher data rates, the details that contribute to data rate increase is less trivial to understand. In the rest of this chapter, we shall examine the factors that contribute to data rate enhancements for each of these amendments.
Table 5.1 Chronological list of mainstream IEEE 802.11 amendments
802.11 Amendment
802.11-1997 802.11b-1999
Key Enhancements
Legacy WiFi in 2.4GHz (now extinct!) Higher speed modulation in 2.4GHz
Max. Data Rate
2 Mbps 11 Mbps
© , Wireless and Mobile Networking, 2022, CRC Press
802.11a-1999 802.11g-2003 802.11n-2009 802.11ac-2013
802.11ax-2020
802.11be-2024 (expected)
Higher speed PHY (OFDM) in 5GHz Higher speed PHY (OFDM) in 2.4GHz Higher throughput in 2.4/5GHz
Very high throughput in 5GHz
High efficiency in 2.4/5GHz
Extremely high throughput in 2.4/5/6GHz
54 Mbps 54 Mbps 600 Mbps ~7 Gbps
~9.6 Gbps ~30 Gbps
5.2 Basics of WiFi Data Rates
Each WiFi version supports a range of specific data rates. For example, 802.11a [802-11a] supports 8 different data rates: 6, 9, 12, 18, 24, 36, 48, and 54 Mbps. Fundamentally, data rate of WiFi or for any other communications technology is derived as:
Data rate = symbol rate x data bits per symbol
While the symbol rate is defined by the PHY, the number of data bits carried in a symbol depends on the choice of modulation and coding. It should be noted that only 802.11b used DSSS, while all subsequent WiFi amendments used OFDM as their PHY. Usually, many different combinations of modulation and coding are available for a given PHY, which leads to a range of specific available data rates for a given WiFi. When MIMO is employed, data rates can be further increased linearly with the number of independent spatial streams supported by the MIMO system. Next, we are going to examine each of the mainstream WiFi amendments, discuss the main enhancements they introduce compared to their predecessor, and how these enhancements increase their achievable data rates.
5.3 Data Rate in DSSS-based WiFi: IEEE 802.11-1997 and 802.11b-1999
Recall that the original 802.11 released in 1997 supported only 2 Mbps for 22 MHz channels using the Direct Sequence Spread Spectrum (DSSS) technique at the physical layer. In this section, we will examine how the data rate for DSSS is computed and how the 802.11b was able to increase the DSSS rate to 11 Mbps for the same 22 MHz channel.
First, let us look at the use of chips in a DSSS system as illustrated in Figure 5.1, where a binary coded symbol (single bit per symbol) is spread with 10 chips. Both the original 802.11 and 802.11b [802-11b] operate at 1/2 chip per Hz, which gives a chip rate of 11Mchips/s for the 22MHz channel. Second, we note that 802.11 uses a Barker code, which uses 11 chips per symbol. On the other hand, to increase the data rate,
© , Wireless and Mobile Networking, 2022, CRC Press
802.11b employs Complementary Code Keying (CCK), which employs only 8 chips to code a symbol. This means we have a symbol rate of 1 Msps (mega symbol per second) for the 2Mbps rate and 1.375 Msps for the 11 Mbps data rate. The third factor that determines the final data rate is the symbol coding, which determines how many data bits are conveyed per symbol. For 2-Mbps rate, 802.11 uses 2 bits per symbol, whereas for the 11 Mbps, it uses 8 bits per symbol.
Now we can verify the final data rates by multiplying the symbol rates with the bits per symbol for each data rates. Specifically, for the 2 Mbps, we have 1Msps x 2bits/symbol=2Mbps. For the 11 Mbps, we have 1.375 Msps x 8 bits/symbol = 11Mbps
Figure 5.1 An example of DSSS with binary modulated symbols spread with a chip rate of 10 chips per symbol.
Example 5.1:
A WLAN standard is employing a spread spectrum coding with only 1⁄2 rate, which produces chips at a rate of 1⁄2 chips per Hz. It uses 8 chips to code a symbol and 16 QAM modulation to modulate the symbol stream. What would be the data rate for 22 MHz channels?
Chip rate = 1⁄2 x 22 = 11 Mcps (cps = chips per second)
Symbol rate = 11/8 = 1.375 Msps (sps = symbols per second)
Bits per symbol = log2(16) = 4 [16 QAM produces 4 bits per symbol] Data rate = symbol rate x bits per symbol = 1.375 x 4 = 5.5 Mbps
5.4 Data Rate in OFDM-based WiFi
OFDM, which was adopted in WiFi from 802.11a onwards, has a completely different structure than its predecessor, DSSS. The symbol rate in OFDM is obtained as the inverse of the symbol interval (a.k.a. symbol length or duration), which includes a data interval followed by a guard interval. The actual symbol is contained within the data interval, whereas the guard interval is used to avoid inter-symbol interference. The longer the delay spread, the longer the guard interval and the lower the symbol rate.
© , Wireless and Mobile Networking, 2022, CRC Press
The number of bits carried in an OFDM symbol depends on the subcarrier structure and the modulation order of the symbol. OFDM divides a WiFi channel into many subcarriers. All these subcarriers are divided into three categories: data subcarriers, pilot subcarriers, and guard subcarriers. Only the data subcarriers carry the OFDM symbols. Pilots estimate the wireless channel, while the guards protect the symbol against interference from the adjacent channels. The guard subcarriers are thus equally distributed to the front and rear of the middle subcarriers.
Although the allocation of subcarriers to pilot and guard reduces the total number of data subcarriers, it is interesting to note that each OFDM symbol is carried over all the data subcarriers in parallel, which significantly boosts the effective bits per symbol. For example, an OFDM with N data subcarriers each applying M-ary modulation, the effective number of bits sent per symbol is obtained as Nxlog2M.
Finally, the actual number of data bits per symbol is affected by the choice of error correcting codes and their coding rates. For example, with a coding rate of 3⁄4, 4 bits are actually transmitted for every 3 data bits. Similarly, a coding rate of 2/3 implies 2 data bits for every 3 bits transmitted, and so on. The number of data bits per OFDM symbol therefore is obtained as:
Data bits per OFDM symbol = coding rate ✘ log2M ✘ #-of-data-subcarriers Example 5.2:
What is the data rate of an OFDM WiFi applying 64-QAM and a coding rate of 3⁄4 to its 48 data subcarriers? Assume a symbol interval of 4μs.
Log2M = log264 = 6
Coded bits per symbol = log2M ✘ #-of-data-subcarriers = 6×48 = 288
Data bits per symbol = coding rate x 288 = 3⁄4 x 288 = 216
Symbol rate = 1/symbol-interval = 1⁄4 Msps (0.25 million symbols per sec.) Data rate = symbol rate x data bits per symbol = 216 x 1⁄4 Mbps = 54 Mbps
Table 5.2 summarises the 5 key parameters that affect data rates in OFDM-based WiFi. In the rest of this chapter, we will examine how the successive amendments exploited these parameters to enhance the data rates from their predecessors.
Table 5.2 Five key parameters affecting WiFi data rates Parameter Description
Guard interval Affects symbol rate; the longer the interval, the lower the © , Wireless and Mobile Networking, 2022, CRC Press
Modulation
Affects the number of bits per symbol; Log2M bits per symbol for M-ary modulation; usually multiple modulation option are available
Error correcting coding affects the actual number of data bits per symbol; usually multiple coding options are available; an integer number, called MCS (modulation and coding system), defines a particular combination of modulation and coding;
symbol rate and vice versa.
Channel Width
Affects the number of achievable OFDM subcarriers and hence ultimately the data rate; channel width can be increased by combining multiple channels into a single one (a.k.a. channel bonding), an option available from 802.11n onwards
MIMO streams
Number of independent data streams that can be sent in parallel; more streams means higher achievable data rates, and vice versa; MIMO available from 802.11n onwards; newer amendments have increased number of MIMO streams compared to their predecessor
5.5 IEEE 802.11a-1999
802.11a is the first amendment to use OFDM, which allowed it to push the date rates to 54 Mbps. Actually, 802.11a supports 8 different data rates, from a mere 6 Mbps up to 54 Mbps, by selecting a combination of modulation and coding to dynamically adjust for the noise and interference.
802.11a divides the 20 MHz channel bandwidth into 64 subcarriers. Out of these 64 subcarriers, 6 at each side are used as guards (a total of 12 guards) and 4 as pilot, which leaves 48 of them to be used to carry data.
802.11a OFDM has a symbol length of 4 microsecond, which gives a symbol rate of 0.25 M symbols/s. Therefore, with a modulation of BPSK for example, there will be 1 coded bit per subcarrier for each OFDM symbol, or 48 coded bits per OFDM symbol in total as the symbol is transmitted over all of the 48 subcarriers in parallel. The actual data bits transmitted per symbol will however depend on the coding used. 802.11a supports three coding rates, 1/2, 2/3, and 3/4.
The modulation schemes are fixed and cannot be changed i.e., to operate at a particular data rate, the corresponding combination of modulation and coding scheme (MCS) has to be selected. Table 5.3 shows the MCS combinations of each data rate in 802.11a. Note that the data bits per symbol has to be multiplied by the symbol rate of 0.25 M symbols/s to obtain the final net data rate shown in the last column.
Table 5.3 Modulation, coding, and data rates for 802.11a
BPSK 1/2 1 48 24 6 BPSK 3/4 1 48 36 9
Modulation
Coding Rate
Coded bits per subcarrier
Coded bits per symbol
Data bits per symbol
Data Rate (Mbps)
© , Wireless and Mobile Networking, 2022, CRC Press
QPSK 1/2 QPSK 3/4 16-QAM 1/2 16-QAM 3/4 64-QAM 2/3 64-QAM 3/4
5.6 IEEE 802.11g-2003
Although 802.11a was able to
and was not compatible with the previous version (802.11b), which was operating in the 2.4GHz and at 11 Mbps. 802.11g [802-11g] achieved 54 Mbps at 2.4 GHz using OFDM, but it could fall back to 802.11b data rates using CCK modulation. More specifically, 802.11g OFDM data rates are identical with 802.11a, i.e., it supports 6, 9, 12, 18, 24, 36, 48, 54 Mbps as per Table 5.3, while CCK supports data rates of 1, 2, 5.5, and 11 Mbps. This seamless backward compatibility made 802.11g very popular, because previous hardware designed to operate in the 2.4GHz band can now benefit from the higher data rates without having to switch to a new spectrum.
5.7 IEEE 802.11e-2005 (Enhanced QoS)
While amendments 802.11a and 802.11g were racing to increase the data rates through enhancements in the PHY layer, 802.11e [802-11e] was released in 2005 to enhance the WiFi medium access control (MAC) for supporting quality of service (QoS). To achieve QoS, delay sensitive traffic, such as voice and video, must be given priority over delay-tolerant traffic, such as web browsing and file transfer. 802.11e provided the necessary protocol support at the MAC layer to achieve that.
802.11e achieves QoS by introducing a Hybrid Coordination Function (HCF) for MAC. HCF allows both contention-free access using a Point Cordination Function (PCF), where the stations are polled by the access point. When PCF is used, stations cannot attempt to access the channel unless the AP provides access to it, which eliminates contentions. On the other hand, the stations can also use a contention-based access, called Enhanced Distributed Control Function (EDCF), where they can contend for the medium access, but with priority assigned for each packet based on the type of service.
Basically, EDCF achieves priority by implementing four separate priority queues within the station (see Figure 5.2). When a packet is delivered to the MAC from the
2 96 2 96 4 192 4 192 6 288 6 288
push the date rates to 54 Mbps, it used
the 5 MHz band
© , Wireless and Mobile Networking, 2022, CRC Press
upper layer, it is queued in the appropriate queue based on the priority required for the service. Each queue has separate values for the transmission parameters, such as the minim and maximum congestion window values.
Figure 5.2 Priority queuing in enhanced DCF
Frame Bursting
802.11e also introduces batch transmission of multiple frames, which is called frame bursting. Instead of sending one frame at a time, a station can request, in an RTS packet for example, for a maximum transmission opportunity (TXOP) duration and send multiple frames back-to-back within that time. The receiver can then acknowledge all the frames together instead of acknowledging them one by one. Voice or gaming has high priority, but allowed to use small TXOP. In contrast, data has low priority but can access long TXOP to send many frames in burst. Figure 5.3 shows how EDCF TXOP works.
Figure 5.3 Frame bursting with TXOP
Direct Link
Another new feature allowed in 802.11e is to allow stations to send packets directly to another station within the same BSS without going through the BS. Figure 5.4 illustrates this feature. This will further reduce the latency for some delay-sensitive communication.
© , Wireless and Mobile Networking, 2022, CRC Press
Figure 5.4 The direct link feature in 802.11e
5.8 IEEE 802.11n-2009
The data rate of 54 Mbps achieved in 1999 with 802.11b served the application demands well at that time. Since then, demand for more bandwidth continued to soar fueled by more devices being connected to the LAN, growing popularity of on- demand video streaming, and so on. In late 2000, it became apparent that new amendments must come forth to boost the speed and capacity of wireless LANs. In fact, some vendors already started to release products with some proprietary enhancements to meet the market demand. It was time to standardize these developments.
In 2009, IEEE introduced 802.11n [802-11n] to significantly increase the data rates of wireless LAN from the previous versions of 802.11a/b/g. The target was to break the 100 Mbps mark and go well beyond that. To achieve this major WiFi data rate boost in history, 802.11n introduced five important techniques, which promised a massive maximum data rate of 600 Mbps.
First, it employed the MIMO technology in WiFi history for the first time to capitalize on the potential of multiple independent streams existing over the same frequency. Second, it reduced the coding overhead by employing a 5/6 coding rate which is much lower than the previous minimum allowed rate of 3⁄4 used in 802.11a. Third, the guard interval and inter-frame spacing were reduced to increase the number of OFDM sub-carriers that can carry data. Fourth, it allowed a new physical layer mechanism, called channel bonding, to combine two consecutive 20-MHz channels into a single 40-MHz channel without any guard intervals between them. Fifth, it promoted reduction of MAC layer overhead by packing multiple frames inside a single frame, called frame aggregation, thereby amortizing the frame header bits over many data bits.
5.8.1 MIMO: Number of antennas and number of streams
Recall from our earlier discussion on MIMO that multiple antennas at the transmitter and the receiver help transmit data in multiple simultaneous independent streams. Clearly, the larger the number of these independent streams, the higher the effective data rates. However, the maximum number of independent streams are limited by the minimum number of antennas available at the transmitter or receiver. The individual
© , Wireless and Mobile Networking, 2022, CRC Press
implementations may further reduce the number of independent streams limiting the total capacity of the MIMO infrastructure.
The convention n x m: k is used to describe the number of antennas and streams in a given system, where n is the number of available antennas in the transmitter and m is the number of antennas in the receiver. The number of streams is represented by k, where k is less than min(n,m). For example, 4 x 2: 2 means that the transmitter has 4 antennas, but the receiver has only 2. Only 2 parallel streams are used to transmit the data in this configuration.
802.11n allows a maximum of 4 x 4: 4 configurations. When there are more receive antennas than the number of streams, then the throughput can be maximized by selecting the best subset of antennas. For example, with a 4 x 3: 2 configuration, the best 2 receive antennas should be selected for processing the received data.
5.8.2 Reduction of Coding Overhead
Recall that the coding rate directly affects the net data rates. Given the raw bit rate, C_raw, of a channel, which shows the number of coded bits transmitted per second, the net data rate, C_data, is derived as C_data = C_raw x C_rate, where C_rate is the fraction representing the coding rate.
Previously, 3⁄4 was the minimum coding rate allowed in any 802.11 amendments. 802.11n allows a coding rate as low as 5/6, which directly increases net data rate by a factor of (5/6)/(3/4) = 11%. Of course, this 11% increase in data rate is entirely due to the reduction of the coding rate. As we will see in the remaining of this section, the ultimate data rate boost will be much more than this, once all other techniques are employed simultaneously.
5.8.3 Reduction of Guard Interval and Increase of Data Sub-carriers
Guard intervals are time-domain guards used between every two consecutive data symbol transmissions to overcome the effect of multipath or inter-symbol interference at the receiver. A direct consequence of guard interval is the reduction of data rates, as no data can be transmitted during the guard interval. Clearly, data rate can be increased by using shorter guard intervals between two data symbols. Figure 5.5 illustrates how by reducing the guard interval slightly, 6 data symbols can be transmitted instead of 5 during the same time interval. 802.11n therefore targets reduction of guard interval as another means for increasing the net data rate.
The rule of thumb is to allow a guard interval four times the multi-path delay spread. Initial 802.11a design assumed 200ns delay spread, which lead to 800 ns guard interval. For 3200 ns data blocks, this incurs a overhead of 800/(800+3200)=20%. Detailed experimental analysis revealed that most indoor environments have a delay spread in the range of 50-75 ns. 802.11n therefore selects a guard interval of 400 ns, which is more than four times this value. Now the guard interval related overhead is reduced
程序代写 CS代考 加微信: powcoder QQ: 1823890830 Email: powcoder@163.com