List of standards for the IEEE 802.11 wireless network. All existing standards of Wi-Fi networks

IEEE 802.11 standard (Wi-Fi)

Wireless networks of the IEEE 802.11 standard operate in two bands: 2.4 …… 2.483 GHz and in several bands around 5 GHz, which are not licensed. In this case, several topology options are possible:

• independent basic service areas (independent basic sets, IBSSs),
• basic service sets (BSSs),
• extended service areas (ESSs).

An independent base coverage area is a group of 802.11-compliant stations communicating directly with one another. IBSS is also called ad-hoc network. In fig. 6.8 shows how three stations equipped with 802.11 wireless network interface cards (NICs) can form an IBSS and communicate directly with one another.

Rice. 6.8. Episodic (ad-hoc) network

Basic service area technology assumes a special station: access points AP (access point). The access point is the central point of communication for all BSSs. Client stations do not communicate directly with each other. Instead, they send messages to the access point, and it already sends information packets to the destination station. The access point may have an uplink port through which the BSS is connected to wired network(for example, an Ethernet uplink for Internet access). Therefore, the BSS is called an infrastructure network. In fig. 6.9 shows a typical BSS infrastructure.

Rice. 6.9. Wireless LAN with infrastructure

Several BSS infrastructures can be connected through their uplink interfaces. Where the 802.11 standard is in effect, the uplink interface connects the BSS to the distribution system (DS). Several BSSs, interconnected through a distribution system, form an Extended Service Area (ESS). The uplink to the distribution system does not need to be wired. The specification of the 802.11 standard allows this channel to be built as wireless. More often, however, the uplinks to the distribution system are wired Ethernet links. In fig. 6.10 shows an example of a practical implementation of ESS.

An area covered by a BSS or ESS with Internet access is called a hot spot. Hot spots are being created in hotels, airports, restaurants, student residences and just on the streets. At the end of 2004, there were about 50,000 hot spots in the world, and the number of their users reached
50 million people. The proliferation of WLAN services and the large number of hardware manufacturers require interoperability between hardware and software vendors. To this end, the Wireless Ethernet Compatibility Alliance (WECA) was established in 1999, which soon became the Wi-Fi Alliance. It includes developers and manufacturers of 802.11 equipment, network operators, experts. The main goal of the alliance is certification of the manufactured equipment in order to ensure the interaction of Wi-Fi devices manufactured by different companies.

Rice. 6.10. Extended service area of ​​ESS wireless LAN

The 802.11 standard has 3 options: 802.11a, b and g. In all cases, information is transmitted in batch mode, by separate frames (packages).

802.11b equipment operates in the 2.4 ... 2.483 GHz band. As mentioned, this range is unlicensed and many other systems and devices operate in it. To reduce the influence of interference in 802.11b networks, 2 methods have been proposed. The first is the use, as in Bluetooth standard, hopping frequency when transmitting each next frame. However, in practice, another method is usually used: direct spreading of the spectrum by filling information symbols with a scrambling code.

In the classic version of 802.11b, information is transmitted in the form of symbols at a rate of 1 Msps. With 2-PM modulation, the data transfer rate in the frame is 1 Mbit / s, and with 4-PM, 2 Mbit / s. When using direct spread spectrum, each symbol is filled with an m-chip sequence of 11 chips (Barker code): +1, -1, +1, +1, -1, +1, +1, +1, -1, -1, -1 ... The chip speed in the radio channel is 11 Mchip / s, and the radio channel width is 22 MHz. In the 2.4 GHz band, the central frequencies of 13 radio channels are fixed: 2412, 2417, 2422, 2427, 2432, 2437, 2442, 2447, 2452, 2457, 2462, 2467 and 2472 MHz. Upon reception, the signal is subjected to correlation processing, which significantly reduces the effect of interference, as in the standards cellular communication with code division of channels.

The use of a wideband channel allows for a high signal-to-noise ratio (15 - 17 dB) to increase the data transmission rate. In this case, scrambling is abandoned, and data is transmitted at a symbol rate of 11 Ms / s with 4-PM modulation. To improve the quality of communication during transmission, redundant coding is used using the Complementary Code Keying (CCK). The frame rate can be 11 or 5.5 Mbps.

The maximum transmit power of 802.11b devices is 100mW in Europe and 1W in the US.

802.11a devices operate in three sub-bands at 5 GHz. In the 5.15 ... 5.25 GHz sub-band, the transmission power is limited to 50 mW, in the 5.25 ... sub-band. 5.35 GHz - 250 mW, and in the sub-band 5.725 ... .5.825 GHz -
1 Watt In these ranges, 12 channels with a width of 20 MHz are allocated.

The advantage of the 802.11a standard in comparison with 802.11b is the increased data transfer rate in the frame: from 6 to 54 Mbps. For this, the standard 802.11a uses OFDM modulation: Orthogonal Frequency Division Multiplexing - orthogonal frequency division multiplexing. This method is used to eliminate intersymbol interference at high data rates. Let's give a typical example.

Let the radio channel transmit at the symbol rate B = 40 Msymb / s. When transmitting on a single carrier, the symbol duration is s. Imagine the transmission of such a signal in a large room (train station, airport, shopping center- rice. 6.11).

Figure 6.11. Multipath signal propagation

In order for the forward and backward rays to arrive with a delay of 1 symbol, the difference in their paths must be only m. Such a delay can be observed even in a sufficiently large room. To eliminate the problem of intersymbol interference, you should increase the length of the symbol by 10, or even better by 100 times. Then, intersymbol interference will be noticeable with a trace difference of 750 m.From this follows the idea behind OFDM: split the high-speed data stream into many separate streams (dozens!), Transmit each of the substreams at its own frequency (subcarrier), increasing the symbol length to one milliseconds.

The generalized character is the sum of characters transmitted to N S subcarriers. All subcarriers can use different modulations: 2-PSK, 4-PSK, 8-PSK, 16-QAM or 64-QAM. The timing diagram of the OFDM signal is shown in Fig. 6.12, where the number i individual subcarriers are marked.

Rice. 6.12. OFDM signal structure

The symbols are specially separated from each other by pauses of duration T p so that in the case of a multipath signal, adjacent symbols do not "creep" on top of each other.

The total OFDM signal at can be represented as:

, (6.1)

where is the complex amplitude of one transmitted signal,

t s- start time of each individual character,

T s- the duration of the symbol.

The spectral pattern of the OFDM signal is shown in Fig. 6.13.

Rice. 6.13. OFDM signal spectrum

In order to distinguish between signals transmitted on adjacent subcarriers during reception, all signals must be mutually orthogonal. This condition is met if the distance between adjacent subcarriers.

When transmitting (forming) an OFDM signal, an inverse discrete Fourier transform (inverse FFT) is used; at reception - the direct discrete Fourier transform (FFT). The OFDM signal is formed at a reduced frequency with the subsequent transfer of the spectrum to the frequency of the radio channel.

In the 802.11a standard, 48 subcarriers are used for information transmission (52 in total). Symbol duration T s= 3.2 μs, pause duration T p= 0.8 μs. Distance between adjacent frequencies MHz. With 2-PM modulation on each subcarrier, data rate (no guard coding)

When switching to multi-position modulation methods

Mbps,

Mbps.

Depending on the interference situation, the 802.11a standard provides for the use of adaptive modulation and coding schemes. The main characteristics of the standard are given in table. 6.4.

Table 6.4

The 802.11g standard combines the capabilities of the 802.11a and b standards in the 2.4 ... 2.483 GHz band. The main characteristics of the standard are given in table. 6.5. In addition to CCK and OFDM, the standard uses redundant packet binary convolutional coding (PBCC) at a number of rates.

Table 6.5

Access to the network of subscriber stations and the ability to transmit frames in 802.11 networks is carried out using coordinate functions. Using distributed coordinate function DCF (distributed coordination function) all stations have the same priority and occupy the channel based on competition with rollback timers. The principle of operation of the DCF is illustrated in Fig. 6.14.

Rice. 6.14. Station operation in DCF mode

Working stations listen to the radio channel and wait until it becomes free (carrier transmission stops). In fig. 6.14, station 3 first transmits, and stations 1, 2 and 5 are ready for transmission. After the frame of station 3 is completed, a mandatory DIFS interframe gap follows (34… .50 μs), after which the stations, ready to transmit their packets, begin the race. Each of the stations starts a race timer, where random numbers are set inside the race window: 0 ... ..7, 0 ... 63, and then up to 127, 255, 511, 1023. From the moment the race starts, the timers are read with a clock speed of 9 ... 20 μs ... The station that is the first to reset the timer occupies the channel (in Fig. 6.14 station 2). The rest remember the contents of their timers (rollback) until the next match. During the transmission, collisions are possible when two stations simultaneously reset their timers (stations 4 and 5 in Fig. 6.14). This leads to an expansion of the race window with subsequent retransmission of frames.

In a real DCF-based access algorithm, a more robust procedure is used (Figure 6.15). The winning station sends a short request packet to the RTS receiver - Request to Send, which receives confirmation of the recipient's readiness to receive CTS - Clear to send... This is followed by the transmission of an information frame. The cycle completes an ACK frame acknowledgment (or non-acknowledgment) packet. This is how the file exchange over the TCP / IP protocol is implemented.

Rice. 6.15. DCF-based network access procedure

In the transmission cycle, RTS - CTS - Data - ACK frames are separated by short SIFS interframe gaps (10 ... 16 μs). Stations not participating in the exchange, according to the information contained in the RTS and CTS frames about the duration of the transmission cycle, set up the NAV vectors (network allocation vector). NAV is the time to read the timer during which the station is in “sleep” mode and does not participate in the contention until NAV is equal to 0.

The considered access method is used when reading files from the Internet. However, it does not allow streaming video and, moreover, IP-telephony, where the permissible signal delays are strictly limited. The new IEEE 802.11e standard provides for support in Wi-Fi networks of four traffic classes, arranged in priority order:

Voice - telephony with transmission quality at the long-distance communication level,

Video - television transmission,

Best Effort - reading Internet files,

Background - low priority file transfer.

This classification corresponds to the classes of network services mobile communications 3rd generation, which allows you to organize the interaction of mobile and Wi-Fi networks. The implementation of the 802.11e standard is possible only in networks with access points, where they use point coordinate function PCF (point coordination function). The principle of operation of a network based on PCF is illustrated in Fig. 6.16.

The transfer process is determined by the AP. The transmission time is divided into superframes, the duration of which is set by the AP adaptively and can be changed during the transmission. At the beginning of each superframe, the AP transmits a beacon frame. It sets the super-frame duration, the maximum data frame size, and the contention-free period. At this time, the exchange of information between the access point and the stations is only by polling the AP (the station itself cannot occupy the channel). Simultaneously with sending a polling frame, the AP can also send a data frame to the station. The end of the no-race period is marked by the AP by sending a CF-End frame. Thereafter, stations, including the AP, occupy the channel on a race basis. This access method allows you to organize the transmission of data packets at a constant rate, which is necessary for telephone and streaming traffic.

Rice. 6.16. Data transfer based on PCF

It must be said that the PCF point coordinate function does not fully provide the QoS parameters. To support the required quality of service, a special 802.11e standard has been developed. It introduces the concept of AC access categories, which derive from the 802.1D group of standards and define priority levels. There are 4 access categories (Table 6.6): Voice, Video, Best Effort and Background. Each category is associated with a corresponding data type.

Table 6.6

The 802.11e standard defines a new type of media access to ensure quality of service - hybrid coordinate function (hybrid coordination function, HCF). HCF defines two mechanisms for accessing the environment:

· Contention-based channel l access;

· Controlled channel access.

Contention-based channel access corresponds to extended distributed channel access ( enhanced distributed channel access, EDCA), and controlled channel access corresponds to HCF-controlled channel access ( HCF controlled channel access, HCCA). In 802.11e, there are still two phases of operation within a superframe — contention periods (CP) and contention-free periods (CFP). EDCA is used only in CP, and HCCA is used in both periods. HCF combines PCF and DCF methods, therefore it is called hybrid. The result of the transformation of the MAC architecture is shown in Fig. 6.17.

Rice. 6.17 MAC architecture

A station that acts as a central coordinator for all stations within a basic service set supporting QoS ( QoS supporting BSS, QBSS), is called a hybrid coordinator ( hybrid coordinator). It, like the point coordinator, is located inside the access point. Client stations that support QoS are called QSTA.

An 802.11e station that is granted access to the medium should not use radio resources longer than that specified in the standard. This new introduction is called the transfer capability ( transmission opportunity, TXOP). TXOP is the interval during which the station has the right to transmit packets. It is determined by its start time and duration. The TXOP that exists in contention-based media access is called EDCA-TXOP. Similarly, a TXOP that exists in controlled media access is called HCCA-TXOP. The duration of EDCA-TXOP is limited by the TXOPlimit parameter, the value of which is constantly transmitted through a certain information element lighthouse frame fields.

Another improvement to the standard is that no station can transmit when it is time to transmit a beacon frame. This reduces the expected latency of the beacon, which gives the hybrid coordinator better control over the environment, especially when the optional CFP is used after the beacon frame.

In the new standard, a station can transmit packets directly to another station in QBSS without having to communicate with an access point. In the old standard, within a network with an infrastructure, all data exchange packets between stations went only through an access point.

QoS support in EDCA provides concepts such as access categories and multiple independent rollback objects ( backoff entities). Each 802.11e station can have several parallel rollback objects, and these objects are assigned different priorities according to a set of specific parameters of access categories ( EDCA parameter set). As mentioned above, there are four access categories, respectively, in each station there are four rollback objects (Fig. 6.18). The EDCA parameter set prioritizes media access by defining individual frame gaps, race windows, and other parameters.

Rice. 6.18. Four access categories in one station

Each access category has its own interframe gaps ( arbitration interframe space, AIFS), similar to DIFS, but of different duration. In addition, the size of the race window changes depending on the priority of the traffic.

6. 5. IEEE 802.16 - WiMAX standard

WiMAX-Worldwide Interoperability for Microwave Access

Table 6.7

Main characteristics of the WiMAX standard

Table 6.8

The non-profit organization WiMAX (World Interoperability for Microwave Access) was formed with the aim of promoting the development of wireless equipment for accessing broadband networks based on the IEEE 802.16 specification for wireless area networks, certifying such equipment for compatibility and interoperability. as well as accelerating its time to market.

The 802.16 standard provides for operation in the ranges 2 ... 11 GHz and 10-66 GHz (Fig. 6.1). In the range 10-66 GHz, radio communication is possible only in case of line of sight between the points. This band uses direct carrier modulation (single carrier mode).

In the range 2 ... 11 GHz, the radio interface specifications allow the possibility of solving the problem of radio communication in multipath propagation conditions and in the absence of line of sight (NLOS - Non-Line-Of-Sight). The WMAN-SC2 radio interface uses single carrier modulation, the WMAN radio interface uses OFDM - Orthogonal Frequency Division Multiplexing (OFDM) with fast Fourier transform for 256 points and up to 2048 points. The certified frequency bands for fixed and mobile WiMAX profiles are shown in Fig. 1.

Fixed WiMAX profiles- 3.5 GHz (FDD): 3.5; 7; (256)

3.5 GHz (TDD): 3.5; 7; (256)

5.8 GHz (TDD): 10 (256)

Mobile WiMAX profiles- 2.3 - 2.4 GHz: 5 (512); 10 (1024); 8.75 (1024);

all TDD 2.305 - 2.320 GHz: 3.5 (512); 5 (512)

2.345 - 2.360 GHz: 10 (1024)

2.496 - 2.69 GHz: 5 (512); 10 (1024)

3.3 - 3.4 GHz: 5 (512); 7 (1024); 10 (1024)

3.4 - 3.8 GHz: 5 (512)

3.4 - 3.6 GHz: 7 (1024)

3.6 - 3.8 GHz: 10 (1024)

In addition to those indicated, it is possible to allocate channels in the 5.7 GHz bands,
1.710 - 1.755: 2.110 - 2.155 GHz.

The following interfaces are used in the 802.16 standard:

1. WirelessMAN-SC (10 - 66 GHz)

2. WirelessMAN-SCa (2-11 GHz; licensed bands)

3. WirelessMAN-OFDM (2-11 GHz; licensed bands)

6. WirelessMAN-OFDMA - Orthogonal Frequency Division Multiple Access

(2-11 GHz; licensed bands)

5.WirelessHUMAN (2-11 GHz; unlicensed bands)

Interfaces 3 and 5 offer Mesh capabilities, a full-fledged network topology to accelerate traffic.

The inverse Fourier transform determines the shape of the OFDM signal. The useful symbol duration is Tb. The last part of the Tg of the symbol period, called the guard interval, is used to eliminate the effects of multipath propagation of the orthogonal signal components (Figure 6.19).

Rice. 6.19. Single frequency symbol format

In the frequency domain, the signal is characterized by spectral characteristics (Figure 6.20). It contains subcarriers for data transmission, pilot signals, and guard intervals are located at the edges of the band.

Rice. 6.20. Frequency Domain Signal Description

The OFDM symbol is characterized by the following parameters:

BW is the nominal channel bandwidth.

Nused - the number of used subcarriers.

N is the sampling rate. This parameter, in conjunction with BW and Nused, determines the subcarrier spacing and symbol duration. The required values ​​for this parameter are defined in Table 6.6.

G is the ratio of the duration of the guard interval (prefix) to the useful time. This value can be 1/4, 1/8, 1/16, 1/32 Tb.

NFFT: number of points Fourier transform,

Sending frequency: Fs = floor (n * BW / 0.008) * 0.008 (BW- bandwidth in MHz),

-∆f: subcarrier spacing, defined as: Fs / NFFT,

Tb = 1 / ∆f - duration of symbol transformation,

Tg = G * Tb - duration of the guard interval (CP),

Ts = Tb + Tg - OFDM symbol duration,

Ts / NFFT - sampling interval.

The main parameters of the OFDM channels of the 802.16a standard are given in table. 6.9.

Table 6.9.

The duration of the symbols depending on the channel bandwidth is given in table. 6.10.

Table 6.10

Modulation and coding schemes l for the 802.16-2004 standard are summarized in Table. 6.11.

Table 6.11

The values ​​of the transmission rates depending on the type of modulation and code rate are given in table. 6.12, and the requirements for the signal-to-noise ratio at the receiver input for various modulation and coding schemes in Table. 6.13.

Table 6.12

Table 6.13

Physical data is transmitted as a continuous sequence of frames. Each frame has a fixed duration (2 (2.5) ... 20 ms), so its information capacity depends on the symbol rate and modulation method. A frame consists of a preamble, a control section, and a sequence of data packets. IEEE 802.16 duplex networks. Both frequency FDD and TDD time division of the uplink and downlink channels are possible.

With time duplexing of channels, the frame is divided into downstream and upstream subframes (their ratio can be flexibly changed during operation depending on the bandwidth needs for upstream and downstream channels), separated by a special guard interval. With frequency duplexing, the upstream and downstream channels are transmitted on two carriers (Figure 6.21)

Rice. 6.21. Frame structure for TDD and FDD

In the downlink, information from the base station is transmitted as a sequence of packets. For each packet, you can specify the modulation method and data coding scheme - i.e. choose between speed and reliability of transmission. TDM - packets are transmitted simultaneously for all subscriber stations, each of them receives the entire information flow and selects "its" packets. In order for subscriber stations to distinguish one packet from another, downlink (DL-MAP) and uplink (UL-MAP) maps are transmitted in the control section (Fig. 6.22).

Figure 6.22. Downlink structure.

The downlink map shows the frame duration, frame number, number of packets in the downlink subframe, and the start point and profile type of each packet. The starting point is counted in so-called physical slots, each physical slot is equal to four modulation symbols.

A packet profile is a list of its parameters, including the modulation method, the type of FEC-coding (with the parameters of the coding schemes), as well as the range of values ​​of the signal-to-noise ratio in the receiving channel of a particular station, in which this profile can be applied. The base station periodically broadcasts the list of profiles in the form of special control messages (downlink and uplink DCD / UCD descriptors), and each profile is assigned a number that is used in the downlink map.

Subscriber stations get access to the transmission medium through the TDMA (Time Division Multiple Access) mechanism. For this, in the ascending subframe for the AU, the base station reserves special time intervals - slots (Fig. 6.23). The slot allocation information between the SSs is recorded in the UL-MAP uplink channel map broadcast in every frame. UL-MAP - functionally similar to DL-MAP - it reports how many slots are in a subframe, the starting point and connection ID for each of them, as well as the types of profiles of all packets. The UL-MAP message of the current frame can be related to both this frame and the next one. The modulation rate (symbol rate) on the upstream channel must be the same as on the downstream channel. Note that, unlike TDM downlink packets, each uplink packet starts with a preamble, a 16 or 32 QPSK symbol sync sequence.

Rice. 6.23. Uplink channel structure

Examples of the structure of a frame with TDD are shown in Fig. 6.24.

Rice. 6.24. Example of an OFDM frame structure with TDD

In the upstream channel, in addition to the slots assigned by the BS for certain SSs, there are intervals during which the SS can send a message for initial registration in the network or for requesting a change in channel bandwidth (provision of channels on demand DAMA - Demand Assigned Multiple Access).

The physical layer of the IEEE 802.16 standard provides direct delivery of data streams between BS and SS. All tasks related to the formation of these data structures, as well as the management of the system operation, are solved at the MAC (Medium Access Control) level. Equipment of the IEEE 802.16 standard forms a transport medium for various applications (services).

WiMAX networks support 4 types of traffic that differ in reliability and latency requirements:

UGS - Unsolicited Grant Service - real-time transmission of signals and streams of telephony (E1) and VoIP. The admissible delay is less than 5 - 10 ms in one direction at BER = 10 -6 ... 10 -6.

rtPS - Real Time Polling Service - real-time streams with variable length packets (MPEG video).

nrtPS - Non-Real-Time Polling Service - support for variable length streams when transferring files in wideband mode.

BE - Best Effort - the rest of the traffic.

Wireless networks of the IEEE 802.11 standard operate in two bands: 2.4 …… 2.483 GHz and in several bands around 5 GHz, which are not licensed. In this case, several topology options are possible:

• independent basic service areas (independent basic sets, IBSSs),
• basic service sets (BSSs),
• extended service areas (ESSs).

Rice. 4.8. Episodic (ad-hoc) network

Rice. 4.9. Wireless LAN with infrastructure

Rice. 4.10. Extended service area of ​​ESS wireless LAN

The 802.11 standard has 3 options: 802.11a, b and g. In all versions, information is transmitted in batch mode, in separate frames (packets).

802.11b equipment operates in the 2.4 ... 2.483 GHz range

In the classic version of 802.11b, information is transmitted in the form of symbols at a rate of 1 Msps. With 2-PM modulation, the data transfer rate in the frame is 1 Mbit / s, and with 4-PM, 2 Mbit / s. When using direct spread spectrum, each symbol is filled with an m-chip sequence of 11 chips (Barker code): +1, -1, +1, +1, -1, +1, +1, +1, -1,

-1, -1 ... The chip speed in the radio channel is 11 Mchip / s, and the radio channel width is 22 MHz. In the 2.4 GHz band, the central frequencies of 13 radio channels are fixed: 2412, 2417, 2422, 2427, 2432, 2437, 2442, 2447, 2452, 2457, 2462, 2467 and 2472 MHz. When received, the signal is subjected to correlation processing, which significantly reduces the effect of interference, as in the standards of cellular communication with code division multiplexing.

802.11a devices operate in three sub-bands at 5 GHz. In the 5.15 ... 5.25 GHz sub-band, the transmission power is limited to 50 mW, in the 5.25 ... sub-band. 5.35 GHz - 250 mW, and in the 5.725 ... 5.825 GHz sub-band - 1 W. In these ranges, 12 channels with a width of 20 MHz are allocated.

The advantage of the 802.11a standard in comparison with 802.11b is the increased data transfer rate in the frame: from 6 to 54 Mbps. For this, the standard 802.11a uses OFDM modulation: Orthogonal Frequency Division Multiplexing - orthogonal frequency division multiplexing. This method is used in order to eliminate intersymbol interference due to multipath propagation of signals at high data rates (Figure 4.11).

Figure 4.11. Multipath signal propagation

The idea behind OFDM is to split the high-speed data stream into many separate streams (tens, hundreds, thousands!), Transmit each of the substreams at its own frequency (subcarrier), increasing the symbol length to units and tens of milliseconds.

OFDM technology (Orthogonal Frequency Division Multiplexing) - Orthogonal frequency diversity, used to eliminate intersymbol interference in high-speed radio channels. Instead of passing n information symbols of a digital information signal (DIS) on one carrier frequency (Fig.4.12a), they are transmitted simultaneously on n subcarriers located in the radio channel band (Fig. 4.12b). Protective gaps of such duration are introduced between the symbols. T g so that the symbols arriving due to multipath propagation of radio waves with delay do not “crawl” onto the next ones. Moreover, the length of each character T b increases in comparison with the duration of the symbol in the original sequence in nT b / (T b + T g) once.

u cis (t)

n info symbols

u 1 t

u 2 t

u k t

u n t

Rice. 4.12. OFDM technology principle

The transmission of information symbols over a communication channel is the transmission of complex numbers. Signal constellations for different types of modulation are shown in Fig. 4.13.

Consider an example with the transmission of symbols with 16-QAM modulation (Fig. 4.14).

Rice. 4.13. Constellations of signals used in Wi-Fi, WiMA, LTE

Figure 4.14. 16-QAM signal constellation

Symbol S k transmitted on the k-th subcarrier can be represented as

where the amplitude of the symbol

and symbol phase

.

In the example in Fig. 4.14,

glad

Analytically, an OFDM signal is the sum of harmonics:

(4.1)

All subcarriers are harmonics of the fundamental frequency F 1: F k = kF 1 and the frequency F 1 is rigidly related to the duration of the symbol: F 1 = 1 / T b... Therefore, on the time interval T b fits k subcarrier waves F k... Each character S k can be viewed as a discrete sample of the spectrum on a subcarrier F k... Amplitude of the k-th subcarrier - a phase - When forming a signal u OFDM use the inverse (fast) Fourier transform procedure. In fig. 4.15 shows subcarriers with frequencies F 1 and F 2 and zero initial phases in the time interval T b.

Figure 4.15. Two subcarriers in the interval 0 - T b

The main problem when using OFDM technology is to provide a high signal-to-noise ratio in the receiver. Formally, when receiving signals n subcarriers should work n independent receivers. However, the spectra of signals on adjacent subcarriers are superimposed on each other (Fig. 4.16). Therefore, the reception of the OFDM signal and the selection of individual symbols are carried out using the direct (fast) Fourier transform procedure.

Figure 4.16. Spectrum of an OFDM signal fragment

Let's see how the receiver works k th subcarrier. It performs the direct Fourier transform procedure:

(4.2)

At the frequency F k = kF 1

On any other subcarrier F p= pF 1

Since the integral (area) of the sinusoid during one period is equal to 0 (Figure 4.17), and on the interval T b stacked integer │p-k│ periods of a sinusoid.

Figure 4.17. To determine the area of ​​a sinusoid

Therefore, with an accurate choice of the integration time, the interference from signals of other subcarriers is equal to 0. However, when calculating integrals (4.2), it is necessary to run functions with a zero initial phase, i.e. to provide coherent signal reception For this purpose, the access point (AP) in the radio channel down and and the subscriber terminal in the radio channel up, in addition to information symbols, transmit reference signals , i.e. pre-known complex numbers C (n), accepting which the receiver provides the necessary phase correction and scaling of the amplitudes of the received signals.

When transmitting (forming) an OFDM signal, an inverse discrete Fourier transform (inverse FFT) is used; at reception - the direct discrete Fourier transform (FFT). The OFDM signal is formed at a reduced frequency with the subsequent transfer of the spectrum to the frequency of the radio channel.

In the guard interval T g between the characters (Fig. 4.12), a cyclic prefix (CP - Cyclic Prefix) is transmitted - the end of the next character with a duration T g(fig. 4.18).

Rice. 4.18. Cyclic prefix OFDM symbol

This is done for reduction of intra-symbol interference (intra-symbol interference). If there was no cyclic prefix, then, when calculating the integral (4.2), the retarded rays arriving after the start of integration would lie on the time interval 0‒ T b, non-integer number of subcarrier periods. As a result, an error would appear when calculating integral (4.3), and integrals (4.4) would not vanish. When transmitting an SR with a beam delay of no more than T g, on the integration interval T b any subcarrier has an integer number of its periods and integrals (4.4) are equal to zero.

In the 802.11a standard, 48 subcarriers are used for information transmission (52 in total). The 4 subcarriers carry reference signals. Symbol duration T s= 3.2 μs, pause duration T p= 0.8 μs. Distance between adjacent frequencies MHz. With 2-PM modulation on each subcarrier, data rate (no guard coding)

When switching to multi-position modulation methods

Mbps,

Mbps.

The main characteristics of the 802.11a standard are shown in table. 4.4.

The new wireless standard IEEE 802.11n has been talked about for years. It is understandable, because one of the main drawbacks of the existing IEEE 802.11a / b / g wireless communication standards is that the data transfer rate is too low. Indeed, the theoretical throughput of the IEEE 802.11a / g protocols is only 54 Mbps, while the actual data transfer rate does not exceed 25 Mbps. The new standard for wireless communication IEEE 802.11n should provide a transmission speed of up to 300 Mbps, which looks very tempting against the background of 54 Mbps. Of course, the actual data transfer rate in the IEEE 802.11n standard, as shown by the test results, does not exceed 100 Mbps, but even in this case, the real data transfer rate is four times higher than in the IEEE 802.11g standard. The IEEE 802.11n standard has not yet been finally adopted (this should happen before the end of 2007), however, almost all manufacturers of wireless equipment have already begun to release devices compatible with the Draft version of the IEEE 802.11n standard.
In this article, we will look at the basic provisions of the new IEEE 802.11n standard and its main differences from the 802.11a / b / g standards.

We have already covered the 802.11a / b / g wireless standards in some detail on the pages of our magazine. Therefore, in this article we will not describe them in all details, however, in order for the main differences of the new standard from its predecessors to be obvious, we will have to make a digest of previously published articles on this topic.

Looking at the history of wireless standards used to create wireless local area networks(Wireless Local Area Network, WLAN), perhaps it is worth remembering the IEEE 802.11 standard, which, although it is no longer found in its pure form, is the progenitor of all other wireless communication standards for WLAN networks.

IEEE 802.11 standard

The 802.11 standard provides for the use of a frequency range from 2400 to 2483.5 MHz, that is, a range of 83.5 MHz wide, divided into several frequency subchannels.

The 802.11 standard is based on Spread Spectrum (SS) technology, which implies that an initially narrowband (in terms of spectrum width) useful information signal during transmission is converted in such a way that its spectrum is much wider than the original signal spectrum. Simultaneously with the broadening of the signal spectrum, a redistribution of the spectral energy density of the signal occurs - the signal energy is also "smeared" over the spectrum.

802.11 uses Direct Sequence Spread Spectrum (DSSS) technology. Its essence lies in the fact that in order to broaden the spectrum of an initially narrow-band signal, a chip sequence is embedded in each transmitted information bit, which is a sequence of rectangular pulses. If the duration of one chip pulse is n times less than the duration of the information bit, then the width of the spectrum of the converted signal will be in n times the spectrum width of the original signal. In this case, the amplitude of the transmitted signal will decrease by n once.

Chip sequences embedded in information bits are called noise-like codes (PN-sequences), which emphasizes the fact that the resulting signal becomes noise-like and difficult to distinguish from natural noise.

How to broaden the signal spectrum and make it indistinguishable from natural noise is understandable. For this, in principle, you can use an arbitrary (random) chip sequence. However, the question arises as to how to receive such a signal. After all, if it becomes noise-like, then it is not so easy, if not impossible, to extract a useful information signal from it. Nevertheless, this can be done, but for this you need to choose the appropriate chip sequence. Chip sequences used to spread the signal spectrum must satisfy certain autocorrelation requirements. Autocorrelation in mathematics means the degree of similarity of a function to itself at different points in time. If we choose such a chip sequence for which the autocorrelation function will have a pronounced peak for only one moment in time, then such an information signal can be isolated at the noise level. For this, the received signal is multiplied by the chip sequence in the receiver, that is, the autocorrelation function of the signal is calculated. As a result, the signal becomes narrowband again, so it is filtered in a narrow bandwidth equal to twice the transmission rate. Any interference that falls into the band of the original wideband signal, after multiplying by the chip sequence, on the contrary, becomes wideband and is cut off by filters, and only a part of the interference falls into the narrow information band, which is much less powerful in power than the interference acting at the receiver input.

There are quite a lot of chip sequences that meet the specified autocorrelation requirements, but the so-called Barker codes are of particular interest to us, since they are used in the 802.11 protocol. Barker codes have the best noise-like properties among the known pseudo-random sequences, which led to their widespread use. The protocols of the 802.11 family use a Barker code that is 11 chips long.

In order to transmit the signal, the information bit sequence in the receiver is added modulo 2 (mod 2) to the 11-chip Barker code using an XOR (exclusive OR) gate. Thus, logical one is transmitted by the direct Barker sequence, and logical zero is transmitted by the inverse sequence.

The 802.11 standard provides two speed modes - 1 and 2 Mbps.

At an information rate of 1 Mbit / s, the repetition rate of individual chips in the Barker sequence is 11x106 chips per second, and the spectrum width of such a signal is 22 MHz.

Considering that the width of the frequency range is 83.5 MHz, we find that in total in this frequency range, three non-overlapping frequency channels can be accommodated. The entire frequency range, however, is usually divided into 11 overlapping frequency channels of 22 MHz, spaced 5 MHz apart. For example, the first channel covers the frequency range from 2400 to 2423 MHz and is centered on the frequency of 2412 MHz. The second channel is centered around 2417 MHz, and the last, 11th channel, is centered around 2462 MHz. With this consideration, the 1, 6 and 11 channels do not overlap with each other and have a 3 MHz gap relative to each other. These three channels can be used independently of each other.

Differential Binary Phase Shift Key (DBPSK) is used to modulate a sinusoidal carrier signal at a data rate of 1 Mbps.

In this case, the information is encoded due to the phase shift of the sinusoidal signal with respect to the previous state of the signal. Binary phase modulation provides two possible values ​​for the phase shift - 0 and p. Then a logical zero can be transmitted in-phase signal (phase shift is 0), and one can be transmitted by a signal that is phase-shifted by p.

The data rate of 1 Mbit / s is mandatory in the IEEE 802.11 standard (Basic Access Rate), but an optional speed of 2 Mbit / s (Enhanced Access Rate) is also possible. To transfer data at this rate, the same DSSS technology is used with 11-chip Barker codes, but to modulate the carrier wave, a relative quadrature phase shift key (Differential Quadrature Phase Shift Key) is used.

In conclusion of the examination of the physical layer of the 802.11 protocol, we note that at an information rate of 2 Mbit / s, the repetition rate of individual chips of the Barker sequence remains the same, that is, 11x106 chips per second, and therefore the width of the transmitted signal spectrum does not change either.

IEEE 802.11b standard

The IEEE 802.11 standard was replaced by the IEEE 802.11b standard, which was adopted in July 1999. This standard is a kind of extension of the basic protocol 802.11 and, in addition to speeds of 1 and 2 Mbit / s, provides speeds of 5.5 and 11 Mbit / s, for which the so-called Complementary Code Keying (CCK) codes are used.

Complementary codes, or CCK-sequences, have the property that the sum of their autocorrelation functions for any cyclic shift other than zero is always zero, so they, like Barker codes, can be used to distinguish a signal against a background of noise.

The main difference between CCK sequences and the previously considered Barker codes is that there is not a strictly defined sequence by means of which either logical zero or one can be encoded, but a whole set of sequences. This circumstance makes it possible to encode several information bits in one transmitted symbol and thereby increases the information transmission rate.

The IEEE 802.11b standard deals with complex complementary 8-chip sequences defined on a set of complex elements taking the values ​​(1, –1, + j, –j}.

Complex signal representation is a convenient mathematical tool for representing a phase modulated signal. Thus, a sequence value equal to 1 corresponds to a signal in phase with the generator signal, and a sequence value equal to –1 corresponds to an antiphase signal; sequence value equal to j- a signal out of phase by p / 2, and a value equal to - j, - signal, phase-shifted by –p / 2.

Each element of the CCK sequence is a complex number, the value of which is determined using a rather complex algorithm. There are 64 sets of possible CCK-sequences in total, and the choice of each of them is determined by the sequence of input bits. Six input bits are required to uniquely select one CCK sequence. Thus, in the IEEE 802.11b protocol, each character is encoded using one of 64 possible eight-bit CKK sequences.

At a rate of 5.5 Mbit / s, 4 data bits are encoded in one symbol, and at a rate of 11 Mbit / s - 8 data bits. In this case, in both cases the symbol rate of transmission is 1.385x106 symbols per second (11/8 = 5.5 / 4 = 1.385), and considering that each symbol is specified by an 8-chip sequence, we obtain that in both cases the repetition rate of individual chips is 11x106 chips per second. Accordingly, the signal spectrum width at both 11 and 5.5 Mbit / s is 22 MHz.

IEEE 802.11g standard

The IEEE 802.11g standard, adopted in 2003, is a logical development of the 802.11b standard and assumes data transmission in the same frequency range, but at higher speeds. In addition, the 802.11g standard is fully compatible with 802.11b, which means that any 802.11g device must support 802.11b devices. The maximum data transfer rate in the 802.11g standard is 54 Mbps.

Two competing technologies were considered in the development of the 802.11g standard: the orthogonal frequency division OFDM borrowed from the 802.11a standard and proposed by Intersil, and the PBCC binary packet convolutional coding method proposed by Texas Instruments. As a result, the 802.11g standard contains a compromise solution: OFDM and CCK technologies are used as the basic ones, and the use of PBCC technology is optionally provided.

The idea behind Packet Binary Convolutional Coding (PBCC) is as follows. The incoming information bit sequence is converted in a convolutional encoder so that each input bit corresponds to more than one output bit. That is, the convolutional encoder adds some redundant information to the original sequence. If, for example, each input bit corresponds to two output bits, then we speak of convolutional coding at a rate r= 1/2. If each two input bits correspond to three output bits, then it will already be 2/3.

Any convolutional encoder is built on the basis of several sequentially connected storage cells and logical elements XOR. The number of memory cells determines the number of possible states of the encoder. If, for example, six memory cells are used in a convolutional encoder, then the encoder stores information about six previous signal states, and taking into account the value of the incoming bit, we find that seven bits of the input sequence are used in such an encoder. Such a convolutional encoder is called a seven-state encoder ( K = 7).

The output bits generated in the convolutional encoder are determined by XOR operations between the values ​​of the input bit and the bits stored in the memory cells, that is, the value of each generated output bit depends not only on the input information bit, but also on several previous bits.

PBCC technology uses seven-state convolutional encoders ( K= 7) with speed r = 1/2.

The main advantage of convolutional encoders is the noise immunity of the sequence they generate. The fact is that with coding redundancy, even in the event of receiving errors, the original bit sequence can be accurately recovered. A Viterbi decoder is used to restore the original bit sequence on the receiver side.

The dibit formed in the convolutional encoder is used in what follows as a transmitted symbol, but first it is subjected to phase modulation. Moreover, depending on the transmission rate, binary, quadrature or even eight-position phase modulation is possible.

Unlike DSSS technologies (Barker codes, CCK sequences), the convolutional coding technology does not use spectrum broadening through the use of noise-like sequences, however, spectrum broadening to standard 22 MHz is provided in this case as well. For this, variations of the possible QPSK and BPSK constellations are used.

The considered PBCC coding method is optionally used in the 802.11b protocol at 5.5 and 11 Mbps. Similarly, in the 802.11g protocol for transmission rates of 5.5 and 11 Mbps, this method is also used optionally. In general, due to the compatibility of the 802.11b and 802.11g protocols, the encoding technologies and rates provided by the 802.11b protocol are supported in the 802.11g protocol. In this regard, up to 11 Mbps, the 802.11b and 802.11g protocols coincide with each other, except that the 802.11g protocol provides speeds that are not in the 802.11b protocol.

Optionally, in the 802.11g protocol, PBCC technology can be used at transmission rates of 22 and 33 Mbps.

For a speed of 22 Mbit / s, compared to the PBCC scheme we have already considered, data transmission has two features. First of all, 8-position phase modulation (8-PSK) is used, that is, the phase of the signal can take eight different values, which makes it possible to encode three bits in one symbol. In addition, a Puncture encoder has been added to the scheme, except for the convolutional encoder. The meaning of this solution is quite simple: the redundancy of the convolutional encoder, equal to 2 (for each input bit there are two output ... For this, of course, you can develop an appropriate convolutional encoder, but it is better to add a special puncture encoder to the circuit, which will simply destroy the extra bits.

Let's say a puncture encoder removes one bit from every four input bits. Then every four incoming bits will correspond to three outgoing ones. The speed of such an encoder is 4/3. If such an encoder is used in tandem with a rate 1/2 convolutional encoder, then the total coding rate will already be 2/3, that is, each two input bits will correspond to three output bits.

As noted, PBCC is optional in the IEEE 802.11g standard, and OFDM is mandatory. In order to understand the essence of OFDM technology, let us consider in more detail the multipath interference arising from the propagation of signals in an open environment.

The effect of multipath signal interference is that multiple reflections from natural obstacles can result in the same signal reaching the receiver in different ways. But different propagation paths differ from each other in length, and therefore the signal attenuation will not be the same for them. Consequently, at the receiving point, the resulting signal is the interference of many signals having different amplitudes and offset from each other in time, which is equivalent to the addition of signals with different phases.

Multipath interference results in distortion of the received signal. Multipath interference is inherent in any type of signal, but it has a particularly negative effect on broadband signals, because when using a wideband signal, interference causes certain frequencies to add in phase, which leads to an increase in the signal, and some, on the contrary, out of phase, causing the signal to weaken at a given frequency.

When talking about multipath interference that occurs during signal transmission, two extreme cases are noted. In the first of them, the maximum delay between signals does not exceed the duration of one symbol and interference occurs within one transmitted symbol. In the second, the maximum delay between signals is greater than the duration of one symbol, therefore, as a result of interference, signals are added that represent different symbols, and so-called Inter Symbol Interference (ISI) occurs.

It is intersymbol interference that affects signal distortion most negatively. Since a symbol is a discrete state of a signal, characterized by the values ​​of the carrier frequency, amplitude and phase, the amplitude and phase of the signal change for different symbols, and therefore, it is extremely difficult to restore the original signal.

For this reason, at high data rates, a data coding technique called Orthogonal Frequency Division Multiplexing (OFDM) is used. Its essence lies in the fact that the stream of transmitted data is distributed over many frequency subchannels and the transmission is carried out in parallel on all such subchannels. In this case, a high transmission rate is achieved precisely due to the simultaneous transmission of data through all channels, while the transmission rate in a separate subchannel may be low.

Due to the fact that in each of the frequency subchannels the data transmission rate can be made not too high, prerequisites are created for effective suppression of intersymbol interference.

Frequency division multiplexing requires that the individual channel be narrow enough to minimize signal distortion, yet wide enough to provide the required bit rate. In addition, in order to economically use the entire channel bandwidth, divided into subchannels, it is desirable to place the frequency subchannels as close to each other as possible, but at the same time to avoid inter-channel interference in order to ensure their complete independence. Frequency channels that meet the above requirements are called orthogonal. The carriers of all frequency subchannels are orthogonal to each other. It is important that the orthogonality of the carrier signals guarantees the frequency independence of the channels from each other, and hence the absence of inter-channel interference.

The considered method of dividing a wideband channel into orthogonal frequency subchannels is called orthogonal frequency division multiplexing (OFDM). To implement it in transmitters, the inverse fast Fourier transform (IFFT) is used, which translates the pre-multiplexed into n-channels signal from time O th representation in frequency.

One of the key advantages of OFDM is the combination of high bit rate with effective multipath resistance. Of course, OFDM technology itself does not exclude multipath propagation, but it creates the prerequisites for eliminating the effect of intersymbol interference. The fact is that an integral part of OFDM technology is a Guard Interval (GI) - a cyclic repetition of the end of a symbol, added at the beginning of a symbol.

The guard interval creates pauses between individual symbols, and if its duration exceeds the maximum signal delay time as a result of multipath propagation, then intersymbol interference does not occur.

When using OFDM technology, the duration of the guard interval is one fourth of the duration of the symbol itself. In this case, the symbol has a duration of 3.2 μs, and the guard interval is 0.8 μs. Thus, the duration of the symbol together with the guard interval is 4 μs.

Speaking about the frequency orthogonal division of channels OFDM, applied at different rates in the 802.11g protocol, we have not yet touched on the question of the method of modulating the carrier signal.

802.11g uses BPSK and QPSK binary and quadrature phase modulation at low bit rates. When using BPSK modulation, only one information bit is encoded in one symbol, while with QPSK modulation, two information bits are encoded. BPSK modulation is used for data transmission at 6 and 9 Mbit / s, and QPSK modulation at 12 and 18 Mbit / s.

For transmission at higher rates, QAM (Quadrature Amplitude Modulation) is used, in which information is encoded by changing the phase and amplitude of the signal. The 802.11g protocol uses 16-QAM and 64-QAM modulation. The first modulation assumes 16 different signal states, which allows you to encode 4 bits in one symbol; the second - 64 possible signal states, which makes it possible to encode a sequence of 6 bits in one symbol. 16-QAM modulation is used at 24 and 36 Mbps, and 64-QAM is used at 48 and 54 Mbps.

In addition to the use of CCK-, OFDM- and PBCC-coding, the IEEE 802.11g standard also provides various options for hybrid coding.

In order to understand the essence of this term, remember that any transmitted data packet contains a header (preamble) with service information and a data field. When it comes to a packet in CCK format, it means that the header and frame data are transmitted in CCK format. Similarly, when using OFDM technology, the frame header and data are transmitted using OFDM coding. Hybrid coding implies that different coding technologies can be used for the frame header and data fields. For example, when CCK-OFDM technology is applied, the frame header is encoded using CCK codes, but the frame data itself is transmitted using OFDM multifrequency coding. Thus, CCK-OFDM technology is a kind of hybrid of CCK and OFDM. However, this is not the only hybrid technology - when using packet coding PBCC, the frame header is transmitted using CCK codes, and the frame data is encoded using PBCC.

IEEE 802.11a standard

The IEEE 802.11b and IEEE 802.11g standards discussed above refer to the 2.4 GHz frequency range (2.4 to 2.4835 GHz), and the IEEE 802.11a standard, adopted in 1999, assumes the use of a higher frequency range (from 5 , 15 to 5.350 GHz and 5.725 to 5.825 GHz). In the United States, this range is referred to as the Unlicensed National Information Infrastructure (UNII) range.

In accordance with FCC rules, the UNII frequency range is divided into three 100 MHz sub-bands, differing in terms of maximum emission power. The lower range (5.15 to 5.25 GHz) provides a power of only 50 mW, the middle (from 5.25 to 5.35 GHz) - 250 mW, and the upper (from 5.725 to 5.825 GHz) - 1 W. The use of three frequency subbands with a total width of 300 MHz makes the IEEE 802.11a standard the widest in the family of 802.11 standards and allows the entire frequency range to be divided into 12 channels, each of which is 20 MHz wide, eight of which lie in the 200 MHz range from 5 , 15 to 5.35 GHz, and the remaining four channels are in the 100 MHz range from 5.725 to 5.825 GHz (Fig. 1). In this case, the four high frequency channels, providing the highest transmission power, are used mainly for signal transmission outdoors.

Rice. 1. Division of the UNII range into 12 frequency sub-bands

The IEEE 802.11a standard is based on Orthogonal Frequency Division Multiplexing (OFDM). For channel separation, an inverse Fourier transform with a window of 64 frequency subchannels is applied. Since each of the 12 channels defined in 802.11a has a width of 20 MHz, each orthogonal frequency subchannel (subcarrier) is 312.5 kHz wide. However, out of 64 orthogonal subchannels, only 52 are used, with 48 of them being used for data transmission (Data Tones), and the rest - for transmission of service information (Pilot Tones).

In terms of modulation technique, the 802.11a protocol is not much different from 802.11g. At low bit rates, BPSK and QPSK binary and quadrature phase modulations are used to modulate the subcarriers. With BPSK modulation, only one information bit is encoded per symbol. Accordingly, when using QPSK modulation, that is, when the signal phase can take on four different values, two information bits are encoded in one symbol. BPSK modulation is used for data transmission at 6 and 9 Mbit / s, and QPSK modulation at 12 and 18 Mbit / s.

For higher transmission rates, the IEEE 802.11a standard uses 16-QAM and 64-QAM quadrature amplitude modulation. In the first case, there are 16 different signal states, which allows you to encode 4 bits in one symbol, and in the second, there are already 64 possible signal states, which allows you to encode a sequence of 6 bits in one symbol. 16-QAM modulation is used at 24 and 36 Mbit / s, and 64-QAM at 48 and 54 Mbit / s.

The information capacity of an OFDM symbol is determined by the modulation type and the number of subcarriers. Since 48 subcarriers are used for data transmission, the OFDM symbol capacity is 48 x Nb, where Nb is the binary logarithm of the number of modulation positions, or, more simply, the number of bits that are encoded in one symbol in one subchannel. Accordingly, the capacity of the OFDM symbol is 48 to 288 bits.

The processing sequence of input data (bits) in the IEEE 802.11a standard is as follows. Initially, the input data stream undergoes a standard scrambling operation. The data stream then enters the convolutional encoder. The convolutional coding rate (combined with puncture coding) can be 1/2, 2/3, or 3/4.

Since the convolutional coding rate can be different, then when using the same type of modulation, the data transmission rate is different.

Consider, for example, BPSK modulation, where the data rate is 6 or 9 Mbps. The duration of one symbol together with the guard interval is 4 μs, which means that the pulse repetition rate will be 250 kHz. Considering that one bit is encoded in each subchannel, and there are 48 such subchannels in total, we get that the total data rate will be 250 kHz x 48 channels = 12 MHz. If in this case the convolutional coding rate is 1/2 (one service bit is added for each information bit), the information rate will be half the full rate, that is, 6 Mbit / s. At a convolutional coding rate of 3/4, one overhead is added for every three information bits, so in this case the useful (information) rate is 3/4 of the full rate, that is, 9 Mbps.

Similarly, each type of modulation corresponds to two different transmission rates (Table 1).

Table 1. Relationship between transmission rates
and the type of modulation in the 802.11a standard

After convolutional coding, the bit stream is interleaved, or interleaved. Its essence lies in changing the order of the bits within one OFDM symbol. For this, the sequence of input bits is divided into blocks, the length of which is equal to the number of bits in the OFDM symbol (NCBPS). Further, according to a certain algorithm, a two-stage permutation of bits in each block is performed. In the first step, the bits are swapped so that adjacent bits are transmitted on non-contiguous subcarriers when transmitting the OFDM symbol. The bit swap algorithm at this stage is equivalent to the following procedure. Initially, a block of bits of length NCBPS is written row by row into a matrix containing 16 rows and NCBPS / 16 rows. Then the bits are read from this matrix, but already in rows (or in the same way as they were written, but from the transposed matrix). As a result of this operation, initially adjacent bits will be transmitted on non-contiguous subcarriers.

This is followed by a second bit swap step, the purpose of which is to ensure that adjacent bits do not end up in the least significant bits of the groups defining the modulation symbol in the constellation at the same time. That is, after the second stage of the permutation, adjacent bits are alternately in the most significant and least significant bits of the groups. This is done in order to improve the noise immunity of the transmitted signal.

After interleaving, the bit sequence is divided into groups according to the number of positions of the selected modulation type and OFDM symbols are formed.

The generated OFDM symbols undergo a fast Fourier transform, as a result of which the output in-phase and quadrature signals are generated, which are then subjected to standard processing - modulation.

IEEE 802.11n standard

The development of the IEEE 802.11n standard officially began on September 11, 2002, that is, a year before the final adoption of the IEEE 802.11g standard. In the second half of 2003, an IEEE 802.11n (802.11 TGn) Task Group was created to develop a new wireless standard for communication at speeds over 100 Mbps. Another target group, 802.15.3a, was also involved in the same task. By 2005, the processes of developing a single solution in each of the groups had reached an impasse. In the 802.15.3a group, there was a confrontation between Motorola and all other group members, and the members of the IEEE 802.11n group split into two roughly identical camps: WWiSE (World Wide Spectrum Efficiency) and TGn Sync. WWiSE was led by Aigro Networks, while TGn Sync was led by Intel. In each of the groups, for a long time, none of the alternative options could gain the necessary 75% of the votes for its approval.

After nearly three years of unsuccessful opposition and attempts to come up with a compromise solution that would suit everyone, the 802.15.3a group members voted almost unanimously to eliminate draft 802.15.3a. The members of the IEEE 802.11n project proved to be more flexible - they managed to agree and create a joint proposal that would suit everyone. As a result, on January 19, 2006, at a regular conference held in Kona, Hawaii, a preliminary (draft) specification of the IEEE 802.11n standard was approved. Of the 188 members of the working group, 184 were in favor of the standard and four abstained. The main provisions of the approved document will form the basis of the final specification of the new standard.

The IEEE 802.11n standard is based on OFDM-MIMO technology. Many of the technical details implemented in it are borrowed from the 802.11a standard, but the IEEE 802.11n standard provides for the use of both the frequency range adopted for the IEEE 802.11a standard and the frequency range adopted for the IEEE 802.11b / g standards. That is, devices that support the IEEE 802.11n standard can operate in either the 5 or 2.4 GHz frequency range, with the specific implementation depending on the country. For Russia, devices of the IEEE 802.11n standard will support the 2.4 GHz frequency range.

The increase in the transmission speed in the IEEE 802.11n standard is achieved, firstly, due to the doubling of the channel width from 20 to 40 MHz, and secondly, due to the implementation of MIMO technology.

MIMO (Multiple Input Multiple Output) technology involves the use of multiple transmit and receive antennas. By analogy, traditional systems, that is, systems with one transmitting and one receiving antenna, are called SISO (Single Input Single Output).

Theoretically, a MIMO system with n transmitting and n receiving antennas is capable of providing a peak bandwidth of n times more than SISO systems. This is achieved by the transmitter splitting the data stream into independent bit sequences and transmitting them simultaneously using an array of antennas. This transmission technique is called spatial multiplexing. Note that all antennas transmit data independently of each other in the same frequency range.

Consider, for example, a MIMO system consisting of n transmitting and m receiving antennas (Fig. 2).

Rice. 2. The principle of the implementation of MIMO technology

The transmitter in such a system sends n independent signals using n antennas. On the receiving side, each of m antennas receive signals that are superposition n signals from all transmitting antennas. Thus, the signal R1 received by the first antenna can be represented as:

Writing down similar equations for each receiving antenna, we get the following system:

Or, rewriting given expression in matrix form:

where [ H] - transfer matrix describing the MIMO communication channel.

In order for the decoder on the receiving side to be able to correctly reconstruct all signals, it must first of all determine the coefficients hij characterizing each of m x n transmission channels. To determine the coefficients hij MIMO uses a packet preamble.

Having determined the coefficients of the transfer matrix, one can easily restore the re this signal:

where [ H] –1 - matrix inverse to the transfer matrix [ H].

It is important to note that in MIMO technology, the use of several transmitting and receiving antennas allows increasing the throughput of a communication channel by implementing several spatially separated subchannels, while data is transmitted in the same frequency range.

MIMO technology does not affect the data encoding method in any way and, in principle, can be used in combination with any methods of physical and logical data encoding.

For the first time, MIMO technology was described in the IEEE 802.16 standard. This standard allows the use of MISO technology, that is, several transmitting antennas and one receiving one. The IEEE 802.11n standard allows up to four antennas at the access point and wireless adapter... Mandatory mode implies support for two antennas at the access point and one antenna and wireless adapter.

The IEEE 802.11n standard provides for both standard 20 MHz communication channels and double-wide channels. However, the use of 40 MHz channels is an optional feature of the standard, since the use of such channels may be contrary to the laws of some countries.

The 802.11n standard provides two transmission modes: standard transmission mode (L) and high throughput (HT) mode. In traditional transmission modes, 52 OFDM frequency subchannels (subcarriers) are used, of which 48 are used for data transmission, and the rest are used for transmission of service information.

In modes with increased capacity with a channel width of 20 MHz, 56 frequency subchannels are used, of which 52 are used for data transmission, and four channels are pilot channels. Thus, even when using a 20 MHz channel, increasing the frequency subchannels from 48 to 52 can increase the transmission rate by 8%.

When using a channel of double width, that is, a channel with a width of 40 MHz, in the standard transmission mode, broadcasting is actually carried out on a dual channel. Accordingly, the number of subcarriers is doubled (104 subchannels, of which 96 are informational). This increases the transmission speed by 100%.

When using a 40MHz channel and high-throughput mode, 114 frequency subchannels are used, of which 108 are data subchannels, and six are pilot ones. Accordingly, this allows you to increase the transmission speed by 125%.

Table 2. Relationship between bit rates, modulation type
and the convolutional coding rate in the 802.11n standard
(20 MHz channel, HT-mode (52 frequency subchannels))

Two more circumstances, due to which the transmission rate increases in the IEEE 802.11n standard, are a reduction in the duration of the GI guard interval in OGDM symbols from 0.8 to 0.4 μs and an increase in the convolutional coding rate. Recall that in the IEEE 802.11a protocol, the maximum convolutional coding rate is 3/4, that is, one more bit is added to every three input bits. In the IEEE 802.11n protocol, the maximum convolutional coding rate is 5/6, that is, every five input bits in the convolutional encoder are converted into six output bits. The relationship between transmission rates, modulation type and convolutional coding rate for a standard 20 MHz channel is shown in Table. 2.

The ubiquity of wireless networks, the development of hotspot infrastructure, the emergence of mobile technologies with an embedded wireless solution (Intel Centrino) has led end users (not to mention enterprise customers) to pay more and more attention to wireless solutions. Such solutions are considered primarily as a means of deploying mobile and fixed wireless local area networks and as a means of online access to the Internet.

however, the end user who is not a network administrator is usually not very knowledgeable about network technologies, so it is difficult for him to do right choice when purchasing a wireless solution, especially considering the variety of products on offer today. The rapid development of wireless technology has led to the fact that users, not having time to get used to one standard, are forced to switch to another, with even higher transmission rates. We are, of course, talking about a family of wireless communication protocols known as IEEE 802.11, which includes the protocols 802.11, 802.11b, 802.11b +, 802.11a, 802.11g, 802.11g +, and a new standard, 802.11n, is on the horizon. And if you add such security and QoS protocols as 802.11e, 802.11i, 802.11h, etc. to this numerous family, it becomes clear that it is not easy to figure it out.

To make life easier for those who want to join the world of wireless communication, but do not know where to start, we decided to compile a quick guide, after reading which the reader will be able to understand the main differences between the wireless protocols of the 802.11 family and understand the basic principles of wireless networks.

Physical layer of the 802.11 family of protocols

The main difference between the standards of the 802.11 family lies in the way information is encoded and in the resulting differences in transmission / reception rates. All wireless protocols are based on Spread Spectrum (SS) technology, which implies that the initially narrow-band (in the spectrum width) useful information signal during transmission is converted in such a way that its spectrum is much wider than the spectrum of the original signal, that is, the signal spectrum as if smeared over the frequency range. Simultaneously with the broadening of the signal spectrum, a redistribution of the spectral energy density of the signal occurs - the signal energy is also "smeared" over the spectrum. As a result, the maximum power of the converted signal is significantly lower than the power of the original signal. In this case, the level of the useful information signal can literally be compared with the level of natural noise, as a result of which the signal becomes, in a sense, "invisible" - it is simply lost at the level of natural noise.

For unlicensed use in Europe and the United States (it is in this spectral range that the protocols of the 802.11 family operate), a radio band from 2400 to 2483.4 MHz is allocated, intended for use in industry, science and medicine (Industry, Science and Medicine, ISM) and called ISM- range), as well as from 5725 to 5875 MHz, but at the same time the transmitter power is strictly regulated, which is limited to 100 mW in Europe (ETSI limits) and 1 W in the USA (FCC limits). Spectrum broadening technology is used to organize the joint use of the radio range in such harsh conditions. 802.11b / g protocols use Direct Sequence Spread Spectrum (DSSS) technology.

IEEE 802.11 standard

The very first wireless networking standard that served as the basis for a whole family of wireless protocols was IEEE 802.11. Today there are no solutions based exclusively on this protocol, but it deserves a separate discussion, if only because it is included as a subset of the 802.11b and 802.11g protocols.

The 802.11 standard provides for the use of the frequency range from 2400 to 24 835 MHz and transmission rates of 1 and 2 Mbps. The data is encoded using the DSSS method with 11-chip Barker codes. At an information rate of 1 Mbit / s, the repetition rate of individual chips in the Barker sequence is 11Ѕ106 chips / s, and the spectrum width of such a signal is 22 MHz.

A Differential Binary Phase Shift Key (DBPSK) is used to modulate a sinusoidal carrier signal (a process required to fill a carrier signal).

The data rate of 1 Mbit / s is mandatory in the IEEE 802.11 standard (Basic Access Rate), but an optional speed of 2 Mbit / s (Enhanced Access Rate) is also possible. To transmit data at this rate, DSSS technology with 11-chip Barker codes is used, but to modulate the carrier wave, Differential Quadrature Phase Shift Key is used.

At an information rate of 2 Mbit / s, the repetition rate of individual chips of the Barker sequence remains the same, that is, 11Ѕ106 chips / s, and, therefore, the width of the spectrum of the transmitted signal does not change either.

IEEE 802.11b standard

The IEEE 802.11b protocol, adopted in July 1999, is a kind of extension of the basic 802.11 protocol and, in addition to the speeds of 1 and 2 Mbit / s, provides speeds of 5.5 and 11 Mbit / s. To operate at speeds of 5.5 and 11 Mbit / s, instead of noise-like Barker sequences, so-called eight-chip Complementary Code Keying (CCK) sequences are used to spread the spectrum.

The use of CCK codes can encode 8 bits per symbol at 11 Mbps and 4 bits per symbol at 5.5 Mbps. Moreover, in both cases, the symbol rate is 1.385Ѕ106 symbols per second (11/8 = 5.5 / 4 = 1.385).

The phase values ​​defining the elements of the CCK sequence depend on the sequence of the input information bits. At a transmission rate of 11 Mbit / s, knowledge of 8 bits (4 dibits) of the input data is required to unambiguously determine the CCK sequence. The first dibit of the input data determines the phase shift of the entire symbol relative to the previous one, and the remaining 6 bits are used to set the CCK sequence itself. Since 6 data bits can have 64 different combinations, the IEEE 802.11b protocol uses one of 64 possible eight-bit CKK sequences to encode each character, and this allows 6 bits to be encoded in one transmitted symbol. Since each symbol is additionally shifted in phase relative to the previous symbol depending on the value of the first dibit and the phase shift can take four values, we obtain that 8 information bits are encoded in each symbol.

At a transmission rate of 5.5 Mbit / s, 4 bits are already encoded in one symbol, which determines twice the information rate. At such a transmission rate, the already considered CCK sequences are used, which are formed according to the same rules - the only difference is the number of used CCK sequences and the rule for their selection.

To set all members of the CCK-sequence, 4 input information bits are used, that is, 2 dibits. The first dibit, as before, specifies the phase shift value of the whole symbol, and the second dibit is used to select one of four possible CCK sequences. Considering that each symbol is additionally shifted in phase relative to the previous one by one of four possible values, then this makes it possible to encode 4 information bits in one symbol.

Considering the possible transmission rates of 5.5 and 11 Mbit / s in the 802.11b protocol, we have so far ignored the question of why we need a speed of 5.5 Mbit / s, if the use of CCK sequences allows data to be transmitted at a speed of 11 Mbit / s ... In theory, this is true, but only if the interference environment is not taken into account. In real conditions, the noise level of the transmission channels and, accordingly, the ratio of noise and signal levels may turn out to be such that transmission at a high information rate (that is, when many information bits are encoded in one symbol) will become impossible due to their erroneous recognition. Without going into mathematical details, we only note that the higher the noise level of the communication channels, the lower the information transmission rate. At the same time, it is important that the receiver and transmitter correctly analyze the interference environment and select an acceptable transmission rate.

In addition to CCK sequences, the 802.11b protocol optionally provides an alternative coding method at transmission rates of 5.5 and 11 Mbps - PBCC packet convolutional coding. And it is this coding mode that formed the basis of the 802.11b + protocol - an extension of the 802.11b protocol. Actually, the 802.11b + protocol as such does not officially exist, however, this extension was at one time supported by many manufacturers. wireless devices... The 802.11b + protocol provides for another data rate of 22 Mbps using PBCC technology.

At a bit rate of 5.5 Mbit / s, BPSK is used to modulate the dibit generated by a convolutional encoder with a convolutional coding rate 1/2, and at a rate of 11 Mbit / s, QPSK is used. In this case, for a speed of 11 Mbit / s, one input bit is encoded in each symbol and the bit rate corresponds to the symbol rate, and at a rate of 5.5 Mbit / s, the bit rate is equal to half the symbol rate (since each input bit in this case match two output characters). Therefore, for both 5.5 Mbps and 11 Mbps, the symbol rate is 11Ѕ106 symbols per second.

For a speed of 22 Mbit / s, compared to the PBCC scheme we have already considered, data transmission has two differences. Firstly, 8-position phase modulation (8-PSK) is used, that is, the signal phase can take eight different values, which makes it possible to encode 3 bits in one symbol. Secondly, in addition to the convolutional encoder, a puncture encoder (Puncture) was added to the scheme for the following reason: the redundancy of the convolutional encoder equal to 2 (there are two output bits for each input bit) is high enough and, under certain conditions of the interference environment, is unnecessary, so redundancy can be reduced so that, for example, every two input bits correspond to three output bits. For this purpose, it is, of course, possible to develop a corresponding convolutional encoder with a convolutional coding rate of 2/3, but it is better to add a special puncture encoder to the scheme, which will simply delete the extra bits.

Having dealt with the principle of operation of the puncture encoder, let's return to the consideration of PBCC coding at a rate of 22 Mbit / s in the 802.11b + protocol.

The convolutional encoder (r = 1/2) receives data at a rate of 22 Mbps. After adding redundancy in a convolutional encoder, 44 Mbps bits are fed to a punctuated encoder, in which the redundancy is reduced so that there are three output bits for every four input bits. Consequently, after the punctured encoder, the flow rate will already be 33 Mbit / s (not the informational rate, but the total rate taking into account the added redundant bits). The resulting sequence is sent to an 8-PSK phase modulator where every three bits are packed into one symbol. In this case, the transmission rate will be 11Ѕ106 symbols per second, and the information rate - 22 Mbit / s.

The relationship between the transmission rates and the type of coding in the 802.11b / b + standard is shown in Table. 1.

* 22Mbps rate applies to 802.11b + protocol only.

IEEE 802.11g standard

The 802.11g standard is a logical evolution of the 802.11b standard and assumes data transmission in the same frequency range, but at higher speeds. In addition, the 802.11g standard is fully compatible with 802.11b, which means that any 802.11g device must support 802.11b devices. The maximum transmission rate in the 802.11g standard is 54 Mbps.

In the development of 802.11g, two competing technologies were considered: the orthogonal frequency division OFDM method and the binary packet convolutional coding method PBCC, optionally implemented in the 802.11b standard. As a result, the 802.11g standard is based on a compromise solution: OFDM and CCK technologies are used as the basic ones, and the use of PBCC technology is optionally provided.

In the 802.11g protocol, PBCC coding technology can optionally (but not necessarily) be used at 5.5 rates; eleven; 22 and 33 Mbps. In general, in the standard itself, transmission rates of 1 are mandatory; 2; 5.5; 6; eleven; 12 and 24 Mbps, and higher bit rates of 33, 36, 48 and 54 Mbps are optional. In addition, the same transmission rate can be realized with different modulation techniques. For example, a transmission rate of 24 Mbps can be achieved with both multi-frequency OFDM coding and hybrid CCK-OFDM coding techniques.

The only thing we haven't mentioned yet is the hybrid coding technique. To understand the essence of this term, remember that any transmitted data packet contains a header / preamble with service information and a data field. When it comes to a packet in CCK format, it means that the header and frame data are transmitted in CCK format. Similarly, when using OFDM technology, the frame header and data are transmitted using OFDM coding. With CCK-OFDM technology, the frame header is encoded using CCK codes, but the frame data itself is transmitted using OFDM multifrequency coding. Thus, CCK-OFDM technology is a kind of hybrid of CCK and OFDM. However, CCK-OFDM is not the only hybrid technology: with packet coding PBCC, the frame header is transmitted using CCK codes and the frame data is encoded using PBCC.

IEEE 802.11a standard

The 802.11b and 802.11g standards discussed above refer to the 2.4 GHz frequency range (2.4 to 2.4835 GHz), while the 802.11a standard assumes the use of a higher frequency range (from 5.15 to 5.350 GHz and from 5.725 to 5.825 GHz). In the United States, this range is referred to as the Unlicensed National Information Infrastructure (UNII) range.

In accordance with FCC rules, the UNII frequency range is divided into three 100 MHz sub-bands, differing in terms of maximum emission power. The lower range (5.15 to 5.25 GHz) provides power of only 50 mW, the middle range (5.25 to 5.35 GHz) - 250 mW, and the upper range (from 5.725 to 5.825 GHz) - up to 1 W. The use of three frequency sub-bands with a total width of 300 MHz makes the 802.11a standard the widest in the family of 802.11 standards and allows you to split the entire frequency range into 12 20 MHz channels, eight of which lie in the 200 MHz range from 5.15 to 5.35 GHz and the other four are in the 100 MHz range from 5.725 to 5.825 GHz. In this case, the four high frequency channels, providing the highest transmission power, are used mainly for signal transmission outdoors.

The 802.11a protocol is based on Orthogonal Frequency Division Multiplexing (OFDM) techniques. For channel separation, an inverse Fourier transform with a window of 64 frequency subchannels is used. Since each of the 12 channels defined in 802.11a is 20 MHz wide, each orthogonal frequency sub-channel is 312.5 kHz wide. However, out of 64 orthogonal subchannels, only 52 are used, with 48 of them being used for data transmission (Data Tones), and the rest - for transmission of service information (Pilot Tones).

In terms of modulation technique, the 802.11a protocol is not much different from 802.11g. BPSK and QPSK are used at low bit rates, while 16-QAM and 64-QAM are used at high bit rates. In addition, convolutional coding is provided in the 802.11a protocol to improve noise immunity. Since the convolutional coding rate can be different, then when using the same type of modulation, the transmission rate is different.

In the OFDM method, the duration of one symbol together with the guard interval is 4 μs, and therefore the pulse repetition rate will be 250 kHz. Considering that one bit is encoded in each subchannel, and there are 48 such subchannels in total, we get that the total transmission rate will be 250 kHz 48 channels = 12 MHz. If the rate of the convolutional encoder is 1/2, then the data bit rate will be 6 Mbps. If the convolutional coding rate is 3/4, then the data bit rate will be 9 Mbps. In total, the 802.11a protocol provides for the use of eight different transmission modes, differing from each other by the rate, the type of modulation and the used convolutional coding rate (Table 2). At the same time, we emphasize that in the 802.11a protocol itself, only the speeds of 6, 12 and 24 Mbit / s are mandatory, and all others are optional.

Shared access mechanisms in 802.11 networks

So far, considering various wireless protocols of the 802.11 family, we have concentrated precisely on the physical (PHY) layer, which determines the methods of encoding / decoding and modulating / demodulating a signal during its transmission and reception. However, issues such as media sharing regulation are defined at a higher level - the media access layer, which is called the MAC layer (Media Access Control). It is at the MAC level that the rules for the sharing of the transmission medium by several nodes simultaneously are established. wireless network.

The need for regulatory rules is clear. Imagine what it would be like if every node of the wireless network, without observing any rules, began to transmit data over the air. As a result of the interference of several such signals, the nodes for which the sent information was intended could not only receive it, but generally understand that this information is addressed to them. That is why it is necessary to have strict regulatory rules that should govern the collective access to the data transmission medium. Such rules of collective access can be figuratively compared to the rules of the road, which regulate the joint use of roads by all road users.

The 802.11 MAC layer defines two types of shared media access: the Distributed Coordination Function (DCF) and the Point Coordination function (PCF).

Distributed Coordination Function DCF

At first glance, organizing shared access to the data transmission medium is not difficult: for this it is only necessary to ensure that all nodes transmit data only when the environment is free, that is, when none of the nodes is transmitting data. However, such a mechanism will inevitably lead to collisions, since there is a high probability that two or more nodes at once, trying to access the data transmission medium, will decide that the medium is free and start simultaneous transmission. That is why it is necessary to develop an algorithm that can reduce the likelihood of collisions and at the same time guarantee all network nodes equal access to the data transmission medium.

Distributed Coordination Function (DCF), based on Carrier Sense Multiple Access / Collision Avoidance (CSMA / CA), is one option for such peer-to-peer media access. With such an organization, each node, before starting transmission, listens to the medium, trying to detect the carrier signal, and only if the medium is free can start transmitting data.

However, as we have already noted, in this case, there is a high probability of collisions, and in order to reduce the likelihood of such situations, the Collision Avoidance (CA) mechanism is used. The essence of this mechanism is as follows. Each node in the network, making sure that the medium is free, waits for a certain period of time before starting transmission. This interval is random and consists of two components: a mandatory DIFS interval (DCF Interframe Space) and a randomly selected countdown interval (Backoff Time). As a result, each network node waits for a random period of time before starting transmission, which, naturally, significantly reduces the likelihood of collisions, since the probability that two network nodes will wait for the same period of time is extremely small.

To ensure that all nodes in the network have equal access to the data transmission medium, it is necessary to appropriately determine the algorithm for choosing the length of the countdown interval. Although this interval is random, it is selected from a set of some discrete time intervals, that is, it is equal to an integer number of elementary time intervals, called time slots (SlotTime). To select the countdown interval, each network node forms a so-called Contention Window (CW), which is used to determine the number of time slots during which the station waited before transmitting. The minimum window size is 31 timeslots and the maximum is 1023 timeslots.

When a host tries to access the data transmission medium, after the mandatory waiting period for DIFS, a countdown procedure is started, that is, the countdown of the time slot counter starts from the selected window value. If during the entire waiting period the medium remains free, then the node starts transmitting.

After successful transmission, the window is formed again. If, during the waiting time, another node in the network starts transmitting, the value of the countdown counter stops and data transmission is postponed. After the environment becomes free, this node starts the countdown procedure again, but with a smaller window size determined by the previous countdown counter value, and, accordingly, with a shorter timeout value. Moreover, it is obvious that the more times a node postpones transmission due to a busy environment, the higher the likelihood that the next time it will get access to the data transmission medium.

The considered algorithm for implementing shared access to the data transmission medium guarantees equal access of all network nodes to the medium. However, with this approach, the likelihood of collisions still exists. It is clear that it is possible to reduce the likelihood of collisions by increasing maximum size the generated window, however, this will increase the transmission delay times, thereby reducing the network performance. Therefore, the DCF method uses the following algorithm to minimize collisions. After each successful reception of a frame, the receiving side after a short period of SIFS (Short Interframe Space) confirms successful reception by sending a response receipt - an ACK (ACKnowledgement) frame. If a collision occurs during data transmission, then the transmitting side does not receive an ACK frame about successful reception, and then the window size for the transmitting node is almost doubled. So, if for the first transmission the window size is 31 slots, then for the second transmission attempt it is already 63, for the third - 127, for the fourth - 255, for the fifth - 511, and for all subsequent ones - 1023 slots. Consequently, the increase in the size of the window occurs dynamically, as the number of collisions increases, which allows, on the one hand, to reduce time delays, and on the other hand, to reduce the likelihood of collisions.

The considered mechanism for regulating collective access to the data transmission medium has one bottleneck. This is the so-called hidden node problem. Due to the presence of natural obstacles, it is possible that two network nodes cannot hear each other directly; such nodes are called hidden nodes. To solve the problem of hidden nodes, the DCF function optionally provides the possibility of using the RTS / CTS algorithm.

RTS / CTS Algorithm

In accordance with the RTS / CTS algorithm, each network node, before sending data, first sends a special short message, which is called RTS (Ready-To-Send) and means that this node is ready to send data. Such an RTS message contains information about the duration of the upcoming transmission and about the addressee and is available to all nodes in the network (unless, of course, they are hidden from the sender). This allows other nodes to delay transmission for a time equal to the advertised message duration. The receiving station, having received the RTS signal, responds by sending a CTS (Clear-To-Send) signal, indicating that the station is ready to receive information. After that, the transmitting station sends a data packet, and the receiving station must transmit an ACK frame, confirming the error-free reception.

Now consider the situation when the network consists of four nodes: A, B, C and D (Fig. 1). Suppose that node C is within the reach of only node A, node A is within reach of nodes C and B, node B is within reach of nodes A and D, and node D is within reach of only node B, that is, in the network there are hidden nodes: node C is hidden from nodes B and D, and node A is hidden from node D.

In such a network, the RTS / CTS algorithm makes it possible to cope with the problem of collisions, which cannot be solved by the considered basic method of organizing shared access in DCF. Let node A try to transmit data to node B; To do this, it sends an RTS signal, which, in addition to Node B, also receives Node C, but does not receive Node D. Node C, having received this signal, is blocked, that is, it suspends attempts to transmit the signal until the end of transmission between nodes A and B. Node B, in response to the received RTS signal, sends a CTS frame, which is received by nodes A and D. Node D, having received this signal, is also blocked for the duration of transmission between nodes A and B.

The RTS / CTS algorithm has, however, its pitfalls, which in certain situations lead to a decrease in the efficiency of using the data transmission medium. For example, sometimes such a phenomenon as the spread of the effect of false blocking of nodes is possible, which ultimately can lead to a stupor in the network.

Consider, for example, the network shown in Fig. 2. Let Node B try to transmit data to Node A by sending it an RTS frame. Since this frame is also received by node C, the latter is blocked for the duration of transmission between nodes A and B. Node D, trying to transmit data to node C, sends an RTS frame, but since node C is blocked, it does not receive a response and begins the countdown procedure with an increased window size. At the same time, the RTS frame sent by the node D is received by the node E, which, incorrectly assuming that this will be followed by a data transfer from the node D to the node C, is blocked. However, this is a false blocking, since there is really no transmission between nodes D and C, and this phenomenon of false blocking of nodes can lead to a short-term stupor of the entire network.

Centralized coordination function PCF

The above DCF distributed coordination mechanism is basic for 802.11 protocols and can be used both in wireless networks operating in Ad-Hoc mode and in networks operating in Infrastructure mode, that is, in networks whose infrastructure includes an access point (Access Point, AP ).

However, for networks in Infrastructure mode, a slightly different mechanism for provisioning shared access, known as the Point Coordination Function (PCF), is more natural. Note that the PCF mechanism is optional and applies only to networks with an access point. In the case of the PCF mechanism, the access point is the Point Coordinator (PC). The coordination center is responsible for managing the collective access of all other network nodes to the data transmission medium based on a specific polling algorithm or based on the priorities of the network nodes. The Coordination Center polls all the network nodes included in its list, and on the basis of this poll organizes the data transfer between all network nodes. It should be noted that this approach completely eliminates concurrent access to the medium, as in the case of the DCF mechanism, and makes collisions impossible.

The centralized coordination function does not replace the distributed coordination function, but rather complements it by overlapping it. Within a certain period of time, the PCF mechanism is implemented, then DCF, and then everything is repeated anew.

To be able to alternate between the PCF and DCF modes, it is necessary that the access point performing the functions of the coordination center and implementing the PCF mode has priority access to the data transmission medium. This can be done by using concurrent media access (as in the DCF method), but allow the coordination center to use a latency less than DIFS. In this case, if the coordination center tries to access the medium, then it waits for the end of the current transmission, and since the minimum standby mode is determined for it after detecting "silence" in the air, it is the first to access the medium.

IEEE 802.11 - a set of communication standards for communication in the wireless local area network area of ​​frequency ranges 0.9; 2.4; 3.6 and 5 GHz.

Better known to users by the name Wi-Fi, which is actually a brand proposed and promoted by the Wi-Fi Alliance. It became widespread thanks to the development in mobile electronic computing devices: PDAs and laptops.

Institute of Electrical and Electronics Engineers - IEEE (I triple E - "I triple and") is an international non-profit association of specialists in the field of engineering, a world leader in the development of standards for radio electronics and electrical engineering.

Initially, the IEEE 802.11 standard assumed the ability to transmit data over a radio channel at a speed of no more than 1 Mbit / s and, optionally, at a speed of 2 Mbit / s. One of the first high-speed wireless networking standards - IEEE 802.11a - defines transmission rates up to 54 Mbps gross. The working range of the standard is 5 GHz.

Contrary to its name, the IEEE 802.11b standard, adopted in 1999, is not a continuation of the 802.11a standard, since they use various technologies: DSSS (more precisely, its improved version of HR-DSSS), DSSS technology (Direct Sequence Spread Spectrum), in 802.11b against OFDM, OFDM (Orthogonal frequency-division multiplexing), in 802.11a. The standard provides for the use of the unlicensed 2.4 GHz frequency band. Transfer rates up to 11 Mbps.

IEEE 802.11b products from various manufacturers are tested and certified by the Wireless Ethernet Compatibility Alliance (WECA), now known as the Wi-Fi Alliance. Compatible wireless products tested by the Wi-Fi Alliance program may carry the Wi-Fi mark.

For a long time, IEEE 802.11b was the prevalent standard on which most wireless LANs were built. Now its place has been taken by the IEEE 802.11g standard, which is gradually being replaced by the high-speed IEEE 802.11n.

The draft IEEE 802.11g standard was approved in October 2002. This standard uses the 2.4 GHz frequency band, providing connection speeds of up to 54 Mbps (gross), and thus surpassing the IEEE 802.11b standard, which provides connection speeds of up to 11 Mbps. In addition, it guarantees backward compatibility with the 802.11b standard. Backward compatibility of the IEEE 802.11g standard can be implemented in the DSSS modulation mode, and then the connection speed will be limited to eleven megabits per second, or in the OFDM modulation mode, in which the speed can reach 54 Mbps. Thus, this standard is the most suitable for the construction of wireless networks.

The massiveness of wireless communication technologies in our time is simply amazing. The IEEE 802.11 technology deserves a separate topic. It is almost impossible to find a place in the city where a laptop or tablet “did not find” at least one Wi-Fi network... In any cafe, multi-storey building or office, you can find several broadcasts. It is very difficult to underestimate the convenience that this technology provides us.

The Wi-Fi that we use today has come a long and thorny path for the convenience of the user, to which we are all accustomed to. Many standards with their own transmission characteristics and frequency ranges formed something without which the life of an IT specialist or simply modern man it is hard to imagine. We will not plunge into history, but only note that at the moment 802.11g and 802.11n standards are actively used, which operate in the 2.4 GHz band. There are many sources of interference to wireless networks in the home, but they are not the main problem. The culprit for most of the inconvenience is the Wi-Fi point itself, or to be more precise, a large number of them close to each other. Due to the popularity of this technology and the high saturation of broadcasting places, users may encounter some difficulties in their work. A large congestion of wireless networks can cause the effect of overlapping frequencies, which causes a decrease in transmission speed or loss of connection altogether. This significant disadvantage caused by the popularization of wireless technology was one of the big bells at WECA for the implementation of the IEEE 802.11ac standard.