IEEE 802.11 synopsis

IEEE 802.11 Industry overview Standards

Wireless Ethernet

The IEEE 802.11 specification is a WLAN standard that defines a set of specifications for physical layers (PHYs) and a medium access control (MAC) layer. On the high-rate side, the 802.11b spec defines a set of re-quirements for a new PHY as an ex-tension to the legacy direct sequence spread spectrum (DSSS) PHY.

IEEE 802.11 MAC layer

The MAC layer of the IEEE 802.11b specifications handles many functions in a WLAN system. On one hand, the 802.11b MAC layer (the second layer of the OSI model) serves as the interface between the PHY and the host device. On the other, the MAC supports both ad-hoc and client/server infrastructure networks.

The IEEE 802.11b MAC layer is similar to the IEEE 802.3 Ethernet wired LAN standard. IEEE 802.3 Ethernet uses carrier sense multiple access/collision detection (CSMA/CD) as its baseline protocol. IEEE 802.11, however, employs carrier sense multiple access/collision avoidance (CSMA/CA) as its protocol mechanism because it is difficult to detect collisions over a wireless medium.

CSMA/CA is referred to as the distributed coordination function (DCF) and uses a random back-off algorithm to avoid collisions over the medium. This algorithm is referred to as the virtual carrier-sense mechanism.

The CSMA/CA protocol also includes an optional point coordination function (PCF), which is used to set up an access point as a point coordinator. In this function, the point coordinator assigns priority to each client in a given transmission frame. The PCF option is extremely powerful because it can be used for time-bounded services such as voice, voice over IP (VoIP), and multimedia traffic.

In addition to the CSMA/CA protocol, the MAC layer supports authentication network management and privacy. Specifically, the MAC layer supports two flavors of authentication: open and shared key.

Of the two authentication schemes, the shared key approach has captured the most headlines in the WLAN design community. Shared key authentication uses wired equivalent privacy (WEP) to secure the WLAN system. In this authentication approach, the access points and clients are configured using shared encryption keys. A 64-bit encryption algorithm known as RC4 is used to add a level of security and privacy between packets.

Frame structure

The 802.11b spec also details frame format structure. This structure is designed to support DCF and PCF operation. The frame format consists of a MAC header and a frame body. The frame body contains the MAC service data unit (MSDU) from the higher layer protocols and has a maximum length of 2,048 bytes. The frame structure also sports a MAC header, which contains information on the frame type, destination addresses, and the length of the data payload.

Within the frame structure are a series of control frames. Overall, IEEE 802.11b supports six types of control frames: RTS, CTS, AC, PS-Poll, CF-End, and CF+End+ACK. The operation of these control frames is complex and each warrants its own discussion. Therefore, I suggest consulting the standard or 802.11 handbooks for more information.

Modulation selections

When developing any WLAN system, modulation choices should be a key concern. This is especially true when engineering IEEE 802.11b products.

Traditionally, IEEE 802.11 systems employed either a frequency-hopping spread spectrum (FHSS) or DSSS approach. Both approaches were good solutions for delivering data rates in the 1- to 2-Mbps range.

As the market has evolved, however, there has been a clear call for WLAN systems that deliver higher data rates. To do this, the IEEE 802.11 committee unleashed the 802.11b spec, which supports 11-Mbps data rates and employs a DSSS scheme.

Legacy IEEE 802.11 DSSS systems employ differential binary phase-shift keying (DBPSK) and differential quadrature phase-shift keying (DQPSK) modulation techniques for delivering data packets at data rates of 1 and 2 Mbps. In DSSS systems, data is spread using an 11-bit Barker word prior to transmission.

DSSS systems operating in the 2.4-GHz band feature channels that occupy 22 MHz of bandwidth. This channel bandwidth scheme was chosen to allow three independent, non-interfering WLAN networks across 83.5 MHz of spectrum. An example of the channel spacing and shaping of the waveform for DSSS operation is illustrated in Figure 1 .

Figure 1: Channel spacing and shaping for DSSS operation

A new approach Top

As the IEEE 802.11 committee began to look at higher speed operation, it was clear that supporting interoperability with legacy 1- and 2-Mbps DSSS systems was extremely important, and changes to the current DSSS channel scheme were not possible. The committee realized it had to move toward a new modulation technique that would allow higher speed operation without widening channel bandwidths.

During the evaluation process, the IEEE 802.11 committee considered a variety of modulation scheme proposals. These proposals included M-ary bi-orthogonal keying (MBOK), quadrature amplitude modulation (QAM), Barker code pulse position modulation (BCPM), packet binary convolutional coding (PBCC), M parallel orthogonal spreading signals, and carrier frequency offset spread spectrum.

After some heated battles, the committee adopted a joint proposal put forward by Lucent and Intersil Corporation for the 802.11. Under this proposal, designers could achieve higher-speed operation in DSSS WLAN systems using a complementary code keying (CCK) modulation scheme.

In addition to CCK, the committee adopted PBCC, a modulation proposed by Alantro Communications as an option in the standard for future applications requiring higher performance and robustness. Systems implementing PBCC must support CCK to be interoperable with 802.11b 11-Mbps WLAN systems. A control bit is provided in the preamble and PLCP header to indicate whether PBCC is supported.

More on CCK CCK was chosen over other modulations for its superior performance when combating multipath. CCK is a form of vector modulation and a variation of M-ary orthogonal keying (MOK) modulation which uses I&Q modulation with complex symbol structures. The spreading employed by CCK offers the same chipping rate and spectrum shape as the legacy Barker word spreading functions as shown in Figure 1 . The CCK modulation is not a new concept in the communication market. In fact, this approach has been around for decades and is based on a complex code set known as complementary codes. The origins of complementary codes date back to the 1950s, when M.J.E. Golay conceived them for infrared multislit spectrometry. Over the years, complementary codes found their way into radar and communication applications for their rich autocorrelation and cross-correlation properties.

The IEEE committee drew on past complementary code efforts to develop the CCK modulation scheme for the 802.11b spec. By doing this, the committee developed a CCK scheme that uses complementary codes derived from Walsh/Hadamard functions. Walsh/Hadamard functions are similar to those used in cellular phone CDMA applications. They have excellent complex code properties. Using complex codes, we can get a larger set of orthogonal codes to select from and still get the same number of bits transmitted per symbol. These codes, used together with CCK, suffer less from multipath distortion in the form of cross coupling of I&Q channel information.

The spreading function for CCK was chosen from a set of a "M" nearly orthogonal vector by the data word. CCK uses one vector from a set of 64 complex (QPSK) vectors for the symbol, and thereby modulates 6 bits (one of 64) on each 8-chip spreading code. Two additional bits are sent by QPSK, modulating the entire code symbol and thus modulating 8 bits onto each symbol. The net effect is a symbol that has 16 bits of complexity.

Overall, the symbol rate for CCK is 1.375 Msamples/sec with a chipping rate of 11 Mchips/sec for 8 bits per symbol. This translates into a data rate of 11 Mbps.

PHY preamble structure

The MAC layer was not the only important issue the 802.11 committee considered when developing the high-rate specification. The committee also spent time working on the PHY preamble structure.


          Figure 2: Long and short preamble structure described in 802.11b
                      specification.

The IEEE 802.11b protocol supports data rates of 1, 2, 5.5, and 11 Mbps. To achieve such data rates, two PHY preamble header structures are supported. These structures include a long and short preamble (see Figure 2 ). The long preamble uses the legacy 1- and 2-Mbps DSSS header to allow interoperability with other legacy systems. The preamble consists of a 128-bit sync field containing a string of scramble logic 1s followed by a 16-bit start-of-frame (SFD) field. Acquisition of the received signal for long preambles and headers are processed at 1 Mbps. Data packets contained in the PLCP service data unit (PSDU) can be received at 1-Mbps-DBPSK, 2-Mbps-DQPSK, and 5.5- and 11-Mbps CCK. The rate at which the PSDU is received is determined by the value stored in the signal field. To ensure interoperability with legacy systems, all compliant 802.11b systems must support the long preamble.

The short preamble option was provided in the standard to improve the efficiency of a network's throughput. The short preamble uses a 56-bit sync field, which is scrambled with logic 0s, followed by a 16-bit SFD field. The acquisition of the preamble signal is processed at 1 Mbps. The header is processed at 2 Mbps and the payload in the PSDU can be processed at 2, 5.5, or 11 Mbps. It takes a maximum of 192 ms to process the long preamble and 96 ms to process the short preamble. The short preamble yields a 50% savings in overhead in a PLCP data unit (PPDU) frame on a per packet basis. The short preamble can improve the throughput of a network when transmitting control frames and fragmented data packets carrying time-bounded content such as voice, VoIP, and streaming video.

TOP