Going Wi-Fi: A Practical Guide to Planning and Building an 802.11 Network

The most commonly deployed WLANs conform to the IEEE 802.11b specification. Not only are they increasingly deployed in private enterprise applications, but also in public applications such as airports and coffee shops. This standard includes three transmission options, one of which is infrared-based, and two of which are RF-based. (This book covers only the RF options.)

This PHY specification provides High Rate Direct Sequencing Spread Spectrum (DSSS) modulation in the 2.4 GHz band (the same ISM band that the original 802.11 uses), and offers the potential of three simultaneous channels. The IEEE 802.11b Task Group also enhanced the original 802.11's DSSS PHY to include 5.5 Mbps and 11 Mbps data rates, in addition to the one Mbps and two Mbps data rates of the original 802.11 standard. The Task Group did this by employing DSSS modulation using the Barker code chipping sequence. Each bit is encoded into an 11-bit Barker code (e.g., 10110111000), with each resulting data object forming a chip. The chip is put on a carrier frequency (i.e., a small frequency range that carries the signal) in the 2.4 GHz range, and the waveform is modulated using one of several techniques. 802.11b systems running at one Mbps use Barker code and binary phase shift keying (BPSK) modulation, and those running at two Mbps use Barker code and quadrature phase-shift keying (QPSK) modulation.

Then in 2001, the FCC indicated in a Notice of Proposed Rule Making from the FCC (ET 99-231; FNPRM & ORDER 05/11/01; FCC 01-158 Amendment of Part 15 of the Commission's Rules Regarding Spread Spectrum Devices, Wi-LAN, Inc. et al.), that it would consider relaxing the spread spectrum requirement on the ISM band. This Amendment abandons the peaceful "coexistence of equipment" requirement (interference rejection) in favor of support for greater wireless network capacity (higher bit-rate transmissions). The 802.11b Task Group responded to this action by modifying the specification. For high bit rates above 2 Mbps (i.e. 5.5 Mbps to 11 Mbps), 802.11b's purely spread spectrum techniques was supplanted by complementary code keying (CCK) modulation so as to provide 4 or 8 bits per transmission symbol, which when combined with QPSK, allows 802.11b to easily reach its maximum data rate of 11 Mbps.

Thus current 802.11b networks use CCK to provide the higher data rates. To support very noisy environments as well as extended range, 802.11b wireless LANs also are capable of "dynamic rate shifting," which allows data rates to be automatically adjusted to compensate for the varying nature of the radio channel. Initially, the equipment tries to connect at the full 11 Mbps rate. If the devices move beyond the optimal range for 11 Mbps operation, or if considerable interference is encountered, then the 802.11b devices will "fall back" and transmit at lower speeds; first to 5.5, then to 2, and finally to 1 Mbps. Likewise, if the device moves closer or if the interference disappears, then the connection will automatically increase to 11 Mbps. Rate shifting is a Physical Layer mechanism that's transparent to the user and the upper layers of the OSI protocol stack.

Furthermore, systems running at 5.5 Mbps and 11 Mbps use a combination of CCK and QPSK. CCK involves 64 unique code sequences, each of which supports 6 bits per code word. The CCK code word is then modulated onto the RF carrier using QPSK, which allows another two bits to be encoded for each 6-bit symbol. Therefore, each 6-bit symbol contains 8 bits (i.e. 1 byte).

Because of the complex encoding, it is much more difficult to discern which of the 64 code words is coming across the airwaves. Furthermore, the radio receiver design is significantly more difficult. In fact, while a 1 Mbps or 2 Mbps radio has one correlator (the device responsible for lining up the various signals bouncing around and turning them into a bitstream), 11 Mbps radio must have 64 such devices.

PHY and MAC

Like 802.11a, the 802.11b PHY Layer is split into two sublayers: PLCP (Physical Layer Convergence Protocol) and the PMD (Physical Medium Dependent). The PMD takes care of the encoding. The PLCP presents a common interface for higher-level drivers to write to, and provides carrier sense and CCA (Clear Channel Assessment), which is the signal that the MAC sublayer needs to determine whether the medium is currently in use. (See Fig. 6.10.)

Figure 6.10: The IEEE 802.11 PHY frame using DSSS.

The PLCP consists of a 144-bit preamble that is used for synchronization to determine radio gain and to establish CCA. The preamble comprises 128 bits of synchronization (scrambled 1 bits), followed by a 16-bit field consisting of the pattern 1111001110100000. This sequence, which is called the SFD or start frame delimiter, marks the start of every frame.

The next 48 bits are collectively known as the PLCP header. The header contains four fields: signal, service, length, and header error check (HEC). The signal field indicates how fast the payload will be transmitted (1, 2, 5.5 or 11 Mbps). The service field is reserved for future use. The length field indicates the length of the ensuing payload, and the HEC is a 16-bit CRC of the 48-bit header.

The PLCP is always transmitted at 1 Mbps, which serves not only to complicate things, but also degrades performance, because 24 bytes of each packet are sent at 1 Mbps. This means that the PLCP introduces 24 bytes of overhead into each wireless packet before the packet's destination is even considered. In comparison, Ethernet introduces only 8 bytes of data. Because the 192-bit header payload is transmitted at 1 Mbps, 802.11b is at best only 85 percent efficient at the Physical Layer.

The MAC sublayer's most basic ability is to sense a quiet time on the network before transmitting. Once the host has determined that the medium has been idle for a minimum time period, it may transmit a packet. This minimum time period is known as "distributed coordination function inter-frame spacing" or "DIFS." If the medium is busy, the node must wait for a time equal to DIFS, plus a random number of slot times. The time between the end of the DIFS period and the beginning of the next frame is known as the contention window.

Each station listens to the network, and the first station to finish its allocated number of slot times begins transmitting. If any other station hears the first station talk, it stops counting down its back-off timer. When the network is idle again, it resumes the countdown. In addition to the basic back-off algorithm, 802.11b adds a back-off timer that ensures fairness. Each node starts a random back-off timer when waiting for the contention window. This timer ticks down to zero while waiting in the contention window. Each node gets a new random timer when it wants to transmit. This timer isn't reset until the node has transmitted (see Fig. 6.11).

Figure 6.11: 802.11b contention window.

The Hidden-Node Problem

The hidden-node problem can occur where walls and other structures create obscure radio coverage areas. To handle this situation, an RTS/CTS (request to send/clear to send) is specified as an optional feature of the IEEE 802.11b standard. RTS/CTS solves the hidden-node problem in the following fashion:

Referring to Fig. 6.12, when node A wants to transmit data to node B, it first sends an RTS packet. The RTS packet includes the address of the receiver of the data transmission ensuing and the duration of the whole transmission, including the ACK related to it. Node B hears this request (as do nodes D and E). Node A must use the standard transmission method to obtain access to send the RTS packet. Once the receiving host receives the packet, that host replies with a CTS message that includes the same duration of the session about to happen. When node B replies with this CTS message, node C (and F and G) hears this response and is made aware of the potential collision, and will hold its data for the appropriate amount of time, preventing a collision. If every node on the network uses RTS/CTS, collisions are guaranteed to occur only while in the contention window. Access points also participate in the RTS/CTS process when necessary.

Figure 6.12: The hidden node problem. Workstations A, B and C can all see wireless access point P. Workstations A and B can see one another, and B and C can see one another, but A can't see C.

RTS/CTS, however, adds significant overhead to the 802.11b protocol, especially at small packet sizes. If used, RTS/CTS thresholds must be set on both the access point and the client side.

Power Level Influences

The FCC limits the power output of the 802.11b system to one watt EIRP equivalent (or effective) isotropic (or isotropically) radiated power. At this low power level the physical distance between the transmitting devices becomes an issue due to signal attenuation, with error performance suffering as the distance increases. (Note: 100 meters / 328 feet is a pretty good rule of thumb for an 802.11b WLAN with clear line-of-sight.)

Any dense physical obstructions between transmitter and receiver add considerably to the problem. Therefore the devices adapt to longer distances, physical obstructions and other factors that impact signal strength by using a less complex encoding technique, resulting in lower signaling speed, which translates into a lower data rate.

For example, a system running at 11 Mbps using CCK and QPSK might throttle back to 5.5 Mbps by halving the signaling rate as the distances increase, as doors and walls get in the way, and as bit errors increase.

The situation may get even worse when you move your laptop out to poolside on a sunny summer afternoon. The distance from an access point and interference encountered as the signal travels that distance affects a signal's quality. Thus the system might throttle back to two Mbps using only QPSK, or even one Mbps using BPSK, to keep the connection from degrading. This process is much the same as that used by conventional fallback modems that might initiate a call at 56 Kbps (actually 53.3 Kbps), and "fall back" to rates of perhaps 28.8 Kbps or 14.4 Kbps as the quality of the dial-up PSTN connection degrades.

If there are a lot of errors in transmission, throughput drops precipitously, as the receiving station must advise the transmitting station of the errored frames and then wait for retransmissions. If, for example, the error rate is 50%, the actual throughput drops to about 2 Mbps.

This scenario is a blend of a best-case 11 Mbps and a worst-case error rate. In actuality, such an error rate would cause the system to fall back to a lower transmission rate of perhaps 2 Mbps, at which rate the combination of the Barker code and QPSK would be used and the error rate would drop.

Power Savings and DTIMs

By default, 802.11b systems use a constant access mode (CAM) to listen constantly to the network and get the data they need. When power utilization is an issue, however, the workstations and access points can be configured for polled access mode (PAM). With this, the client devices on the network wake up at a regular interval and listen for a special packet called a traffic information map (TIM) from the access point. In between TIMs, the client radio shuts off and thus conserves power. All the devices on the network share the same wake-up period, as they must all wake up at exactly the same time to hear the TIM from the access point.

The TIM informs certain clients that data is waiting at the access point. A client card stays awake when the TIM indicates it has messages buffered at the access point until those messages are transferred, and then the card goes to sleep again. The access point buffers the data for each card until it receives a poll request from the destination station. Once the data is exchanged, the station goes back into power-saving mode until the next TIM is transmitted. Tests run by Network Computing magazine discovered that PAM mode could save power by as much as 1000 percent, depending on the volume of traffic on the network.

The access point indicates the presence of broadcast traffic with a delivery traffic information map (DTIM) packet. The DTIM timer is always a multiple of the TIM timer and is often adjustable at the access point. Setting that value high cuts down on the amount of time the station must stay awake checking for broadcast traffic. However, a higher DTIM timer means that the radio will stay on longer to receive DTIM traffic when it does come up in the time cycle.

In a typical enterprise environment, two or more access points will provide signals to a single client. The client is responsible for choosing the most appropriate access point based on the signal strength, network utilization and other factors. When a station determines the existing signal is poor, it begins scanning for another access point. This can be done by listening passively or actively probing each channel and waiting for a response.

Once information has been received, the station selects the most appropriate signal and sends an association request to the new access point. If the new access point sends an association response, the client has successfully transferred to a new access point. This is called "make, then break" behavior.

The Channels

The 802.11b specification divides the assigned RF spectrum into 14 channels. The FCC allows the use of 11 channels (1 through 11). Since the U.S. 2.4 GHz band is only 83 MHz wide, and the 802.11b channels are 25 MHz wide, only three channels can be used simultaneously. So not only is the amount of spectrum highly limited, but not all of it is used.

Furthermore, there also is overlap between adjacent channels (e.g., channels two and three), which affects performance. This overlap, therefore, requires that any given system maintain maximum channel separation from other systems in proximity.

When designing international wireless LANs, you must choose channels with the least common denominator because different local regulatory bodies allow the use of a varying number of channels. For example, Japan allows the use of only one channel, but the U.K. allows the use of channels 1 through 13. However, none allow the use of all 14.

The transmitter's modulator translates the spread signal into an analog form with a center frequency corresponding to the radio channel chosen by the user. The following table identifies the center frequency of each channel:

The Future

It is expected that over the next few years, 802.11b networks will be slowly phased out. There is still much debate over whether this popular WLAN standard's successor will be 802.11a or 802.11g.