IEEE 802.11b

IEEE wireless local area network (WLAN)specification.

802.11b is capable of 11 Mbps, which happens to be faster than my school's wired LAN. Products conforming to 802.11b have a maximum range of about 1650 feet (503 meters) outdoors, but at this range, the bandwidth usually drops to about 1 Mbps (still great if you're just using the Internet). Indoors, at the full 11 Mbps, range is approximately 150 feet (46 meters).

This standard runs on the 2.4GHz frequency range. 802.11b uses direct sequence spread spectrum (DSSS). At high data rates, DSSS uses a set of 64, 8-bit codewords (Complementary Code Keying (CCK). These codewords are recognizable even in the presence of significant interference. To compensate for range and noise variation, it can dynamically shift between 1, 2, 5.5, and 11 Mbps.

Actual operation at the Data Link Layer of the OSI Reference Model is very similar to IEEE 802.3, with the addition of some additional functionality for wireless.

Wireless networks can operate in a variety of modes. The first is Ad-hoc, which creates a wireless network from client stations. These stations do not tie into a larger network and no base station is needed. In Infrastructure mode, the wireless clients communicate with a base station or access point, that may be connected to a wired LAN or the Internet. Basic mode uses only one base sation; extended allows for roaming between multiple base stations.

802.11b also encludes Layer 1 and Layer 2 security functionality. Wired Equivalant Privacy (WEP) uses a extended session identity that wireless clients must know in order to connect to it (if enabled). Keys are RC4; in either 64 or 128-bit flavors. Keys for the client and base station must match in order for a client to associate and gain access to a base station. The WEP security is relatively weak, and additional security should be used for any businesses installing a wireless LAN.

A quick note on the WEP encryption:

There is some confusion out there about the levels of encryption available in 802.11b. Some vendors list 40-bit encryption while others list 64-bit enrcyption.

The 40-bit encryption and the 64-bit encryption are the same thing. In effect, this scheme uses 40-bit secret key, with a 24-bit initialization vector. So some vendors list it as 40-bit and others as 64-bit.

128-encryption also uses the same scheme: 104-bit encryption, 24-bit initialization vector. The funny thing is, all vendors state they support 128-bit encryption. They don't get caught up in the 104-bit issue.

Operational Outline

Although 802.11 can be configured as an Ad Hoc peer-to-peer network consisting of only a few nodes, the primary business application will utilize WLANs as an extension of the traditional wired LAN, using a concentrator called an Access Point (AP).

A traditional wired distribution system brings network access to each WLAN in a network, terminating in Access Points (APs). The APs bridge the wired network with the WLANs and act as the concentrator for WLAN traffic. End devices such as laptops or workstations are outfitted with standard PCMCIA or PCI expansion cards to provide access to the WLAN.

This configuration, if properly designed, will allow users to roam in-between APs, and allow IT managers to load balance between overlapping AP coverage areas on-the-fly based on traffic or noise, and grow their network easily and at minimal cost. Best of all, all of these advantages are gained transparently to the user.

Another configuration option that can be employed by IT managers is the Point Co-ordination Function (PCF). Normally the WLAN will function like an Ethernet in that intelligence is distributed among each device, this is called Distributed Co-ordination Function (DCF). However, for time-constrained applications that need a deterministic delivery system, PCF can be employed and the WLAN will function more like a Token Ring, where intelligence will come from one device who polls each station to determine if they are allowed to transmit / receive data. This guarantees a maximum latency.

Most WLAN devices are manageable through a Telnet, SNMP or web browser interface, or some combination of the three. In addition to traffic statistics, the access point also includes management features such as mapping of access points and their associated clients, and controlling access and traffic flow.

802.11b Brief Technical Summary

The 802.11b standard encompasses the physical and Data Link layers of the OSI model. It uses the 802.2 LLC and 802.3 MAC addressing system, using the same 48-bit addressing scheme. This makes interoperability with a pre-existing Ethernet network extremely convenient.

As far as the actual media access mechanism, 802.11b again uses a very similar method to CSMA/CD, in fact it's just a modified version of this protocol to account for the different medium. Carrier Sense Multiple Access / Collision Avoidance (CSMA/CA) follows this algorithm: Listen to the medium for a free moment to transmit, when it's clear wait for a random time interval, and then if it's still clear transmit. Since the stations cannot listen for a collision while transmitting, the receiving station sends an explicit ACK for each frame received. This adds to the overall overhead but makes the network necessarily robust for wireless use.

The physical layer uses a technique called Direct Sequence Spread Spectrum and operates in the unlicensed 2.4GHz radio band. Quadrature Phase-Shift Keying (QPSK) is used to squeeze 8bits into every symbol (with a base signaling rate of 1.375 MSps (Million Symbols per Second) that makes 8 * 1.375 = 11Mbps).

Stupid 802.11b Tricks

The only thing limiting the range of 802.11b (in the U.S.) is Part 15.247 of FCC Rules and Regulations. This regulates the 900mhz, 2.4 GHz (802.11b), and 5.7 Ghz (802.11a) unlicensed bands. Concerning 2.4 GHz, Part 15.247 limits the power output of the Intentional Radiator (your 802.11b wireless card) to one watt of power.

There is, of course, work arounds.

The FCC will let you have about four times the base power (6dBi) as long as you are using a directional antenna. This is nice, especially for mobile applications that aren't within five hundred meters of your base station. This is allowed because with a directional antenna, you aren't just dumping your power in all directions, but aiming it at a certain location. Cleaner, and less noisy.

If you wanted to set up two fixed wireless locations communicating with each other, you can hack it up a little more, and get around ten miles at full 11mb/s bandwidth. The catch is, the two directional antennas need to stay pointing at each other, setting up a permanent wireless LAN. Otherwise, you would have to stick with the above, "directional antenna", restrictions.

Once you go past a certain wattage of output from your directional antenna (30dBm base, 6dBi gain, 36dBm altogether), you have to start pulling back on power output from your intentional radiator. Strangely enough, this isn't a problem for most 802.11b users, because their cards don't put out more than 15dBm, or .03 Watts of power (3% of the maximum).

There is good smarts behind the restrictions, though. If you were to use a discarded Primestar Dish (which I use) at the full one watt output of your intentional radiator, you would, with the parabolic dish's gain (around 27-31 dBi well configured), achieve 1000 watts in a concentrated area. My microwave puts out 1200 watts. Knowledge (or information) truly is dangerous.