Wireless LANs have been around for a few years, but a lack of standards required that only components from the same manufacturer could be used. Today, we have two standards, Wi-Fi and Home RF; a third standard, Bluetooth, is expected to hit the market in early 2002. In order to understand this technology better, let's discuss the fundamentals behind it.
In 1997, a standard, known as IEEE 802.11, for wireless LANs (WLANs) that defined both the physical and the IEEE-approved Media Access Control (MAC) layer (part of the Data-Link Layer) was approved. This was recently updated to 802.11b.
802.11b Basic Theory
The Data-Link layer, comprised of the MAC and Logical Link Control (LLC) sub-layers, is responsible for packaging data for transmission and error checking. The LLC ensures the integrity of the data sent, while the MAC layer monitors and controls how the data accesses the transmission medium.
In a traditional wired-Ethernet environment, data packets from each attached device are randomly transmitted and, in the event two devices transmit simultaneously, the devices will wait and retransmit until successful. In this scheme, the devices detect the collision and provide a means to get the data across reliably. This protocol is called Carrier Sense Multiple Access with Collision Detection or CSMA/CD. In a wireless environment, the process would get a little more difficult, since we're dealing with a radio system, and it would be impractical for all radios in a network to be transmitting at the same time. The creators of the standard felt the price of the radios would be too high if all the radios worked on different frequencies in full-duplex mode. In addition, there could be some physical limitations to the amount of nodes a wireless network can handle in a point-to-multipoint network.
The 802.11b standard pertains to radios operating in half-duplex mode (the radio is either transmitting or receiving.) The transmission protocol used is called CSMA/CA; the CA in this case stands for Collision Avoidance. 802.11b uses this protocol to create a wireless network that can transport data between other wireless and wired devices on a network. Here is how it works: a wireless station (device) intending to transmit sends a Request to Send (RTS) data packet over the network during a quiet time on the network. The RTS consists of a data packet containing source and destination addresses and length of transmission. The destination device responds with a Clear to Send (CTS), informing the originating station that it exists and is ready to receive data. Other stations listening on the network set a Network Allocation Vector (NAV) that tells them to remain quiet until the transmission is complete. This is also called Virtual-Carrier Sensing.
The radio systems
Another point to consider with the implementation of WLANs is the need to make sure the data is transmitted and received reliably under such conditions as weak signals, noise and interference. Perhaps of even greater concern is the ability of a potential hacker to read network traffic without being physically attached to a cable. The simple and economical solution is to use the spread spectrum modulation scheme developed by actress Hedy Lamarr. This solution has provided secure communication for government and military applications since World War II. There are two specific types of spread-spectrum methods defined by the IEEE: Direct Sequence (DSSS) and Frequency Hopping (FHSS). DSSS spreads a signal over a range of frequencies using a signal rate of 1 or 2Mb/s. FHSS breaks the signal into smaller (shorter) packets and sends each burst sequentially over different frequencies. FHSS tends to be more spectrum efficient, although DSSS systems are superior with weak signals and have more noise immunity than FHSS. It should be mentioned that the 802.11b standard also supports optical media, such as the infrared port. These systems typically operate on the unlicensed 2.4 to 2.48GHz Industrial Medical Scientific (ISM) band, although future systems will operate in the 5GHz band.
Simple WLANs consist of an Access Point (AP) and a wireless NIC (client adapter.) The access point has three primary functions: 1) Communicates with other wireless devices. 2) Manages the wireless communications process 3) Acts as a router, routing data to the wireless network from the wired network and vice versa.
Since access points are, in most cases, limited to only a few hundred feet of useful range, it is customary to install several APs in large areas designed to serve wireless clients. Most access points contain an internal antenna; however, many provide for an external antenna.
The Virtual carrier Sensing mentioned above is part of a mechanism called the Basic Service Set (BSS). Among other functions, the BSS allows wireless clients to roam between different access points without interruption to service. The feature can also be configured to limit access for certain clients to specific APs, i.e. the number of physical locations or the number of users on a given AP. This is implemented by permitting a unique ID to be programmed into each wireless NIC, similar to the ESN number used in your mobile phone. Some manufacturers provide access points that can be used with high-gain external antennas that claim useable distances of several miles.
Most currently available WLANs operate with a signal rate of approximately 11Mb/s. This shouldn't be confused with actual data throughput. Similar to that of wired Ethernet, the actual data throughput of these devices is an area of debate in many circles. Manufacturers claim data rates in excess of 6Mb/s; my experience is that the speed of wireless systems tend to operate slightly slower than a properly configured 10baseT network (1.5Mb/s). The reason that it operates slower is due to the additional payload added to the transmitted frames containing the CSMA/CA protocol data, and the delay of the signal due to free-space RF characteristics and multipath. Keep in mind that performance will decrease as the number of users and amount of traffic increases.
One of the problems with comparing the speed of an Ethernet network to that of a dedicated bandwidth transmission method, such as a T1, is that the Ethernet network can only be measured by an average over time, since the data passing through it is sent in bursts. Depending on the type of data passed, that average could vary greatly. Some manufacturers also use aggressive data compression techniques, which yield up to a 3-to-1 data transfer boost. The problem is that most data is already stored in a compressed format, thus nullifying any advantage of further compression.
Next month we'll look closer at the standards and some applications.
Kevin McNamara, BE Radio's consultant on computer technology, is president of Applied Wireless Inc., New Market, MD.
All of the Networks articles have been approved by the SBE Certification Committee as suitable study material that may assist your preparation for the SBE Certified Broadcast Networking Technologist exam. Contact the SBE at (317) 253-1640 or go to www.sbe.org for more information on SBE Certification.