Infrared (IR) technology is used to implement wireless local area networks (LANs) as well as the wireless interface to connect laptops and other portable machines to the desktop computer equipped with an IR transceiver. IR LANs are proprietary in nature, so users must rely on a single vendor for all the equipment. However, the IR interface for connecting portable devices with the desktop computer is standardized by the Infrared Data Association (IrDA).

IR LANs typically use the wavelength band between 780 and 950 nanometers (nm). This is due primarily to the ready availability of inexpensive, reliable system components. There are two categories of IR systems that are commonly used for wireless LANs. One is directed IR, which uses a very narrow laser beam to transmit data over one to three miles. This approach may be used for connecting LANs in different buildings.

Although transmissions over laser beam are virtually immune to electromechanical interference and would be extremely difficult to intercept, such systems are not widely used because their performance can be impaired by atmospheric conditions, which can vary daily. Such effects as absorption, scattering, and shimmer can reduce the amount of light energy that is picked up by the receiver, causing the data to be lost or corrupted.

The other category is nondirected IR, which uses a less focused approach. Instead of a narrow beam to convey the signal, the light energy is spread out and bounced off narrowly defined target areas or larger surfaces such as office walls and ceilings. Nondirected IR links may be further categorized as either line of sight or diffuse. Line-of-sight links require a clear path between transmitter and receiver but generally offer higher performance.

The line-of-sight limitation may be overcome by incorporating a recovery mechanism in the IR LAN that is managed and implemented by a separate device called a “multiple access unit” (MAU) to which the workstations are connected. When a line-of-sight signal between two stations is temporarily blocked, the MAU’s internal optical link control circuitry automatically changes the link’s path to get around the obstruction.

When the original path is cleared, the MAU restores the link over that path. No data are lost during this recovery process. Diffuse links rely on light bounced off reflective surfaces. Because it is difficult to block all the light reflected from large surface areas, diffuse links are generally more robust than line-of-sight links. The disadvantage of diffused IR is that a great deal of energy is lost, and consequently, the data rates and operating distances are much lower.

System Components

Light-emitting diodes (LEDs) or laser diodes (LDs) are used for transmitters. LEDs are less efficient than LDs, typically exhibiting only 10 to 20 percent electrooptical power conversion efficiency, while LDs offer an electrooptical conversion efficiency of 30 to 70 percent. However, LEDs are much less expensive than LDs, which is why most commercial systems use them.

Two types of low-capacitance silicon photodiodes are used for receivers: positive-intrinsic-negative (PIN) and avalanche. The simpler and less expensive PIN photodiode is typically used in receivers that operate in environments with bright illumination, whereas the more complex and more expensive avalanche photodiode is used in receivers that must operate in environments where background illumination is weak. The difference in the two types of photodiodes is their sensitivity.

The PIN photodiode produces an electric current in proportion to the amount of light energy projected onto it. Although the avalanche photodiode requires more complex receiver circuitry, it operates in much the same way as the PIN diode, except that when light is projected onto it, there is a slight amplification of the light energy. This makes it more appropriate for weakly illuminated environments. The avalanche photodiode also offers a faster response time than the PIN photodiode.

Operating Performance

Current applications of IR technology yield performance that matches or exceeds the data rate of wire-based LANs: 10 Mbps for Ethernet and 16 Mbps for Token Ring. However, IR technology has a much higher performance potential— transmission systems operating at 50 and 100 Mbps have already been demonstrated. Because of its limited range and inability to penetrate walls, nondirected IR can be easily secured against eavesdropping.

Even signals that go out windows are useless to eavesdroppers because they do not travel far and may be distorted by impurities in the glass as well as by the glass’s placement angle. IR offers high immunity from electromagnetic interference, which makes it suitable for operation in harsh environments like factory floors. Because of its limited range and inability to penetrate walls, several IR LANs may operate in different areas of the same building without interfering with each other.

Since there is less chance of multipath fading (large fluctuations in received signal amplitude and phase), IR links are highly robust. Many indoor environments have incandescent or fluorescent lighting that induces noise in IR receivers. This is overcome by using directional IR transceivers with special filters to reject background light.

Media Access Control

IR supports both contention-based and deterministic media access control techniques, making it suitable for Ethernet as well as Token Ring LANs. To implement Ethernet’s contention protocol, carriersense multiple access (CSMA), each computer’s IR transceiver is typically aimed at the ceiling. Light bounces off the reflector in all directions to let each user receive data from other users. CSMAensures that only one station can transmit data at a time. Only the stations to which packets are addressed can actually receive them.

Deterministic media access control relies on token passing to ensure that all stations in turn get an equal chance to transmit data. This technique is used in Token Ring LANs, where each station uses a pair of highly directive (line-ofsight) IR transceivers. The outgoing transducer is pointed at the incoming transducer of a station down line, thus forming a closed ring with the wireless IR links among the computers.

With this configuration, much higher data rates can be achieved because of the gain associated with the directive IR signals. This approach improves overall throughput, since fewer bit errors will occur, which minimizes the need for retransmissions.