Wireless Infrared Communications :A Survey

Wireless Infrared Communications :A Survey

Wireless Infrared Communications :A Survey

Wireless infrared communications refers to the use of free-space propagation of light waves in the near infrared band as a transmission medium for communication as shown in Figure 1.

Fig1:Basic Wireless communication system. The communication can be between one portable communication device and another or between a portable device and a main device, called an access point or base station. Typical portable devices include laptop computers, personal digital assistants, and portable telephones, while the base stations are usually connected to a computer with other networked connections. Although infrared light is usually used, other regions of the optical spectrum can be used so the term wireless optical communications" instead of wireless infrared communications" is sometimes used. Wireless infrared communication systems can be characterized by the application for which they are designed or by the link type, as described below.

The primary commercial applications are as follows:

Short-term cable- less connectivity for information exchanges e.g. business cards, schedules, file sharing between two users. The primary example is IrDA systems (see Section 4). Wireless local area networks (WLANs) provide net-work connectivity inside buildings. This can either be an extension of existing LANs to facilitate mobility or to establish ad hoc" networks where there is no LAN. The primary example is the IEEE 802.11standard building-to-building connections for high- speed network access or metropolitan or campus area networks. wireless input and control devices, such as wireless mice, remote controls, wireless game controllers, and remote electronic keys.

Another important way to characterize a wireless infrared communication system is by the link type", which means the typical or required arrangement of receiver and transmitter.

Fig 2: Common Applications

Figure 2 depicts the two most common conjurations: the point-to-point system and the diffuse system. The simplest link type is the point-to-point system. There, the transmitter and receiver must be pointed at each other to establish a link. The line-of-

sight (LOS) path from the transmitter to the receiver must be clear of obstructions, and most of the transmitted light is directed toward the receiver. Hence, point-to-point systems are also called directed LOS systems. The links can be temporarily created for a data exchange session between two users, or established more permanently by aiming a mobile unit at a base station unit in the LAN replacement application. In diffuse systems, the link is always maintained between any transmitter and any receiver in the same vicinity by reflecting or bouncing" the transmitted information bearing light off reflecting surfaces such as ceilings, walls, and furniture. Here, the transmitter and receiver are non-directed; the transmitter employs a wide transmit beam and the receiver has a wide field-of- view. Also, the LOS path is not required. Hence, diffuse systems are also called non- directed non-LOS systems. These systems are well suited to the wireless LAN application, freeing the user from knowing and aligning with the locations of the other communicating devices.

Most wireless infrared communications systems can be modeled as having an output signal Y (t) and an input signal X(t) which are related by Y (t) = X(t) * c(t) + N(t) Where * denotes convolution, c(t) is the impulse response of the channel and N(t) is additive noise. This article is organized around answering key questions concerning the system as represented by this model. In Section 2, we consider questions of optical design. What range of wireless infrared communications systems does this model apply to? How does c(t) depend on the electrical and optical properties of the receiver and transmitter? How does c(t) depend on the location, size, and orientation of the receiver and transmitter? How do X(t) and Y (t) relate to optical processes? What wavelength is used for X (t)? What devices

produce X (t) and Y (t)? What is the source of N(t)? Are there any safety considerations? In Section 3, we consider questions of communications design. How should a data symbol sequence be modulated onto the input signal X(t)? What detection mechanism is best for extracting the information about the data from the received signal Y (t)? How can one measure and improve the performance of the system? In Section 4, we consider the design choices made by existing standards such as IrDA and 802.11.Finally, in Section 5, we consider how these systems can be improved in the future.

1. Modulation and demodulation

   What characteristic of the transmitted wave will be modulated to carry information fro m the transmitter to the receiver? Most communication systems are based on phase, amplitude, or frequency modulation, or some combination of these techniques However, it is difficult to detect such a signal following non directed propagation, and more expensive narrow line width sources are required(2). An effective solution is to use intensity modulation, where the transmitted signal's intensity or power is proportional to the modulating signal. At the demodulator (usually referred to as a detector in optical systems) the modulation can be extracted by mixing the received signal with a carrier light wave. This coherent detection technique is best when the signal phase can be maintained. However, this can be difficult to implement and additionally, in non directed propagation, it is difficult to achieve the required mixing efficiency. Instead, one can use direct detection using a photo detector. The photo detector current is proportional to the received optical signal intensity, which for intensity modulation, is also the original modulating signal. Hence, most systems use intensity modulation with direct detection

   (IM/DD) to achieve optical modulation and demodulation. In a free-space optical communication system, the detector is illuminated by sources of light energy other than the source. These can include ambient lighting sources, such as natural sunlight, fluorescent lamp light, and incandescent lamp light. These sources cause variation in the received photo current that is unrelated to the transmitted signal, resulting in an additive noise component at the receiver. We can write the photocurrent at the receiver as Y (t) = X(t) * Rh(t) + N(t)

   where R is the responsivity of the receiving photodiode (A/W). Note that the electrical impulse response c(t) is simply R times the optical impulse response h(t). Depending on the situation, some authors use c(t) and some use h(t) as the impulse response.

2. A transmitter or source converts an electrical signal to an optical signal. The two most appropriate types of device are the light- emitting diode (LED) and semiconductor laser diode (LD). LEDs have a naturally wide transmission pattern, and so are suited to non directed links. Eye safety is much simpler to achieve for an LED than for a laser diode, which usually have very narrow ransmit beams. The principal advantages of laser diodes are their high energy- conversion efficiency, their high modulation bandwidth, and their relatively narrow spectral width. Although laser diodes o ffer several advantages over LEDs that could be exploited, most short-range commercial systems currently use LEDs. A receiver or detector converts optical power into electrical current by detecting the photon flux incident on the detector surface. Silicon p-i-n photodiodes are ideal for wireless infrared communications as they have good quantum efficiency in this band and are in expensive (4). Avalanche photodiodes are not used here since the dominant noise

   source is back-ground light- induced shot noise rather than thermal circuit noise.

3. combination with base line restoration(6), or by careful selection of the modulation type.

4. There are two safety concerns when dealing with infrared communication systems. Eye safety is a concern because of a combination of two effects: the cornea is transparent from the near violet to the near IR. Hence, the retina is sensitive to damage from light sources transmitting in these bands. However, the near IR is outside the visible range of light, and so the eye does not protect itself from damage by closing the iris or closing the eyelid. Eye safety can be ensured by restricting the transmit beam strength according to IEC or ANSI standards (7, 8).Skin safety is also a possible concern. Possible short term effects such as heating of the skin are accounted for by eye safety regulations (since the eye requires lower power levels than the skin). Long term exposure to IR light is not a concern, as the ambient light sources are constantly submitting our bodies too much higher radiation levels than these communication systems do. Communications Design Equally important for achieving the design goals of wireless infrared systems are communications issues. In particular, the modulation signal format together with appropriate error control coding is critical to achieving power efficiency. Channel characterization is also important for understanding performance limits. A Modulation Techniques to understand modulation in IM/DD systems, we must look again at the channel model Y (t) = X(t)

   * c(t) + N(t) and consider its particular characteristics. First, since we are using intensity modulation, the channel input X(t) is optical intensity and we have the constraint X(t) ¸ 0. The average transmitted optical power PT is the time average of X(t). Our goal is to minimize the transmitted power required to attain a certain probability of bit error Pe, also known as a bit error rate

   (BER).It is useful to define the signal-to- noise ratio SNR as SNR =

   where H(0) is the d.c. gain of the channel,

   i.e. it is the Fourier transform of h(t) evaluated at zero frequency, so H(0) = The transmitted signal can be represented as X(t) = .

   The sequence fang represents the

   digital information being transmitted, where an is one of L possible data symbols from 0 to L¡1. The function si(t) represents one of L pulse shapes with duration Ts, the symbol time. The data rate (or bit rate) Rb, bit time T, symbol rate Rs, and symbol time Ts are related as follows: Rb = 1/T, Rs = 1/Ts, and Ts = log2 (L) T.

   There are three commonly used types of modulation schemes: on-off keying (OOK) with non-return to- zero pulses, OOK with return-to-zero pulses of normalized width (RZ ) and pulse position modulation with L pulses (L-PPM). OOK and are simpler to implement at both the transmitter and receiver than L-PPM. The pulse shapes for these modulation techniques are shown in Figure 3. Representative examples of the resulting transmitted signal X(t) for a short data sequence are shown in Figure4. We compare modulation schemes in Table 1 by looking at measures of power efficiency and bandwidth efficiency. Bandwidth efficient is measured by dividing the zero-crossing bandwidth by the data rate. Bandwidth efficient schemes have several advantages the receiver and transmitter electronics are cheaper, and the modulation scheme is less likely to be affected by multipath distortion. Power efficient is measured by comparing the required transmit power to achieve a target probability of error Pe for different modulation techniques. Both RZ-± and PPM are more power efficient than OOK, but at the cost of reduced bandwidth efficient.

   However, for a given bandwidth efficient PPM is more power efficient than RZ-±, and so PPM is most commonly used. OOK is most useful at very high data rates, say 100 Mb/s or greater. Then, the effect of multipath distortion is the most significant effect and bandwidth efficiency becomes of paramount importance (9).

Error control coding is an important technique for improving the quality of any digital communication system. We concentrate here on forward error correction channel coding, as this specifically relates to wireless infrared communications; source coding and ARQ coding are not considered here. Trellis-coded PPM has been found to be an effective scheme for multipath infrared channels (10, 11).The key technique is to recognize that although on a distortion- free channel, all symbols are orthogonal and equidistant in signal space; this is not true on a distorting channel. Hence, trellis-coding using set partitioning designed to separate the pulse positions of neighboring symbols is an effective coding method. Coding gains of 5.0 dB electrical have been reported for rate 2/3-coded 8-PPM over un coded 16- PPM, which has the same bandwidth (11).

C. Channel impulse response characterization Impulse response characterization refers to the problem of understanding how the impulse response c(t) in Equation (1) depends on the location, size, and orientation of the receiver and transmitter. There are basically three classes of techniques for accomplishing this: measurement, simulation, and modeling. Channel measurements have been de-scribed in several studies (9, 12, 2), and these form the fundamental basis for understanding the channel properties. A particular study might generate a collection of hundreds or thousands of example impulse responses ci(t) for configuration i. The collection of measured impulse responses ci(t) can then

be studied by looking at scatter plots of path loss 0versus distance, scatter plots of delay spread versus distance, the effect of transmitter and receiver orientations, robustness to shadowing, and so on. Simulation methods have been used to allow direct calculation of a particular impulse response based on a site specific characterization of the propagation environment (13, 14). The transmitter, receiver, and there reflecting surfaces are described and used to generate an impulse response. The basic assumption is that most interior surfaces reflect light diffusely in a Lambertian pattern, i.e. all incident light, regardless of incident angle, is rejected in all directions with intensity proportional to the cosine of the angle of the rejection with the surface normal. The difficulty with existing methods is that accurate modeling requires extensive computation. A third technique attempts to extract knowledge gained from experimental and simulation-based channel estimations into a simple to use model. In (15), for example, a model using two parameters (one for path loss, one for delay spread) is used to provide a general characterization of all diffuse IR channels.

Methods for relating the parameters of the model to particular room characteristics are given, so that system designers can quickly estimate the channel characteristics in a wide range of situations.

We examine the details of the two dominant wireless infrared technologies, IrDA and IEEE 802.11, and other commercial applications. Infrared Data Association Standards (IrDA) The Infrared Data Association (16), an association of about one hundred member companies, has standardized low-cost optical data links. The IrDA link transceivers or ports appear on many portable devices including notebook compters, personal digital assistants, and also computer peripherals such as printers.

The current version of the physical layer standards is IrDA PHY 1.3. Data rates from

2.4 kb/s to 4 Mb/s are supported. The link speed is negotiated by starting at 9.6 kb/s. Most of the transmission standards are for short-range, directed links which an operating range from0 m to 1 m. The transmitter half angle must be between 15 and 30 degrees, and the receiver field-of- view half angle must be at least 15 degrees. The transmitter must have a peak-power wavelength between 850 nm and 900 nm. IEEE 802.11 and wireless LANs The IEEE has published a set of standards for wireless LANs, IEEE 802.11 (17). The IEEE 802.11 standard is designed into the structure of the suite of 802 LAN standards. Hence, it determines the physical layer (PHY) and medium-access control layer (MAC) leaving the logical- link control (LLC) to 802.2. The MAC layer uses a form of carrier-sense multiple access with collision avoidance (CSMA/CA). The original standard supports both radio and optical physical layers with a maximum data rate of 2Mb/s. The 802.11b standard adds a

2.4 GHz radio physical layer at up to 11 Mb/s and 802.11a standard adds a 5.4 GHz radio physical layer at up to 54 Mb/s. The two supported data rates for infrared 802.11 LANs are 1 Mb/s and 2 Mb/s. Both systems use PPM but share a common chip rate of 4 M chips/s, as explained below. Each frame begins with a preamble encoded using 4 Mb/s OOK. In the preamble, a three-bit field indicates the transmission type, either1 Mb/s or 2 Mb/s (the six other types are reserved for future use). The data is then transmitted at 1Mb/s using 16-PPM or 2 Mb/s using 4- PPM. 16-PPM carries log2(16)/16=1/4 bits/chip, and 4-PPMcarries log2(4)=4 2 bits/chip, resulting in the same chip time for both types. The transmitter must have a peak-power wavelength between 850 nm and 950 nm. The required transmitter and receiver characteristics are intended to allow

for reliable operation at link lengths up to 10m.

Long range (greater than 10 m) infrared links must be directed LOS systems in order to ensure a reasonable path loss. The emerging products for long range links are typically designed to be placed on roof tops (18, 19), as this provides the best chance for establishing line-of-sight paths from one location to another in an urban environment. These high data rate connections can then be used for enterprise network access or metropolitan or campus-area networks. There are several design issues specific to these systems that are unique to these long- range systems(3).The first is atmospheric path loss, which is a combination of clean- air absorption from the air and absorption and scattering from particles in the air, such as rain, fog, and pollutants. Secondly, an effect called scintillation, which is caused by temperature variations along the LOS path, causes rapid fluctuations in the channel quality. Finally, building sway can affect alignment and result in signal loss unless the transceivers are mechanically isolated or active alignment compensation is used.

Wireless infrared communication has found several markets in and around the home, car, and office) which fall outside the traditional telecommunications markets of voice and data networking. These can either be classified as wireless input devices, or as wireless control devices, depending on one's perspective. Examples include wireless computer mice and keyboards, remote controls for entertainment equipment, wireless video-game controllers, and wireless door keys for home or vehicle access. All such devices use infrared communication systems due to the attractive combination of low cost, reliability, and lightweight in a transmitter/receiver pair that

achieves the required range, data rate, and data integrity required.

In this section, we discuss how competition from radio and developments in research will impact the future uses of wireless infrared communication systems.

A. Comparison to radio Wireless infrared communication systems enjoy significant

advantages over radio systems in certain environments. First, there is an abundance of unregulated optical spectrum available. This advantage is shrinking somewhat as the spectrum available for licensed and unlicensed radio systems increases due to modernization of spectrum allocation policies. Radio systems must make great efforts to over-come or avoid the effects of multipath fading, typically through the use of diversity. Infrared systems do not suffer from time-varying fades due to the inherent diversity in the receiver. This simplifies design and increases operational reliability. Infrared systems provide a natural resistance to eaves dropping, as the signals are confined within the walls of the room. This also reduces the potential for neighboring wireless communication systems to interfere with each other, which is a significant issue for radio-based communication systems. In band interference is a significant problem for both types of systems. A variety of electronic and electrical equipment radiates in transmission bands of current radio systems; microwave ovens are a good example. For infrared systems, ambient light, either man- made or natural, is a dominant source of noise. The primary limiting factor of infrared systems is their limited range, particularly when no good optical path can be made available. For example, wireless communication between conventional rooms with opaque walls and doors cannot be accomplished; one must resort to using either a radio-based or a wire line network to bypass the obstruction.

A variety of techniques have been considered to improve upon the performance of wireless infrared communication systems. At the transmitter, the radiation pattern can be optimized to improve performance characteristics such as range. Some optical techniques for achieving this are diffusing screens, multiple-beam transmitters, and computer- generated holographic images.

At the receiver, performance is ultimately determined by signal collection (limited by the size of the photo detector) and by ambient noise filtering. Optical interference filters can be used to reduce the impact of background noise; the primary difficulty is in achieving a wide field-of-view. This can be done using non-planar filters or multiple narrow FOV receiving elements. Some recent developments and research programs are described in (20), and an on- line resource guide is maintained in (21).

Wireless infrared communication systems provide a useful complement to radio-based systems, particularly for systems requiring low cost, light weight, moderate data rates, and only requiring short ranges. When LOS paths can be assured, range can be dramatically improved to provide longer links. Short-range wireless networks are poised for tremendous market growth in the next decade, and wireless infrared communications systems will compete in a number of arenas. Infrared systems have al- ready proven their effectiveness for short- range temporary communications and in high data rate longer range point-to-point systems. It remains an open question whether infrared will successfully compete\in the market for general-purpose indoor wireless access.

1. F. R. G feller and U. H. Bapst, Wireless in- house data communication via diffuse

   infrared radiation," Proceedings of the IEEE, vol. 67, pp. 1474{1486, Nov. 1979.

2. J. M. Kahn and J. R. Barry, \Wireless infrared communications," Proceedings of the IEEE, vol. 85, pp. 265{98, Feb. 1997.

3. D. Heatley, D. Wisely, I. Neild, and P. Cochrane, Optical wireless: The story so far," IEEE Communications Magazine, pp. 72{82, Dec. 1998.

4. J. R. Barry, Wireless Infrared Communications. Boston: K luwer Academic Publishers, 1994.

5. J. R. Barry and J. M. Kahn, \Link design for non-directed wireless infrared communications," Applied Optics, vol. 34, pp. 3764{3776, July 1995.

6. R. Narasimhan, M. D. Audeh, and J. M. Kahn, ffect of electronic-ballast fluorescent lighting on wireless infrared links," IEE Proceedings Optoelectronics, vol. 143, pp. 347{354, Dec. 1996.

7. International Electro technical Commission, CEI/IEC 825-1: Safety of Laser Products, 1993.

8. ANSI-Z136-1, American National Standard for the Safe Use of Lasers, 1993.

9. J. M. Kahn, W. J. Krause, and J. B. Carruthers, \Experimental characterization of non-directed indoor infrared channels," IEEE Transactions on Communications, vol. 43, pp. 1613{1623, February-March-April 1995.

10. D. Lee and J. Kahn, \Coding and equalization for PPM on wireless infrared channels," IEEE Transactions on Communications, pp. 255{260, Feb. 1999.

11. D. Lee, J. Kahn, and M. Audeh,

    \Trellis-coded pulse position modulation for indoor wireless infrared communications," IEEE Transactions on Communications, pp. 1080{1087, Sept. 1997.

12. H. Hashemi, G. Yun, M. Kavehrad, F. Behbahani, and P. Galko, \Indoor propagation measurements at infrared frequencies for wireless local area networks applications,"IEEE Transactions on

    Vehicular Technology, vol. 43, pp. 562{576,

    Aug. 1994.

13. J. R. Barry, J. M. Kahn, W. J. Krause,

    E. A. Lee, and D. G. Messerschmitt,

    \Simulation of multipath impulse response for indoor wireless optical channels," IEEE Journal on Selected Areas in Communications, vol. 11, pp. 367{379, Apr.

    1993.

14. M. Abtahi and H. Hashemi, \Simulation of indoor propagation channel at infrared frequencies in furnished office environments," in PIMRC, pp. 306{310, 1995.

15. J. B. Carruthers and J. M. Kahn,

    \Modeling of non directed wireless infrared channels," IEEE Transactions on Communications, pp. 1260{1268, Oct. 1997.

16. http://www.irda.gov.

17. IEEE Std 802.11-1999 Wireless LAN Medium Access Control (MAC) and Physical Layer (PHY) specifications.

18. http://www.canobeam.com.

19. http://www.terabeam.com.

20. Special issue of IEEE Communications Magazine on 'Optical Wireless Systems and Networks', Dec. 1998.

21. http://www.bu.edu/wireless/irguide. Cross-references Wireless Optical Communications. IrDA. See Wireless Infrared Communications.