Deployment

---

The software side of wireless bridging is fairly simple, although the physical installation of these bridges can be complicated. Choosing the right antenna and correct installation is crucial. With the wrong antenna, you'll end up with poor performance as well as inefficient use of available bandwidth.

Certain antennae are designed for straight point-to-point applications, while others are omni-directional, allowing point-to-multipoint bridging. An antenna installation consists of mounting the antenna, along with a lightning arrester, usually on the roof of the building. Consider using a professional installer to ensure line of sight, although to make the task easier, some vendors, such as Enterasys and RadioLAN, provide utilities to correctly aim the antenna.

In addition to the ability to connect two network segments, wireless bridge products offer other features; for example, performing both protocol and broadcast packet filtering. The configuration software that comes with many bridging products lets you choose which protocols you want filtered, and whether to allow broadcast packets to be bridged. By restricting certain types of Ethernet packets, you can save considerable bandwidth.

Many devices also offer an option that lets the bridge function as a router. And one of the most powerful options available on some of the bridges is the ability to do point-to-multipoint bridging. Moreover, Cisco, Lucent, and Pinnacle offer products that can accommodate more than one radio. By assigning each radio a different frequency, the capacity of the bridge and even the coverage area can be extended, or multiple buildings can be connected to form a metropolitan area network. But keep in mind that point-to-multipoint configurations can be difficult to engineer because radio interference, bandwidth considerations, and other issues must be taken into account.

How Far Can You Go

Because designers must take into account the RF effects of a system's Fresnel zone (the pattern of electromagnetic radiation that is created between a transmitting antenna to the receiving antenna), bridges should be elevated sufficiently off the ground to ensure reliable operation.

Producing a strong, reliable signal using bridging techniques can be challenging. According to industry sources, 2.4 GHz frequencies need to have clear line of sight, and the Fresnel zone must be at least 80% free of obstacles. There are also other major technical variables that affect range, e.g. radio output power, radio receiver sensitivity, antenna gain, and path loss.

An antenna transmits and receives radio waves. The focused strength of these waves is referred to as "radiated energy," which is measured in terms of gain in decibels (dB). Gain in wireless antennae is typically specified in dBi-the resulting decibel measurement in relation to a theoretic isotropic radiator, which is equal in all directions. The gain in the antenna focuses the transmitted signal towards the targeted area of coverage. It also focuses incoming energy on the receiving side.

Gain must be considered when selecting a bridging antenna. There must be gain on both the broadcasting and receiving side to establish stable links, but not so much as to exceed the legal radiated power limitations of 4 watts (+36 dBm) maximum effective radiated power (ERP). The ERP, the total amount of power actually transmitted through the system's antenna, is the product of the transmitter's power output, the cable's power loss, and the antenna's gain capability.

Figure 9.6: When designing an outdoor wireless link, the Fresnel zone, which is the elliptical area immediately surrounding the visual path as shown in this graphic, must be taken into account. Typically, 20% Fresnel Zone blockage introduces little signal loss to the link. Beyond 40% blockage, signal loss will become significant. The Freznel zone will vary depending on the length of the signal path and the frequency of the signal, although the radius of the Fresnel Zone at its widest point can be calculated by the above formula, where d is the link distance in miles, f is the frequency in GHz, and r is the radius off of the center line of the link in feet. Note that this calculation is based on a flat earth, i.e. it does not take the curvature of the earth into consideration. The effect of this is to budge the earth in the middle of the link. Thus for long links have a path analysis performed that takes the earth's curvature and the topography of the terrain into account.

Path loss is the attenuation that occurs when radio signals pass through air, and while generally easy to calculate, it varies with frequency (the higher the frequency, the higher the path loss per yard or meter). In over-simplified terms, at a given frequency, high radio output power combined with high antenna gain and good receiver sensitivity translates into improved range.

Most professional installers will add a fade margin to their overall range calculations. This margin will vary depending on the geographic and climatic conditions of different geographic areas. This may introduce atmospheric and multipath-related fading that must be added to the free-space path loss. It's not unusual to see engineered fade margins of 20 dBm for high-reliability applications. Your vendor or reseller should be able to help you with this; several of the vendors include technical information and range calculation utilities on their websites.

Of course, in the real world, you are subject to government regulations, which vary somewhat by jurisdiction. In the United States, the FCC imposes a number of restrictions on unlicensed radio operations, and vendors must gain FCC certification of compliance in order to sell their products. Key regulations involve acceptable waveforms, radio output power, and EIRP (effective isotropic radiated power), which is a combination of radio output power and antenna gain. (See Appendix III: Regulatory Specifics re Wi-Fi for more detailed information.)

All else being equal, receiver sensitivity (a radio's ability to separate a very weak signal from lots of noise) is the most important technical design element. This is where RF engineers earn their salaries.

Reliability

Because radio-based fixed-wireless systems (and that's what bridging products are) operate at relatively low frequencies, you won't experience many of the weather-related problems with them that are typical of the higher frequency home-satellite systems. You can further ensure reliability by turning to a professional installer, who can conduct a thorough path analysis prior to installation. By using directional antennae and building in an appropriate fade margin, professionals can guarantee 99.99 percent reliability in most areas.

Interference is the other side of the reliability coin, particularly for devices operating in the 2.4 GHz band. WLAN systems, cordless phones, garage door openers, and myriad other low-cost radio devices can affect a fixed-wireless system. Again, these problems can often be managed by using alternate radio channels or through antenna polarization.

Some of the reliability risk of point-to-point systems is associated with the systems' dependence on line of sight (LoS) communications-if LoS is broken, the link will go down. Syracuse University recently experienced just such an outage when a crane, participating in the construction of a new parking garage, parked directly between two buildings connected via Wi-Fi. Such situations can be dicey to manage, so as with any mission-critical system, you should put contingency plans in place.

Security

Security is always a top concern for wireless implementations, but it's not nearly as big an issue when signals are sent between buildings via a bridging set-up. The reason? First, most bridging products use proprietary radio-signaling schemes, and second, point-to-point systems employ highly directional antennae. Thus the most viable attack is literally a man-in-the-middle approach, though the man would need to be suspended in mid-air to intercept the signals. And if that's not enough, most vendors support some kind of encryption system.

However, there is always the risk of denial of service (DoS) attacks mounted using inexpensive, easy-to-conceal radios and antennae. In fairness, motivated criminals can mount DoS attacks on most information systems, so in that sense, wireless isn't unique. Nonetheless, this is just one more reason why an increasing number of organizations choose to implement their fixed-wireless systems in the 5 GHz bands, where more channels are available and the risk of interference is lower.

---