Designing Outdoor Antenna Systems Part 2: Fresnel Zone Clearance and Fade Margin
Last month, we began a discussion on designing outdoor antenna systems. We talked about how to tell whether a link would have enough power to go a given distance by calculating System Operating Margin (SOM), and we saw how Connect802’s Antenna System Designer could easily perform SOM calculations. Read last month’s article. But all that power won’t do any good if the line-of-sight between the two antennas is blocked by trees, buildings, hilltops, or other obstructions. This month, we discuss how to ensure that RF line-of-sight is clear—it may not be as simple as you think!
If SOM is the first parameter that must be considered when designing an outdoor antenna system, then the Fresnel Zone is the second. The Fresnel Zone refers to a roughly football-shaped volume of space that is centered on the visual line of sight between two antennas. Imagine a laser beam running between the two antennas. This represents visual line of sight. Now imagine a football skewered lengthwise on the laser beam. This roughly represents the shape of the Fresnel Zone.
The Fresnel Zone is important because it contains the majority of the RF energy of the link. Any obstacle that is inside the Fresnel Zone will interfere with the link’s RF energy and may compromise the viability of the link. Fortunately, the size of the Fresnel Zone can be calculated based on the frequency of the RF signal and the distance between the antennas. Once this size is known, it can be determined whether any obstructions impinge on the Fresnel Zone.
The most common method of dealing with Fresnel Zone obstructions is to raise one or both antennas such that the Fresnel Zone goes above the obstruction. Therefore, Fresnel Zone considerations are commonly expressed as a minimum antenna height, meaning the lowest height at which the antennas can be installed while still giving sufficient clearance to avoid obstructing the Fresnel Zone.
The image to the right highlights the parameters of the antenna system designer that relate to the Fresnel Zone. In this mode, the distance to the obstruction is entered, the height of the obstruction is entered, and the designer calculates the minimum height of the antennas to clear the obstruction. Alternatively, the “obstruction height” parameter can be left at its default value of “calculate,” which causes the designer to assume that no obstruction higher than ground level is present. In this case, the minimum calculated height of the antennas is just enough to ensure that the Fresnel Zone clears the ground.
The screen shot above shows the results of the antenna system designer, this time including information about antenna heights. This calculation mode assumes that the two antennas are installed at the same height, but the antenna system designer also includes modes that allow you to enter one fixed antenna height and have the calculator determine the minimum height of the other antenna to clear an obstruction. That mode is useful if one antenna’s location is fixed (e.g. installed on the roof of a building) while the other’s is not fixed (e.g. installed at any point on a tower).
Fade Margin
The System Operating Margin and the Fresnel Zone (minimum antenna height) are the primary factors that are considered when designing an outdoor antenna system. Although these numbers can be calculated very accurately, RF systems are never static. As the environment between the two antennas changes, so does its effect on the RF link. For example, wind blowing against a tower can move the tower a significant amount (as any first-time tower climber finds out the hard way), causing the alignment of two antennas to change, and lowering signal strength below its ideal value.
In order to compensate for environmental fluctuations and inaccuracies in calculations, it is common to add a Fade Margin to RF link calculations. A Fade Margin (also referred to as Rain Fade Margin) is a small amount of extra attenuation that is added to the calculations in order to allow real-world results to be somewhat worse than the calculated results. Typical fade margins for outdoor systems range from 10 dB to 20 dB, depending on expected environmental conditions. The antenna system design has a parameter to enter a fade margin, which will be taken into account during SOM calculations.
Summary and Conclusion
The Connect802 antenna system designer is a powerful tool for quickly and easily performing the basic calculations necessary in designing an outdoor antenna system: System Operating Margin and Fresnel Zone (minimum antenna height). It can even handle situations that would be very difficult to calculate by hand, such as when an obstruction is not mid-way between two antennas or when the antennas are not installed at the same height. In addition to its basic parameters, the designer includes more sophisticated parameters, such as terrain roughness and refractive index, that can increase the accuracy of its calculations. For more basic users, these parameters are set to adequate default values and can be ignored.
Many of the topics that were touched on in this month’s article have been covered in more depth in past articles. Here are some links for your continued education:
Connect802 recently completed the first stage of a market survey comparing the specifications of popular 802.11 radios. The survey examined both client cards and access points. The receive sensitivity and the available transmit power options of the devices were recorded. The primary goal of the survey was to create a database that would allow Connect802 to better create wireless designs for its customers, but in the process of creating the survey, we noticed some trends that we thought would be interesting to our readers.
It should come as no surprise that enterprise-grade devices tended to have better specifications than their consumer-grade siblings. They were more sensitive and had higher output power, meaning that they would achieve better data rates and/or greater range, all other things being equal. Enterprise-grade devices can cost ten times as much as consumer-grade devices. Better radio performance is one way that they justify this cost.
Degree of Disclosure
It was unexpectedly difficult to even find the specifications for most consumer-grade devices. The most common specification made available was the transmit power. All of the devices in the survey published at least one transmit power number. The enterprise-grade devices usually offered adjustable output power and often advertised all available output powers.
At least one enterprise-grade vendor advertised different output powers for different 802.11b (DSSS) and 802.11g (OFDM) data rates. For example, 802.11b had a maximum output power of 100 mW, while 802.11g had a maximum output power of 50 mW. Because both standards operate under the same power output restrictions, one might assume that they would have the same output powers. This does not appear to be the case. In two cases, vendors even advertised different maximum output power for different 802.11g data rates! For example, 54 Mbps might have a maximum output power of 30 mW, 48 and 36 Mbps might have a maximum output power of 40 mW, and 24 Mbps and below might have a maximum output power of 50 mW.
Although Connect802’s engineers are well-versed in power output restrictions for 802.11 devices, we are not aware of any FCC regulation that would require a power reduction as data rates increased or as one moved from DSSS to OFDM. There may be some practical limitation in the engineering of the radio, as opposed to a regulatory limitation. If so, then we wonder whether all 802.11g devices reduce their maximum output power at higher data rates. It seems likely that the story of “output power” is not as simple as a single number. Vendors who only report a single number for output power may not be telling the whole story.
Receive sensitivity numbers were published less often than output power numbers. As with output power, enterprise vendors published more comprehensive information. They generally provided receive sensitivity numbers for all data rates that the device supported. One manufacturer of consumer-grade APs did not appear to publish any receive sensitivity numbers at all for any of the many radios in its product line! A few manufacturers published just a few receive sensitivity numbers—e.g. for 1 Mbps, 11 Mbps, 6 Mbps, and 54 Mbps, but not for any other data rates. This is better than nothing, but not as good as full disclosure.
Summary of Results
The receive sensitivity of access points at 54 Mbps ranged between -64 and -75 dBm. With a 9 dB difference between them, the best APs could be expected to have significantly better range than the worst. The majority of the APs had a receive sensitivity between -72 and -68 dBm at 54 Mbps, a difference of only 4 dB. A similar picture presented itself at the 6 Mbps data rate. Receive sensitivity for APs ranged from -82 to -93 dBm, a 9 dB range. The majority of APs were between -86 and -91 dBm, a 5 dB range. No conclusions could be drawn about receive sensitivity for client devices, due to a lack of published specifications.
Maximum
transmit power for most APs
and client devices was about
15 dBm. Enterprise-grade devices
tended to be closer to 20 dBm,
and a few consumer-grade devices
had transmit power as low as
13 dBm.
Conclusions
Many 802.11 devices have radio specifications that fall within a relatively narrow range. All other things being equal, these devices could be expected to have roughly the same range and achieve roughly the same data rate as each other. In general, enterprise-grade devices had better specifications than consumer-grade devices, but not universally.
Performance specifications alone should not be the only factor that influences a purchasing decision. This survey examined only two factors: transmit power and receive sensitivity. Additional factors that should be considered include: can the device be remotely managed; can the device participate in a light-weight architecture; can the device support the type of authentication and encryption that you want; how rugged is the device; how stable is the device? Because the performance of most 802.11 devices falls within a relatively narrow range, these considerations should take precedence over performance specifications. They are also where enterprise devices tend to have a clear advantage.
The performance of a few devices is much better or much worse than average. For example, Connect802 has worked with a mesh router that has an amazing receive sensitivity of -106 dBm! We know of another AP that outputs the maximum power allowed by the FCC right out of the box, no amplifier required! On the other side of the scale, we know of some devices with abysmal receive sensitivity. Although performance specifications should not be the first thing on your mind, you should make sure that they meet the performance requirements of your network. Connect802 can help you choose the right equipment for your needs.
Ask the Expert
What To Do About Multipath
I have multipath interference in my environment. Can I fix that by forcing my stations to use lower data rates—e.g. 1 and 2 Mbps instead of 11 and 54?
Different 802.11 data rates are created by varying the complexity of the signal. More complex signals can carry more 1's and 0's in the same amount of time, but they are also more susceptible to corruption, and vice versa for less complex signals. Specifically, more complex signals require a greater signal-to-noise ratio in order to receive them correctly. Assuming a constant amount of ambient noise, that statement is equivalent to saying that higher data rates require more signal strength.
Multipath cancellation causes a decrease in signal strength in a specific location. Depending on the amount of the decrease in signal strength, lowering the data rate that the station uses might be sufficient to address the problem, since it will allow the station to receive the packets even though they are weaker. Forcing stations to lower data rates might also address multipath by forcing the station to switch from OFDM to DSSS, although that’s not as certain.
The real question that I would pose is, "What makes you think that multipath is the issue, and not something else, like interference or a shadow from an attenuating obstruction?" In my experience, it's very easy to say, "Signal strength is low here," but very hard to say, "And that's because of multipath." This matters because if you are specifically tackling multipath as the issue, you might not be addressing the real problem, which could be something else.
Whatever the issue, decreasing the distance and obstructions between the client and the AP by relocating the AP or adding another AP will probably solve it. That's not always the preferable solution, but as a last-ditch effort, it usually works, at least up to the point where you have so many APs that you run out of channels and get interference.
In August, Connect802 broke the ONE BILLION SQUARE FOOT mark for indoor and outdoor wireless system design.
The Wi-Fi designs have included in-building Wi-Fi, hot zones, and metro area wireless networks. Various technologies were employed based on the particular project, including 802.11 b/g, 802.11a, 4.9 GHz public safety systems and point-to-multipoint backhaul including WiMAX.
Jason Davies, President of Davies Computer Solutions, Inc., a targeted solutions provider offering customized technology for the business market, summed it up by saying, “Connect802 has helped us succeed in a number of diverse project areas bringing up Wi-Fi systems in offices, hospitals and warehouses. Having the flexibility to use predictive modeling and on-site consulting helps us provide the highest quality, start-to-finish wireless solutions.”
Pre-n Devices Popular; Performance Unpredictable; No Standard Until 2008?
A market survey by In-Stat reports that approximately 300,000 pre-n routers, clients, and APs shipped in Q2, 2006. Buyers of these products are reported to be “early adopters willing to pay two to three times the price of standard 802.11g products.” Pre-n chipsets are expected to make up only 3.6% of total WLAN chipset shipments in 2006, but almost 20% in 2007. Chipset manufacturers Atheros, Broadcom, and Marvell continue to produce chipsets based on the 802.11n v1.0 draft, and Intel is set to release an 802.11n wireless module for mobile devices in early 2007. This indicates that manufacturers and cutting-edge consumers are not overly concerned about the lack of a standard, even though it may hurt interoperability and eventually render pre-n equipment obsolete.
The instability of the 802.11n marketplace is underscored by a PC Magazine review of NetGear and Linksy’s latest pre-n routers. PC Magazine’s “Bottom Line” sums up our point better than anything we could write: “I was extremely disappointed with the WRT300N's results on our performance tests” and “ The WRT300N’s … mediocre performance is disappointing for a product in this class.” Regarding the NetGear WNR854T, they write: “When it comes to performance, the WNR854T was a disappointment in my book. Only after firmware and driver updates did it deliver the numbers you see in this review. Yes, 123.5 Mbps at 10 feet is fantastic, but 27.7 at 120 feet and 2.8 at 160 feet are less than impressive.”
So, pre-n devices have somewhat higher throughput than 802.11g devices at realistic ranges, but does this justify the double or triple price that pre-n devices currently demand? So far, these devices seem to give only a moderate throughput increase AP, require firmware updates out of the box in order to resolve basic connectivity issues, and may become obsolete at some unspecified time in the future. Should you pay two to three times as much for that? Connect802 agrees with the NetGear’s reviewer, who says, “At this time, I can’t recommend that anyone run out and buy [this product].”
Sadly, anyone who is waiting for standards-based 802.11n found out this month that they will have to wait a little longer. You might recall that the first vote on draft 1.0 received 12,000 comments—an unusually large number. ARNnet.com reports that about half of those comments were minor editorial changes and have already been resolved, but apparently the remaining 6,000 will require more serious attention. The IEEE hopes to have a second draft completed by its November, ’06 meeting, for a January, ’07 vote. Even if that happens on schedule, final specifications may not be available until some time in 2008!
802.11 was designed for cases where client devices and APs are moving slowly relative to each other—at speeds of about 30 MPH or less. At higher speeds, reception of RF signals can be more difficult. The robustness of the 802.11 RF architecture was confirmed recently when Caltrain, a commuter train service in northern California, successfully tested an 802.11 system on its trains at speeds of up to 79 MPH. Internet connectivity was not interrupted at any time during the test.
Wi-FI Net News concisely sums up the difference between WiMAX (802.16) and Wi-Fi (802.11): WiMAX is intended for licensed spectrum, so contention between stations is eliminated. Wi-Fi is intended for unlicensed spectrum, so contention is a fact of life. Yes, WiMAX does have profiles for operation in unlicensed bands, but no unlicensed profile has actually been approved, and no unlicensed-band WiMAX equipment currently exists. Read the whole editorial for a deeper explanation.
Linksys and NetGear draft-n routers
Web Searching: The Connect802 Web Presence
At Connect802 we're your PAGE ONE resource for wireless networking!
Connect802 has the experience, expertise, and resources to help you with your wireless network system. Use your favorite search engine and see what Connect802 is doing. Each month we give you some suggested search terms for you to explore. Here's this month's list. As you look down the search engine results you'll find Connect802 either at the top, or on the first page (true for Google and Excite, unknown for the rest).