Wireless Connectivity Update

First responders have long relied on wireless radios for voice communication, but the systems used by different agencies often do not interoperate. Firefighters, police officers, and paramedics may not be able to communicate directly with each other, which can result in sub-optimal coordination that costs lives that could otherwise have been saved. 802.11 access points can provide a common infrastructure that ties together communications from all entities in a disaster area. Because the infrastructure is digital in nature instead of analog, advanced telephony functions can be implemented, further increasing the efficiency and effectiveness of the first responders. For example, one issue with analog “walkie-talkie” radios is that frequencies can become congested when many people talk at the same time. With a digital voice network, a hierarchy of communication can be established. Each individual might only hear communications from other members of his or her group, and a single member within a group can be designated to handle communications with other groups. This allows each individual to focus only on communications that are relevant to his or her immediate environment while still facilitating coordination between groups. Additionally, the 802.11 standard has provisions for cases when there are multiple access points on the same channel in the same location. Even if the access points of two agencies are not coordinated with each other (e.g. configured on separate channels), they should both be able to provide adequate coverage to their users.

Another limitation of analog radios is that they can only carry one type of data: voice. Even when devices are designed for data transmission—for example, the computer that is installed in many police cruisers, which might allow the officer to look up driving records, warrants, and so forth—the system is usually proprietary and custom-designed, which typically results in it being difficult to upgrade and non-interoperable with other devices. By contrast, an 802.11 network provides a standard way of carrying both voice and data traffic, meaning that a single infrastructure can carry whatever type(s) of data the applications want to put on it. New applications can be developed without requiring an upgrade to the installed base of hardware, speeding up the software development cycle and allowing innovative applications to be placed in the first responders’ hands faster.

Its ability to support rapid innovation is probably the biggest reason that 802.11 is being adopted by first responders. Examples include:

· Wireless video cameras that are worn by emergency responders to allow monitoring by off-site personnel. For example, a firefighter encountering a possible chemical spill could have a hazardous materials technician look “over his shoulder” remotely and advise the firefighter whether there was a danger.

· Wireless access to databases. One example is allowing police officers to look up driving records and criminal records from the field. Most modern cruisers have wireless uplinks based on slow first-generation cellular technology. An 802.11 link through a wireless metro mesh provides much faster and more versatile capabilities. Another example is a PDA-based tool that uses wireless links to look up information in a hazardous materials database. Firefighters can look up information on a chemical without returning to the truck or radioing back to base.

· Wireless safety monitoring systems. One example is a “wireless tether” that alerts firefighters when they have wandered too far from the other firefighters. Firefighters often work in low-visibility environments. Getting separated from the group can result in injury or death. Currently, firefighters address this by staying in contact with each other, but this slows down their rescue efforts. Another example is a wireless health monitoring system that reports firefighters’ vital signs back to a central console.

· Location detection. Technology exists that can use 802.11 signals to triangulate the location of a radio to an accuracy of a meter or less. This can be used to locate downed personnel or to quickly find critical pieces of equipment in an emergency situation.

One of the major disadvantages of 802.11 for public safety use is that it uses unlicensed frequencies in the 2.4 GHz and 5 GHz bands. This means that public safety personnel using 802.11 networks must compete with private networks using the same frequencies. Connect802’s experience with RF design leaves us with no doubt that interference with pre-existing transmitters is one of the major factors limiting 802.11’s performance. Clearly, a the task of a first-responder is too critical to allow it to be interfered with by someone’s cordless phone, baby monitor, or Internet browsing. To address this issue, the FCC and the IEEE coordinated to designate a band of frequencies near 4.9 GHz (4940-4990 MHz) specifically for public safety use. Example uses include: homeland security, fire, police, security services, and so forth. Usage of the 4.9 GHz band is tightly regulated, meaning that authorized users should find no interference from private networks. Many 802.11 vendors are transitioning their 802.11a and 802.11b/g devices to the 4.9 GHz band in order to capitalize on this demand.

In this month’s newsletter, we continue our discussion on the metrics of the receivability of an RF signal. Last month, we discussed the most commonly-known metric, Signal Strength. This month, we cover Signal-to-Noise Ratio (SNR) and its less-well-known relative, Carrier-to-Interference Ratio (CIR). Understanding the distinction between SNR and CIR is sometimes the only way to figure out why an 802.11 network isn’t working as expected.

In the context of analyzing an 802.11 network, the terms “Signal” and “Carrier” both refer to the RF waveform that has been transmitted and that must be received. To create that waveform, the transmitter encodes and modulates digital bits—1’s and 0’s—onto a carrier signal. The receiver must lock onto that carrier signal and then decode and demodulate the digital bits, just as they were originally transmitted. In order to successfully lock onto the signal and decode and demodulate the digital bits, the received signal must not only have a certain minimum signal strength, it must also be a certain minimum amount stronger than any interfering transmissions. SNR and CIR measure the strength of the received signal relative to the strength of the interfering transmissions.

Signal-to-Noise ratio measures the strength of the received signal relative to any RF transmitter in the same frequency range as the receiving radio. For example, to an 802.11g radio, a cordless phone or a wireless baby monitor could be a source of noise. Carrier-to-Interference ratio measures the strength of the received signal relative to any RF transmitter that is using the same transmission technology as the receiver. Essentially, to an 802.11g radio, any 2.4 GHz transmitter can create noise, but only other 802.11g radios cause interference.

The distinction between SNR and CIR is important because digital radios have much more tolerance for noise than they do interference. A non-802.11 transmitter has different transmission characteristics than an 802.11 transmitter, making it easier for an 802.11 receiver to filter out the unwanted noise. On the other hand, an unwanted 802.11 transmission has exactly the same characteristics as the desired transmission that the receiver is listening for, making the unwanted transmission much harder to filter out. To put this into numbers, an 802.11b/g radio might only need 6 dB of SNR in order to receive a 6 Mbps signal, but it might need 10 dB or more of CIR to receive the same signal. This can be an issue because manufacturers typically publish their SNR numbers, but seldom publish their more-restrictive CIR numbers.

Therefore, interference is less likely. When two 802.11 devices are configured on distant channels—e.g. 1, 6, and 11—then they are far enough apart that they don’t hear each others’ transmissions and interference is prevented, even when the radios all transmit at the same time.

Of course, in a well-designed 802.11 network, interference won’t be an issue, because all radios will be configured on non-interfering channels. Interference is most likely to result when two different 802.11 networks are installed in proximity to each other and the channel allocation for the two networks is not coordinated. Each network can then act as interference to the other.

Because noise sources are inherently non-802.11 devices, the only way to accurately measure SNR is with a spectrum analyzer. Although many 802.11 cards report a metric called “Signal-to-Noise Ratio,” this metric is not, in fact, a true measurement of SNR, but is usually a synthetic metric that is related to the number of bit-errors in the decoding and demodulation process. Some manufacturers acknowledge this by more-correctly calling the metric “Signal Quality”. To measure SNR with a spectrum analyzer, turn off all of your 802.11 radios and then measure signal strength in the area. Then turn the radios on and perform the measurement again. The difference between the two measurements is the SNR.

Measuring CIR requires that only other 802.11 transmitters be considered. This can be done with careful use of a spectrum analyzer, but distinguishing non-802.11 noise sources from 802.11-based interference sources can be difficult, depending on the nature of the noise sources. Some noise sources look similar to 802.11 transmissions in a spectrum analyzer. An 802.11-card-based tool like NetStumbler can be helpful when measuring CIR because it focuses only on the 802.11 transmitters in the area. At the same time, 802.11 cards have relatively low precision and accuracy, so their measurements of signal strength should not be considered to be highly reliable. To measure CIR, turn all of your radios off, then measure the signal strength of other 802.11 radios in the area. Then turn your radios on and measure the signal strength of your radios. The difference between the two is the CIR.

To Infinity... and Beyond!

Connect802 has written about 802.11n in previous newsletters. Our recommendation has been that enterprise networks should avoid so-called "pre-n" equipment because of uncertainty about whether that equipment would be compatible with possible future changes in the standard. But when we could expect to see standards-compliant 802.11n equipment has been unknown... until now! The IEEE task group working on 802.11n announced that a proposal for the final standard has been confirmed. Although the standard may change before its final ratification, this is a major step forward, representing a resolution of some issues that had been stalling progress on the standard until now. Industry analysts predict that 802.11n will move rapidly forward towards ratification, although it still might take until 2007 before we see a final standard.

Until now, few vendors of “pre-n” equipment were willing to make that guarantee, as too many aspects of the standard had not yet been finalized. The confirmation of an 802.11n proposal makes it much easier for vendors to build chipsets that are guaranteed to be compatible with the ultimate standard. Broadcom announced that chips based on the recent proposal have been incorporated into reference design wireless cards and are available for manufacturer sampling. These “Intensi-fi” chipsets support all mandatory parts of the current, draft 802.11n standard, and are reported to operate at modulation rates of 300 Mbps (actual throughput unknown). Likewise, Marvell announced reference chipsets and predicted the availability of pre-standard equipment this quarter. Marvell’s chips use optional parts of the standard to achieve modulation rates of up to 600 Mbps (analogous to 108 mbps “turbo-802.11g”). Finally, Atheros made a similar announcement, reporting 150 to 180 Mbps of actual throughput at modulation rates of up to 300 Mbps.

We plan to do a full analysis of 802.11n as final ratification approaches and more products hit the market. Until then, our recommendations remain basically the same: avoid “pre-standard” equipment in enterprise installations unless absolutely necessary. This recommendation is tempered somewhat by manufacturers’ new willingness to guarantee future compatibility via firmware upgrade, so the bar of “absolutely necessary” has been lowered a little bit. Nevertheless, issues remain: for example, even if an 802.11n access point is installed, will there be enough 802.11n clients to take advantage of its increased range and throughput? And what will be the effect of mixed-mode 802.11g/802.11n environments? We plan to try to find the answers to some of these questions, and we’ll report them here when we do.

The WiMAX Forum announced this month that equipment had been certified as compliant with the 802.16 standard. In the past, test networks were been built using equipment based on the 802.16 standard, but that had not been certified as compliant. This announcement marks the first time that actual 802.16 equipment can be available on the market. Consumers who buy this equipment can be assured that it will be compatible with 802.16 equipment from other vendors, meaning that the promise of WiMAX—a non-proprietary wireless metropolitan area network technology—can begin to be realized. The WiMAX forum announced that slots had been reserved for 26 more products, so more certifications are likely to follow, and the amount of WiMAX equipment available to the consumer is likely to grow.

The certified equipment included Aperto Networks’ PacketMAX 5000 base station, Redline Communications’ RedMAX AN-100U base station, SEQUANS Communications’ SQN2010 SoC base station solution, and Wavesat’s miniMAX customer premise equipment (CPE) solution. Network engineers in the U.S. should note that this equipment uses exclusively the 3.5 GHz licensed spectrum, which is not used in the U.S., and which is owned by large carriers in other countries, so this equipment isn’t legally usable in the U.S. American users should anticipate certification of WiMAX equipment using the 5 GHz UNII bands.

Connect802 resells equipment from Redline and other WiMAX manufacturers. We can provide design and/or provisioning for your WiMAX project and we would welcome your phone call to discuss the new standards and how they might be applicable to your design requirements.

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).

February 1, 2006