|This is Chapter 27 of Rory Conaway’s “Tales From The Towers” series about 2.4GHz for PTP Mission Critical applications across a city? What are you nuts? If you have not yet done so, read Chapter 26: Everything old is New Again.|
Although I live in Triadland and have been accused of not having a grip in reality…
Everything I talk about has either already been deployed or passed internal testing and is ready for deployment. Back in the real world, the fact is the government hasn’t assigned enough unlicensed bandwidth to make a big dent in competing against DSL, cable, or cellular companies. Even the cellular companies don’t have enough bandwidth to handle their client base and impose over-priced and ridiculous 2GB caps on some plans to keep bandwidth use down. This is despite the fact that they have paid enough money to buy a small country for exclusive rights.
Many ideas have been developed by manufacturers and WISPs over the years to get around the unlicensed RF limitations of lower power and higher frequencies. However, with the increased bandwidth needs of municipalities for video surveillance, traffic, water and sewer, it means that all of the unlicensed frequencies put together make it difficult to find enough bandwidth for everything. Even worse, most WISPs that are using 900MHz, which is the best non-LOS option , quickly get stomped the second a city decides to deploy a 900MHz water meter collection system.
Municipalities have one advantage that WISPs don’t have. Municipalities can use the 4.9GHz band. It’s not LOS and it’s woefully inadequate for high-definition video surveillance, but it’s 50MHz wide. Unfortunately, the high-power mask rules that some vendors pushed to the FCC precluded using mass- produced 5GHz WiFi equipment. That meant that anyone wanting to sell equipment in that band had to design a specific product for a much smaller market. It also meant that equipment costs will be very high when compared to some of the 5.8GHz mass-produced equipment. Basically, it’s pretty impossible to find a CPE in 4.9GHz that’s less than $1000 with any type of range due to the low-power mask limitations. Most of the high-power mask PTMP equipment in that band is limited to either 21Mbps (Motorola) with a 10MHz channel or is based on 802.11a (Solectek, Tranzeo, Alvarion) throughput specification which isn’t much more.
Some recent additions into the band by Radwin got my attention. They finally applied 802.11N 2×2 MIMO to the 4.9GHz band. Unfortunately, they also applied the multi-band licensing tier for the clients meaning they cripple the CPE to 10Mbps to make the entry level point cheaper and then charge you more if you want to upgrade the CPE to 20 or 50Mbps later. I thought that Ubiquiti pretty much killed that model by example. You would think that other companies would take a few notes after they sold more than $150 million dollars in the last fiscal year with around a 25% profit. It looks like it’s going to take Ubiquiti getting into that market next for companies to learn that their clients are tired of getting nickel and dimed for crippled equipment. This is another reason that municipalities are going to continue to use unlicensed frequencies with the lower cost for the equipment.
In one recent upgrade of an existing installation, due to frequency and budget limitations and on the experience of the project manager, we evaluated using 2.4GHz for several PTP backhauls, one of which was 1.7 miles. Yes, we also looked at 4.9GHz but since the vendor was paying for the upgrade out of their pocket, the difference was $40K versus $7K for equipment and there was no way they were going to pay that for less bandwidth. Normally for a public safety design, that option would pretty much be considered suicidal in terms interference, especially from the center of town out to the edges. In reality, most people would dismiss this idea outright from an RF standpoint. Even worse, it would be considered potentially unstable over a long period of time. However, if you start 80’ in the air with a 24dBi dual-polarity parabolic antenna pointing to another location 60+ feet in the air, there are very few retail indoor access points that are going to interfere. Between altitude and building attenuation, the signal is going to be greatly reduced. Notice I’m not suggesting using sector antennas in this equation.
We have to make several assumptions here with this idea. The first is that nobody else has deployed a PTMP system in 2.4GHz in a big way in the area. Second, that there is no mesh 2.4GHz system in place using overpowering APs. Third, assuming both sides are located with buildings more than a couple hundred feet away or more, that’s even more interference protection. Fourth, that hackers really have better things to do and easier targets to go after. The reality is that anyone who wants to block this type of transmission could probably cause a problem whether it was on 900MHz, 2.4GHz, 4.9GHz, or 5.8GHz.
Assuming that the parabolic antennas have between 6.5 and 10 degree beam patterns, the directionality means that anything within several hundred feet at ground level is going to be severely attenuated. Taking that further, keep in mind the 3-1 rule in which we get to increase the EIRP based way above the little Linksys router sitting a block away. With a 24dBi antenna, we get to set our power settings to roughly 24dBm. The EIRP on that is about 48dBm which is pretty cool. Throw in massive attenuation off the side and back and the reality is that for links less than 2 miles, 2.4GHz can still kick it with the big boys.
We are even using two pair of 2.4GHz Ubiquiti NanoBridges to handle relaying from another tower Ľ mile away which in itself is supporting about 20 cameras. In this case, the NanoBridges aren’t bundled together and are on separate switches. Together, they give us over 120Mbps or better of capacity in one direction from the hotel to the tower. The alternative was 4.9GHz with a max capacity around 100Mbps for anywhere from $3000 up to $20,000 or 24GHz unlicensed for about $7000. Considering a total cost of about $350 for this kind of bandwidth, it’s not a bad deal and it’s going to be very hard to interfere with in these locations.
There are other places that this concept could work really well, especially in areas where security and the potential replacement costs of frequently damaged equipment play a big part. I’m looking at some SCADA projects in Middle Eastern countries where this has great value. There is a huge umm, vandalism problem in some of the countries (yes, the ones we can legally ship to although apparently there isn’t a shortage of Ubiquiti equipment in the ones I can’t do business with according to their IPO filing). In some of these environments, replacement costs become critical, especially if you have to replace it on a regular basis due to various failures from outside influences that clearly aren’t covered under warranty. The second problem is that the RF landscape is pretty much the Wild West with no enforcement body to stop illegal transmissions. Low cost WiFi based equipment could be critical to getting municipal infrastructure in place that can be affordably maintained. However, PTMP is useless since the RF rules are enforced about as stringently as the traffic laws. This means that at any time, 3000 scooter drivers could be in your lane playing chicken with you at any moment (just came back from Vietnam and learned that my driving skills aren’t even close to those daredevils). I was hoping to get to GITEX in Dubai in October to talk to some colleagues that are doing projects in the Middle-East, but conflicts with WISPAPALOOZA in Vegas, where I’m doing a presentation, are going to conflict.
When your country is getting rebuilt from scratch on a regular basis, it’s an opportunity to jump ahead and create a foundation that can be deployed quickly and inexpensively to support basic needs. Instead of a bunch of standalone systems built by different entities, a strong foundational system can be built to not only support critical infrastructure such as SCADA and video security, but also to handle governmental data and voice needs or even population needs for Internet, educational, or business purposes. This type of system can be designed for redundancy and multiple points of failure so that catastrophic events can’t take the system down in times of emergency. The only issue is MTBF of the equipment. If you expect it to be stolen or destroyed periodically, an MTBF of 3 years is a pretty acceptable tradeoff for really inexpensive equipment.
As we discussed earlier, PTMP systems based on sector antennas that are difficult to make work in a high-noise environment. However, go back to the dual 2.4GHz directional antennas concept or any AP with a highly directional antenna and short range. High-bandwidth needs can be met with minimal chance of interfere with the signal if designed properly. In addition, nothing says you can’t overbuild a link with bigger antennas. Think of the antennas as shields for your signal as opposed to providing more gain. You can always adjust power, but antenna side and back rejection is fixed. And here is another thought, a 40MHz wide channel with a higher modulation rate normally has a higher potential for interference. However, if you have a lot of extra link path budget, it may be a better option than a narrower channel with lower bandwidth. This depends on the nature and type of interference
In most cases however, basic RF techniques such as narrow-banding, frequency-hopping, and others that we have covered in previous articles can create a system that can securely support critical municipal infrastructure. Because the equipment is so inexpensive and frequencies can be divided down to as little as 2MHz wide, a lot of PTP links could replace a sector based designed. Back it up with some basic network infrastructure such as bonded VPN tunneling, 1+1 architecture, RSTP, or any one of several different redundancy or multi-pathing techniques, and you can create a redundant infrastructure that keeps information flowing regardless of the reasons for failure.
Another alternative is to build a design that is both overt and covert. Using some higher frequencies and very small form factor equipment, you can modify the design to have antennas and radios that are not only physically hard to see, but you can play with the frequencies and broadcast protocol to hide the signals. Again, applying RSTP or VPN bonding, one radio becomes the target while the second radio is still functional. I used the technique on one high value target design. Also don’t be afraid to use the ground as a great place to put radios.
On the other hand, if you can be sure that your infrastructure is going to be around for the next 30 years, then investing in equipment that will last more than a couple summers in the desert (I get local training for that here in Arizona) can also be a good plan. Companies like Motorola (oops, now Cambium) and Alvarion build equipment that might last longer than the buildings they are attached to. Considering these guys use the same processes and manufacturing techniques to build military hardware, this equipment is going to cost a lot less to use and maintain over a long period of time.
This brings me to the article I saw on the Internet-In-A-Suitcase design. The federal government paid $2 million dollars for about $3000 in equipment and is all pretty much off the shelf equipment. This definitely falls under the category of “What are you Nuts?” or even better “Here’s your sign” (Bill Engval) for the federal bureaucrat that approved it. It highlights how clueless the government grant entities are and how good they are at wasting tax dollars (I would like to see a criminal investigation in this but $2,000,000 dollars is a drop in the bucket to these people). To make it even more funny, Iran thought this was a threat and actually came out publicly and announced that they could block it. If Internet-In-A-Suitcase is a big problem for the Iranian government, then I would be worried about them running microwave ovens, let alone nuclear power plants. That and they clearly need a better PR department or should at least talk to Obama’s teleprompter people.
After the tears of laughter ($2,000,000 dollars for a Ubiquiti Picostation and accessories? Again, “What are you nuts?”) finally dried up, I went to work and designed a real system called Technology Enhanced Communication SYStem (TECSYS) that I describe on my website. After evaluating how Iran figured out they were going to block the system from getting out of the country, I redesigned the transport to make it pretty much impossible for the Ayatollah’s boys to block it unless they want to put a force field around the country. Since the Russians and Chinese don’t sell one, I think TECSYS is safe. I’m not guaranteeing that users are going to be watching NetFlix in the middle of the desert but if Aunt Mabel wants to send me a congratulatory birthday email, I’m sure as heck going to get it. The system also has have several times the capacity of the Internet-in-a-Suitcase System. This also depends if you are traveling with a carry-on or plan on checking your baggage. I even added a few other necessary capabilities for a centralized logistics system. Now where is that government official that wants to give me $2,000,000 dollars for the idea?