Hello, I am the Technical Product Manager for Motorola’s RF Design and Management group.
We are investigating the possibility of a “Canopy Planner” product, and would like your feedback regarding
some features we are thinking of. Feel free to volunteer any other features you feel would be useful.
We intend to help Canopy customers with the RF side of planning, i.e. equipment placement
on a map to determine if a link could or could not be made. We actually have some Canopy planning
abilities already built into our MeshPlanner software, however this software is designed with the Mesh
customer in mind, so we would like to expand the capabilities and make a new product that is specific
to Canopy. Here is a brief explanation of what is already available, and would be included in the new product.
The user imports GIS information into the software (terrain, clutter, roads). This can be free map info available
from the USGS (for the USA), or it can use higher resolution data that they would purchase. Obviously, the
more accurate and high-res the map data, the more accurate their results. They can then place one or more
Canopy AP’s. This can be an AP Cluster, a single AP, or anything in between. Power, orientation, height, etc
are all user-configurable. Right now this is limited to Canopy products in the 5.x GHz band. Given these items,
several workflows are available:
- An RSSI prediction grid over the entire map, which would indicate where LoS from AP’s is available at a given
height. This tells the user where it is likely that an SM would be able to make a connection.
- Export the aforementioned grid prediction as a Google Earth layer, so that individual addresses can be checked,
perhaps to determine if they are eligible to become a WISP customer.
- Given a set of potential SM locations, output the best (highest RSSI) link back to a specific AP. Also, output a
spreadsheet showing other potential links as well, if multiple AP coverage is present. Someone in the field could
use this to tell them what azimuth and elevation is necessary to link back to each possible AP.
This is what is available currently. Obviously, the more accurate the map data, the more accurate the results. In the US,
the USGS provides a lot of data for free, albeit low-resolution. Even with this, a lot of useful information can be plotted,
for instance just using terrain will show what areas would be blocked by the earth. Outside the US, there are other free
or low-cost providers of information, and purchasing data from a GIS company is always an option.
For 2009, we would like to take the above feature set as a foundation, and build a stand-alone Canopy Planner product.
Here are the features we are considering:
- Add support for the Canopy 400 OFDM hardware. With the near LoS capabilities, this means that our method of
propagation simulation would be more accurate, since now one tree branch doesn’t ruin a link.
- Possible support for the other frequencies as well. Is 900MHz planning really needed? What about 2.4GHz?
- Application planning. Given a set of radios and map data, give some throughput and capacity estimates, then report
back if this meets the criteria for a given application.
- Automated measurement drivetest, drive around with an SM, collected RSSI link info automatically via GPS, plot this
on a map.
What are your feelings about such a product, what else would it need to do to get you to buy?
Look forward to your feedback,
mike
Personally we already pretty much do this using Radio Mobile to create the map and then overlaying it with Google Earth. It is a little time-consuming at first but we’re using it now almost exclusively for phone site evaluations, and then if it is on the fringe of getting a signal we will still send someone out for an on-site evaluation.
Obviously the costs and additional benefits would be of concern for us.
We are exclusively almost only 900Mhz with our towers and only have a few 5.7 & 2.4Ghz but if this were to be made I would hope all frequencies would be able to be plotted.
I know link there is a link planner for point-to-point backhauls but one overall product that we could use for Point-to-multipoint and all point-to-point devices would be nice.
pcpolo wrote: Personally we already pretty much do this using Radio Mobile to create the map and then overlaying it with Google Earth. It is a little time-consuming at first but we're using it now almost exclusively for phone site evaluations, and then if it is on the fringe of getting a signal we will still send someone out for an on-site evaluation.
Obviously the costs and additional benefits would be of concern for us.
We are exclusively almost only 900Mhz with our towers and only have a few 5.7 & 2.4Ghz but if this were to be made I would hope all frequencies would be able to be plotted.
I know link there is a link planner for point-to-point backhauls but one overall product that we could use for Point-to-multipoint and all point-to-point devices would be nice.
What features/benefits would convince you to use our product instead of Radio Mobile? (Hard to compete with free...)
The PTP Linkplanner product is very good at what it does, and our product would not be able to handle maps as large as what is needed for really long haul links. I'm all for consolidation (I am also the TPM for Wireless Manager, which handles the management/monitoring of all these products) but I don't think I can make the case for consolidating Canopy Planner and PTP Linkplanner. (very different platforms) How common is it that a Canopy WISP would also use Orthogon equipment?
If you can offer a product as reliable as the PtP link planner, Its on! Keep it FREE, radio mobiles headaches are worth not having to pay a dime for something that should compliment a product.
If you could add in a simple way to estimate foliage, it would be awesome since the residential market deals more with that as obstruction than the PtP market.
a ten percent corn loss calculator built right in would be grand
The ability to map our current subs to see what self interference issues are present.
A power management feature where we can key in the installed values, which correlate to the estimated values and give us power recomendations so we can be clutter free by turning down power where possible.
A sales end interface on this would be great - example, a customer calls into sales and says my address is X, sales puts it in, it looks into the tool, if its within the configured range its a positive. Customer estimates of distance to APs is rarely right, this feature of course, would be a pay for add- on
this in my world would be grand
Mike, it is nice to see that Motorola is working on development of a product for the Canopy line. Naturally a product such as what you are proposing will cost something, but as the others have said, free would be nice.
I will agree with everyone else on what they have asked.
I’d like to personally see the 900MHz coverage prediction, with near/non line-of-site prediction and if it is possible, analysis of ducting and reflection and refraction in the 900MHz band. This however could result in extremely high costs associated with the software so take my request with a grain of salt. 
The OFDM platform analysis would be great. Case in point: Last year I spend some time doing transmission engineering on a federal project (I am not going to discuss it here in the open forum as it has not yet been approved) using the 5.4GHz OFDM PTMP platform for coverage. It would have been extremely beneficial to us to have access to a prediction software like you propose, as the area of coverage is extremely dense in foliage coverage. I had asked the federal systems engineer I was working with to please have the area analyzed as part of Motorola’s portion of the job, but I believe he was unable to accomplish this as the 5.4GHz OFDM product line was too new at the time. On our end, I had a preliminary rough (VERY rough) coverage plot developed at 5.4GHz to offer some kind of prediction so we could attempt to proceed with the project. I don’t believe Radiomobile takes into consideration all of the factors surrounding OFDM modulation techniques when it produces a coverage plot, so we had to make some educated estimates on the projected coverage area and the guys at Mother Motorola were satisfied with the data from us.
The application planning would be nice to have, as would be the drive test measurements. Currently, after I turn up a system I manually drive test and plot the as-built coverage so that I can compare it to the system design data. Obviously this is a time-consuming and expensive task (labor intensive), but the resulting empirical data is required for proper system turn-up and acceptance. Usually, it is only the large carriers and providers that do this type of testing. We are not a big company but customer satisfaction, system reliability, and ultimately profitability demand that we do this. Having a tool that is purpose-built and automated for such as task would be nice to have. Perhaps you can have this built in as a module, or option so that those that want to use it can purchase it separately to keep the costs down (I am assuming what you are planning is a pay-for-use product).
If I think of anything else, I’ll be sure to reply.
Obviously, free would be nice. But I think I said that already… 
I normally would never disagree with AMD, but free is the only way to go, the link estimator tool is free, its the same concept, a bunch of point to point links that all happen to have one radio in common.
I cannot stand companies that have products with poor documentation (lets face facts, the canopy manuals are awful, tier 3 support even told me that) then release a tool that would be great to have, but want a bunch of dough for it. Awful customer service if you ask me.
Back to the topic though, will it come with a free tower monkey?
"How common is it that a Canopy WISP would also use Orthogon equipment?"
We are moving all our primaries to PtP500. The link planner tool was a big selling point, had we had to pay for it, we still would have gone with PtP, but we would RM the links instead.
I am getting the general impression that free is the way to go. I suppose we will make up for it in volume…
I have been looking over the Radio Mobile tool, it appears that the map data is all do-it-yourself. I assume that method would be ok for our product as well. Out tool has the ability to bring in low-resolution clutter, cut the streets out of it, and apply different heights and attenuation to create obstructions from the clutter. One idea we had is to make this item somewhat more dynamic. i.e. you can vary the height of clutter on the fly. This would enable the customer to try out best and worst case scenarios, perhaps the truth is somewhere in the middle. Most of the map/clutter stuff is a guesstimate anyway, unless you had a recent LIDAR of the area in question, you’ll never know for sure if something will be in the way. Our goal is to bring the odds a little more in your favor.
I promise the documentation for this product will be stellar ;), I will probably be the one to write it, as I did for our MeshPlanner product. If this is to be a free product, the documentation will most likely be the only support available.
Steve, regarding your idea of mapping current subs, this is exactly the type of item we are looking to add to the Wireless Manager product. All of that data is there, and would be silly not to use it. Your “customer calls in and sales checks them against the map” is something we can already do today, simply by exporting the prediction as a KML in Google Earth.
AMD, yes reflection/refraction is somewhat of a pipe dream. Mainly because the clutter data is not very accurate to begin with, so calculating reflections on it is a fool’s errand. I need to do a bit more market research into the 900MHz variety to determine most bang for our development dollar. A note here, we are not actually in the Canopy division, we are the RF Design team (http://www.motorola.com/rfdesign), btw that’s yours truly in the MeshScanner photo Your Canopy OFDM project sounds interesting. We are in the process of testing some of that equipment ourselves, to determine how best we can model it, and see how accurate we can be.
I figured just as much for the in-depth analysis. No problems, it is just a nice dream.
I look forward to see what you guys come up with. I think whatever it is, it will be a great product, as everything I have seen from Motorola in terms of their major design tools has been excellent.
I’d be surprized if the product was actually free though.
Having just trudged my way through radio mobile a couple months ago, I can say I’m quite pleased with my results, but boy, that is not an easy program to use. I’m glad to see that an alternative is in the works!
This could be immensley useful for someone researching the viability of starting a small WISP. I wouldn’t want to drop a bunch of money on a piece of software that is going to be limited to one type of wireless equipment when I don’t yet know what brand I’m going to actually use. If it’s free though, you might have just encouraged a guy to lean toward Motorola from the get-go…
Anyway, as far as features I’d like to see go:
-2.4, 2.4, 2.4, 2.4! Our current network is all 2.4 with 5.7 for backhauls. We have one 900 (that I still want to move to another location.)
-Google maps integration (don’t know how easy that would be) would be fantastic. I like using it over google earth, just for convenience, half the time. A coverage map on our website would be handy for potential customers to see if we’re in their area at all.
-Keep in mind reflectors, and third party equipment (such as the cyclone 120 and omni’s). Or at least let them be able to put in their own antenna profiles if need be. We don’t have the population density for more than one AP per tower.
Yes, SteveD brought up another point. The ability to put in different types of antennas at the SM level as well as omni-directional antennas on the APs would be helpful.
Being able to import KML’s on Google Maps is a pretty tall order, you have to license the API from Google, and that cost can add up. I expect we’re going to stay with the Google Earth model. If nothing else, customers could download your KML and try it themselves. People love to play with Google Earth
Modeling different antennas for the SM’s is a requirement. We’ll probably have a selection of the common antennas, and the ability to model your own. This functionality already exists in our other design products.
modeling antennas in RM is a pain in the neck, getting antenna patterns is a bear. But if you were too look into the jist of it, I think you could implement an easy antenna pattern integrator, where we key in the values and the frequency, your program creates it. As a cost recouperation tool, you could have an antenna modelling service, where WISP pays you X to create verified accurate pattern for them based on manufacturer specs, even if you locked that pattern down to that user so they couldnt share that (we all know how much moto loves those lescenses) we would probably be willing to pay for that.
The GM API could also be a Paid for service
I think a free basic profiler with add-ons that cost is a great model, so long as the base system actually functions.
Everyone here who uses RM knows the results are not fully reliable, but we all know how to read into the unreliable results to get useful and valuable info. Caveating the profiler with that, its solid gold, and I dont think anyone here would expect the guarantees made on the PtP estimator, especially with the 900 model.
According to our sales rep, 900 is your best selling canopy line, so a 900 model would be well advised. Of course our sales rep knew we were looking into 900. the sales rep when we got our 2.4 told us 2.4 was the cats meow too.
thatoneguysteve wrote: modeling antennas in RM is a pain in the neck, getting antenna patterns is a bear. But if you were too look into the jist of it, I think you could implement an easy antenna pattern integrator, where we key in the values and the frequency, your program creates it. As a cost recouperation tool, you could have an antenna modelling service, where WISP pays you X to create verified accurate pattern for them based on manufacturer specs, even if you locked that pattern down to that user so they couldnt share that (we all know how much moto loves those lescenses) we would probably be willing to pay for that.
The GM API could also be a Paid for service
I think a free basic profiler with add-ons that cost is a great model, so long as the base system actually functions.
Everyone here who uses RM knows the results are not fully reliable, but we all know how to read into the unreliable results to get useful and valuable info. Caveating the profiler with that, its solid gold, and I dont think anyone here would expect the guarantees made on the PtP estimator, especially with the 900 model.
According to our sales rep, 900 is your best selling canopy line, so a 900 model would be well advised. Of course our sales rep knew we were looking into 900. the sales rep when we got our 2.4 told us 2.4 was the cats meow too.
Steve, we already have an "antenna creation wizard" which simply wants the freq, max gain, and vertical and horizontal beamwidths. For most of the simple antennas, this is plenty. The user can also bring in the patten numerically, but yes, that info can be hard to get.
Honestly, I really don't see the ROI on supporting Google Maps. I am rapidly getting the impression that no one is going to want to pay for anything. We'll see.
What do you mean when you say "caveating the profiler"? Are you referring to the path profile utility itself? Are there actually guarantees with the PTP LinkPlanner tool?
I'm investigating the percentages of Canopy sales by frequency as well. I was concentrating on the 5.x varieties, simply because I believed these were more desirable, but harder to implement. With the 900MHz line, what kind of design issues do you encounter?
Thanks everyone for your comments, this is extremely valuable!
jmwatts wrote: Being able to import KML's on Google Maps is a pretty tall order, you have to license the API from Google, and that cost can add up. I expect we're going to stay with the Google Earth model. If nothing else, customers could download your KML and try it themselves. People love to play with Google Earth :)
Modeling different antennas for the SM's is a requirement. We'll probably have a selection of the common antennas, and the ability to model your own. This functionality already exists in our other design products.
Your suggestion of customers downloading the kml and using google earth themselves is kind of a chicken and egg problem... the quality of dialup in this area is atrocious to the point where GE is unusable for the most part, hence why we are here! :)
Nonetheless, I sort of figured it wouldn't be easy to integrate with Google maps for reasons like that, based on prior, brief experiences using GM, but thought I'd throw the idea out there anyway.
We were told if we run a complete link estimator tool that says a link will work, and it doesnt work, we can return the equipment. (I dont know if this came from Moto, or a Vendor).
guys who want the GM API would be willing to pay for it, WISPs are odd characters in what we will and what we wont sink money into.
we just got and deployed our first 900 this week. It would have been really nice to have this tool when we tested the coverage. We could have compared test values to the estimated values, which would than have given us what we needed to know to estimate whether new customers were a good bet.
I didnt have time to do RM on this.
YOu really dont have a marketable product for resale in itself. Additional features is where you could rake in the cash. If the base system is solid, we will buy into features we like. If you go over to dslreports wisp forum, you will see when a WISP finds somehing they like, they get really really loyal to it, moreso than the family dog sometimes.
I couldn’t imagine bothering with this type of software for anything except 900. Currently we use Radio Mobile to pre-qualify everyone on the phone before going on site and attempting an install. The one thing radio mobile cannot do that I could see a use for is the ability to estimate Canopy coverage through foliage. All of our sales, admin, network, and installers know how to use RM and are getting better everyday at reading between the lines to learn what works and what doesn’t.
Radio Mobile’s new version has the ability to do a polar or visual coverage over an area and when you save the picture it also saves a .kml for opening as an overlay with Google Earth. This allows easy geocoding of the customer. You put in the customer address and then can what signal area they are predicted to receive.
I see the following features as what would make your product worth considering over Radio Mobile:
1. The ability to estimate foliage penetration with 900.
2. Ability to put in clusters or omnis (with custom antennas) as AP sites. We use mostly MTI antennas and Cyclone products but have a few other types of antenna as well.
3. Ability to put in current customers or tested locations data and use that data to compare the profile from APs to new SMs.
4. Need to be able to use the 1/3 Arc Second data from USGS.
5. Ability to do a coverage map showing how two 900 sites might interfere with each other in certain areas.
- Matt Jenkins
Steve, I asked around, I don’t think returning the equipment if it doesn’t work in Link Estimator is SOP anymore… I’d check with your rep before depending on this.
Matt, why wouldn’t you use planning software for 2.4 or 5.x?
1. The ability to estimate foliage penetration with 900.
This is doable, would just take a measurement campaign to figure out some averages. We’ve already done this in the mesh world. However, realize that for this to be useful, you’d have to bring in foliage data on your map, and pretty high-res at that. Are you willing to spend the money to buy this data? The software is certainly capable of this, but we were thinking users were more likely to get the low-res 2001 Land Clutter data, which does not differentiate obstacles.
2. Ability to put in clusters or omnis (with custom antennas) as AP sites. We use mostly MTI antennas and Cyclone products but have a few other types of antenna as well.
You’d be able to use whatever antennas you want, and can even make your own. I address this earlier in this thread.
3. Ability to put in current customers or tested locations data and use that data to compare the profile from APs to new SMs.
You could certainly put markers for existing SMs. What do you mean exactly by compare the profile? I’m not sure what you would be comparing. Keep in mind, just because you have an existing customer with a good connection at 123 Main St., doesn’t mean a link would work 2 blocks down, there could be a tree blocking the new location, and this is not going to be reflected on the map unless you buy high-res data.
4. Need to be able to use the 1/3 Arc Second data from USGS.
Can already do this.
5. Ability to do a coverage map showing how two 900 sites might interfere with each other in certain areas.
We can do SIR plots today. If you put two sites on the same channel and plotted it, you certainly see the SIR go bad where they compete.
We already do the KML output.
"Steve, I asked around, I don’t think returning the equipment if it doesn’t work in Link Estimator is SOP anymore… I’d check with your rep before depending on this."
Thats nice to know. Of course if the bosses get a PtP, even if it doesnt work where its intended, we still have 20 some links waitin on em.
were you just taunting us?
Nope, this is going to go forward. Will most likely concentrate on the 5.x GHz and the 5.4/4.9 OFDM products for the first round. We have a tentative release of Q3 of 2009.