Batboard

I'm looking at linking 3 tower sites together using Canopy and based on what I've read, Back Haul Canopy modules would probably be best for this application.

Details of the 1st Canopy shot:
It is line of site, approximately 7-8 miles point to point. I'm not sure if a reflector will need to be used.

Details of the 2nd Canopy shot:
It is line of site, approximately 18-19 miles point to point. Moto claims that some Back Haul modules can do 35 miles LOS with the reflector. I would think for something this far would require a true microwave solution instead of Canopy. Also with this shot it is sometimes difficult to see or impossible to see the other tower site during bad weather and foggy conditions. I also don't know what impact that will have on the Canopy link.

This link will be used for IP connectivity at the tower sites, hopefully to monitor and/or reprogram Quantar stations utilizing ethernet based terminal servers, there is also a possibility of installing some sort of monitoring system at the sites for door contacts and possibly video, but the available bandwidth will determine the video part. Also there are no plans to have a voter or comparator operating off of this system, just IP traffic only.

I'm hoping somebody with Canopy experience will be able to put some closure on this idea. I've never worked with Canopy or any form of microwave before, so this is new learning curve for me. Any info or ideas would be appreciated.

Thanks,

Stu - Personal or professional use? If personal, then you have a lot more latitude than a commercial application. You can put up with outages, low throughput, and intermittent operation. Believe me. When a customer is warned that the equipment they have chosen will not support an application well, their memory suddenly goes blank when it proves true.

First chose your band - 900 has the least throughput but the best range. Does not work well at a busy RF site. Tends to get clobbered by paging and trunking systems. Great for SCADA applications such as door monitoring, but you might get time outs on the Quantar serial port. Will not support video. 2.4Ghz is already very busy with consumer devices. If you are rural and shooting to rural, it might be usable to you with several thousand kilobit throughput and moderate latentcy. 5.7Ghz has your best throughput, but possibly lowest range depending on the model you choose. The OFDM models are far more robust than the standard or Advantage series. OFDM is the model of choice if you want to support video over ip over mountainous terrain with adverse weather conditions. Alignment with the OFDM is simpler as well. It also has better diagnostics than the standard Canopy series.

Second, do your path calcs. Canopy has a good path calculator to download that will help you determine if your shots are actually feasible. Be advised they use the 3 arcsecond uncorrected datum with no terrain datum to keep the software package down to a reasonable size. If you want to do a better study, goggle radio mobile from Cplus, and join their Yahoo group. Excellent free prop study software with lots of Canopy people already using it to help. Either way, do some kind of study to ensure you have sized your antennas gains and power levels for your link budget. Give yourself plenty of margin - at least 20db for fade and normal weather. Also, look out the trees, lakes, and rivers that jump out of nowhere to make your path fail. You will learn how to pronounce fresnal.

For the short shot, a standard 5.7Ghz AP SM pair configured as a backhaul with reflectors or the high gain antenna option for both will serve you well for your limited application minus any hopes of live video. Still frame or buffered narrow band video would be okay. PTZ will be sluggish. The SMMB mounting system for Canopy is a bit winky. You will have to ruggedize it if your area experiences severe weather, or fab your own. I've learned to not fight gravity, and mount the J post upside down with the long side down. Zero opportunity for the top heavy Canopy to flip over and become a water bucket when it is on the low side of the mount already.

For the long shot, I'd advise you to step up to the OFDM 30/60 or 150/300 series connectorized version so you can put high gain antennas on. The connectorized version requires either two antennas, or a dual feed antenna since you will have the vertical and horizontal beams to transmit. You can get away with using just one single feed antenna, but you will lose the advantages of OFDM. The panel antennas will function, but the plastic housing tends to degrade in the sunlight. Over time raptors and other large birds will break the housing and water will destroy the antenna in about five years. A dish antenna is strongly advised for long term use, and especially if you have any opportunity for ice to build up. The 30 mbps can be upgraded to 60, and the 150 can be upgraded to 300 if you need the additional throughput. Latentcy is an issue with the OFDM radios, but they will support video if the path is stable.

I'll let you chew on tht for a bit. Let me know if I can help further.

Ok, Thanks Bill for your input and this is for personal use. I was leaning towards 5.7Ghz but, I wasn't sure if that would be the correct choice for the 19 mile shot because of the long range.

For the 7-8 mile I take it the latency would be to high for video or the throughput would be to low? What about if the video stream was SIF resolution around 1300kbps streaming video? Video link on this setup was an afterthought, what I want to see at both sites is capability to monitor/reprogram the Quantars and maybe the door to the site.

I'll look into joining that Cplus group. As I would prefer feedback on systems similar to this magnitude. Not to say your input was unappreciated. I do appreciate your input and will most likely pursue this 8 mile shot first.

The latentcy is fixed at 20mS for best link conditions at maximum throughput. As conditions change, generally for the worse, your throughput decreases. As jitter increases, and more error correction utilized, latentcy will increase and throughput will decrease. Even under the best conditions, I've never gotten better than 3.5mbps in a network saturation test of Canopy links because they are not true full duplex radios. The OFDM series are, but the lower tier products switch between tx and rx cutting their end to end in half. Notice the literature says aggregate throughput - the sum of both directions. So, under the best conditions with the best link, we start with 3.5mbps, or 3500kbps. Now factor in link losses from distance, and/or atmospherics, and throughput plummets to the point where even 1300kbps could be more than it could handle. People do use Canopy successfully for highway cams with some long hops. What CIF rate they used I don't know. I've worked with some that ran qcif for security cameras quite well, but I let the camera guys worry about how to encode the video. I just provided the infrastructure, and made the necessary changes to the project to carry the loads they speced. So, you may need to work with your encoder / decoder to find acceptable resolutions for your application through a Canopy link.

You should know that going with the Advantage series doesn't necessarilly mean an immediate doubling of throughput. It just means the possibility of 2X mode. Good rssi, low jitter, low multipath interference, no site noise issues or competing systems taking time slices, and a strong wind at your back, you will see increased throughput. But, don't go in starry eyed thinking it is The Solution to your link issues. It will toggle back and forth between 1X and 2X mode as conditions change. OTOH, the Advantage series does seem far more robust than the standard series. I have far fewer problems with Advantage product than standard product.

Definitely get to learn RMD. Great program, and it will take a lot of the mystery out of your link planning. It has a learning curve, and it will eat all the resources you give it trying to plot your paths especially if you get the third arcsecond datum (10m resolution). You can easily fill a hundred gig of hard drive with terrain data. For 8 and 20 mile paths 1 arcsecond is sufficient. Give it a whirl.