Before I throw out questions, I would like to state that I am looking for hardware solutions and suggestions, I would like to steer away from business practices and/or why we can’t compete. We all know whats happening, its just how do we continue to stay competetive, meet our customers needs, and sustain our business model.
Many of us are in the same boat, customers want more for the same amount of money. The same group of customers used 4x the bandwidth in 2010 than in 2009. My competitors can CONSISTENTLY give 10+mb up and down. We are stuck at 3-4mb very inconsitently. Our backhaul infrastructure is just fine. We just cant deliver it to our customers. I feel we have to move to a proprietary solution. I plan to retire our current PtMP networks ( RouterOS, XR2s, and demarc/ns2/rb411 cpes), still support them but with no additions. And build on top of our existing network with something else. Right now I am thinking Canopy in the 5.7 range. Then allowing our customers to upgrade to the new package offerings and upgrade their CPE’s.
I have a full business plan in place to move ahead with Canopy. But then this morning in bed I thought maybe Mikrotik can offer me something if I am willing to move to a proprietary solution. Will nstream w/polling and/or nv2 be something I could look at? Or is there something else? I am sold on canopy, as thats what our competitors use to walk all over us. I have always had the upper hand until recently.
How would you build your PtMP network if you had to do it all over again RIGHT NOW, Mikrotik or Not (as essentially thats what will be happening)?
I am not sure about the rest of the world but in the US unless you are providing service in very very rural areas I think we are all in the same boat.
I have been trying to make the same decisions as you the last few months and I had some high hopes for NV2 but from my testing it is far from stable enough. When it works the results can be pretty good but it suffers from stability issues that range from registered clients just “locking up” and stop passing data to clients that won’t stay registered. It seems to me that it is more susceptible to noise than the standard wireless package. I realize that it is still in “testing” phase but they have added the wireless package to a stable release so it should IMO be usable on a production network. I have tried Nstreme over the years and it never has worked as advertised for PMP, it works well for PTP but every time I have tried it with PMP there have always been latency issues and throughput wasn’t that great. I am not sure at this point in time that MT developers realize how dire the need is for an efficient/stable GPS sync system and even to the point where the semi mock people for asking for it. If they aren’t even working on the system yet it is going to be a year + before they have anything that is semi useable.
Our entire PMP network is MT and due to problems I mentioned above we have been unable to use any of the “proprietary” wireless features of MT. So we are forced to use 802.11 with small channel sizes due to the terrible noise floor in our area. I have stuck completely with MT for AP’s and Clients with the hope that they were going to come out with a proprietary wireless package that would solve the noise/capacity issues, but I just don’t see it happening. I have come to realize that there are more cost effective solutions out there for 802.11 that are just as stable as MT that don’t require assembly and special configuration of every CPE. I have several new tower buildouts this year and it is going to be hard for me to justify building them out with MT gear when there is less expensive gear out there from a company who already has working TDMA and is actively working on a GPS sync system. Not saying that there won’t be an MT router a the bottom of the tower but there may not be 3+ MT AP’s and 100% MT clients fed off these towers. We may go from spending $10000+ on MT for a tower buildout “including clients” to $200 for a router at the bottom.
As far as canopy goes I never have cared for their tiered CPE pricing model and the throughput of the current canopy system isn’t going to cut it in another few years anyway so I am not sure that would be a good system to buy into at this point in time either.
can i ask what happens if congestion or interference effects 5.7 how does canopy handle this, can frequency be moved?
I would not spend any more money than absolutely necessary on equipment used in an unlicensed band?
My suggestion for now is to test Mikrotik at 5.8 and only upgrade sites where competition has a very good signal, or where you are sure if you don’t upgrade customers will be lost.
Long term we may have to consider using licensed band equipment for our networks and have local regulatory bodies to police and sort out any potential interference issues?
Thanks for the input. Anybody have a suggestion on pre-assembled 5ghz cpes out there at a good price and with good installation and maintenance features?
I will continue to repeat a comment i made before,
For those of us who are trying to run a WISP business, reliability is of primary concern, performance is secondary, and I don’t want to waste my time on equipment which spec stating great performance but after a lot of time spent trying to achieve this performance, discover possible incompatibility issues?
At present we purchase antenna “A” with radio card “B” and Mikrotik RB board (?) +ROS Version (?) and config (?) and hope it will work as expected?
It would be great if Mikrotik would test an actual complete kit, for example:
A PTP link of “X” kilometres at “Y” speed used for this test were item A (list), item B (list), item C (list) with RB board (?) + ROS version (?) and config (?),
To summarise Mikrotik need to recommend or offer “a complete integrated kit solutions”,
Other manufacturers are offering complete kit solutions and Mikrotik need to offer similar,
Could the performance or reliability issues be attributed to the current random selection and layout of hardware currently used?
Remember customers in a very competitive marketplace have zero tolerance for network issues caused by us trying to improve our network?
I concurr - I have been asking for this for years! I’m at my max on how far I can go with my current MT network - my AP’s suck and I can only make v3.30 work with any kind of performance level. With more customers coming on - I have to scale back the available bandwidth per AP and add more APs = noise. I don’t have the finances to completely replace my systems.
MT needs to really look at who are using their system.
Is it just a routing platform or is it really a WISP platform that has amazing routing capabilities!
Who buys more equipment? Enterprise level routed networks or WISPs that need advanced routing for reasonable prices? Where is the least competition - enterprise routing or WISP routed networks?
I love this product when it works but I’m being pushed away to less sophisticated equipment or at least being forced to look for it!
I know you guys can make this happen - you have before!
MT needs to really look at who are using their system.
Is it just a routing platform or is it really a WISP platform that has amazing routing capabilities!
Mikrotik need to setup their own specialist WISP Research & Development department, assemble and test working solutions for WISPS, first we need more optimized working equipment and second required is training on how to better our networks using Mikrotik?
Who buys more equipment? Enterprise level routed networks or WISPs that need advanced routing for reasonable prices? Where is the least competition - enterprise routing or WISP routed networks?
Remember Mikrotik, WISP customers purchase in volume and if given good products and support will remain longterm customers,
U can do combination MT for station and some good sector like this http://www.wifi-stock.com/details/airmax19sec.html
for client try UBnt NanoStation M5
U can’t use Airmax Nstreeme or NV2 but N is fully compatible
u can get good performances with that equipment.
I took over a failing WISP last year. It was all motorola canopy. The issue was the config of the network and lack of knowledge. Be sure to read up on canopy!!! At this time i use a combo of MT, canopy and trango equipment.
Main links are MT 5.7 with secondary main links canopy on 5.7 and customers use 2.4 canopy or 900 sm’s. At this time i only offer 3 down 1 up. canopy 900 doesn’t allow anymore unless its advantage equipment and this system i redesigned is not advantage. ( still using firmware 8.2.7, was 7.0.7 before ) Are plans for this company that is also a cable company is going to go fiber to the home. TV, Internet and phone on one line. Have things to take care of first like new office/head end building then start the build out.
I have 13 wireless towers in my county and I have trouble tickets open on most of the towers. Looking over them I find the ack time between 350-450. I knock a few people off here and there and they return with better ack and the bandwidth speed to the customers return… for a while…
I’ve been wondering for a while what others are doing… I thought I would visit the forums to see if there were other spectrum available yet. My back hauls at 5.8 work great but the client’s on 2.4Mhz 10Mhz spacing just is getting too crowded and I’m in a very rural area.
So reading though this topic has really sparked my interest in the subject. I’m very hesitant to do anything at the moment but I do know I need to move forward and expand in order to survive. I upgraded my first router from 4.5 to the 4.16 yesterday and after upgrading it had to be rebooted. Luckily it was here at the office. I can’t upgrade the network to the new version if each unit has to be rebooted manually. Talking about a support nightmare. Not counting the solar powered towers we have that we can’t reach this time of year (due to snow, ice, wind and otherwise bad weather).
It seems to me that 5ghz range for back hauls is just fine but really don’t feel comfortable trying to put clients in that spectrum due to line of sight issues. I was hoping to see something come from the T.V. channel spectrum that would open up to us “unlicensed” operators. I suspect many of you have moved from providing dial-up to wireless in order to survive and stay in business. Things are tight now and I’d like to know how others are handling this very issue.
I feel that TV Whites paces is going to be the solution for this in the US because:
Lower frequency and the benefits that entails
Low usage of mobile type devices “in the home”. The biggest problem with 2.4GHz is that we have to contend with SOHO routers spewing out packets all across the band.
Unlike 3.65GHz, reasonable power limits for the propagation characteristics.
It’s unfortunate that 3.65GHz turned out to be useless for PMP applications (at least in a rural setting where I operate), and unfortunately in most areas 900MHz is even more oversaturated than 2.4GHz so that’s not an option.
I feel that TV White spaces will allow for WISP carriers to operate on frequencies that have good propagation, good power limits, and we won’t have to deal with dozens or hundreds of access points for frequency contention.
To answer your question, if I had to rebuild the PMP network from scratch, right now, I would use a product with robust TDMA+MIMO, or a Beamforming product. I would continue to use Mikrotik for routing and PPPoE termination. Please PM me if you would like specific manufacturers names, as I believe it’s against forum policies to post that out in an open thread.
Anything 802.11 now qualifies for all 50MHz, including 802.11n. There are several products available capable of operating on 3.65GHz capable of 50 to 150Mbps throughput - It’s perfectly usable for backhaul.
The problem as I see it is that it’s not usable for PMP because of the low EIRP limits, thus it doesn’t solve the bigger problem. In my opinion 5GHz is still perfectly fine for backhaul (at least in the rural areas where I operate), so 3.65GHz solves a problem that doesn’t exist for me. Perhaps it’s much more suited to suburban environments with smaller cell sizes than I operate.
I do find a certain TDMA/MIMO 3.65GHz product intriguing, but the output power specs still look weak in terms of PMP. It doesn’t appear that it will get long range, when it’s finally released.
Plus, on top of all of that, you can’t operate a station without FCC registration, which means from time of customer order until truck roll for install is minimum 10 days, and probably more like 14 to 20 days. People simply don’t want to wait two weeks between order and install.