Has anyone deployed this yet in an area with lots of FHSS 900mhz smart meters? Wondering about performance/reliability.
That's a super difficult question to answer as there's so many variables and unknowns at work. As far as really difficult/noisy shots are concerned, I think I'm the only one that's posted a link under such horrible conditions.
We haven't, but we are interested in the performance in the face of the smart meters. In our territory, ComEd is planning to deploy the Smart Meters everywhere in the next 2-5 years. Some operators who have had these meters deployed in their FSK territory simply gave up their efforts on 900MHz. It's hard to compete against the deep deep pockets of the utility companies.
I'm actually surprised the FCC lets this happen. There are plenty of other frequencies they could afford to use.
I have worked with some customers that have the FHSS meters in the area. The PMP450i most certainly worked better than the PMP100 900 MHz in the same location, but probably not as well as they could have without the FHSS radios nearby. In the case I witnessed, the performance was improved enough to work for them. We were seeing 25+ Mbps downlink at 5 miles, and it was solid and usable. Before the changeout, there were speeds all over the place (2 Mbps down to 72 Kbps) and connections going up and down.
It would be irresponsible for the to just say "go for it; it will be great", but so far - so good in the FHSS environemnts where I hae tried it.
I definitely advise testing. We are using it in a smart meter area also, with sectors. We are seeing good performance most the time, however, one variable that we have recently confirmed is we receive enough noise for adverse responses from the radios when the smart meters are 'updating firmware'. This has been confirmed by the power company, and we have now taken the initial steps to open up communications with them, much to our surprise, in agreement with them, to coordinate this better. It's only happened a few times, but mainly, we've been functioning ok.
We also have an experiment at a different location where we are using dual omni's with the 450i 900, and old antennas being recycled for SM installation. So far it's working ok, but the testing is still in progress for that.
Ben, do you have any suggestions on how to approach the utility companies that are using FHSS?
Was there a certain department you had to request?
Also, do you know what manufacturer it is your power company is using?
Ameren Illinois is the company. When it first happened, we reached out to them asking for coordination on if their meters were causing us issue. They denied any possibility to this. In the mean time, we went to work investigating ourselves, and actually tied in with a local electrician/power company guy that has a thing for RF also, and he was willing to help out. We undoubtedly singled it down to noise in the area from the newly installed meters. Then, a few days after it started, it went away, and stayed away, for the most part. Then, out of nowhere, it shows up again. This time, we get confirmation from them that, 'they have been doing some testing to determine if their firmware updates correlate to when we report the interference, and they seem to align', or something to that extent We established this communication by simply reaching out to their engineers via their contact line. The meters in use I believe are Landis Gyr GridStream RF meters.
Ben I'm up at Tuscola, IL and Ameren just deployed smart meters in my area last year. I'm using very old Canopy 900Mhz equipment and ever since that deployment I've seen times of the day (normally starting at 2PM and continuing for a random period of time) and then stop just as soon as when it starts. I've got 900Mhz equipment in three different towns around me and when it starts all the CPE starts re-registering to a greater or lessor degree and then it magically stops. I just wondered what your thoughts are on how well the PMP450 900Mhz Cambium handles this sort of interference? I did catch a 15 min cycle of the interference and included a graph. You can see the band in the middle when the meters are "bursting".
All 900 is going to be susceptible to this interference unfortunately.... The reason for this is, in my opinion, Ameren operates their devices errantly. The meters are allowed higher power because they are supposed to be using the entire band. However, and we've confirmed this with Ameren Engineers. They will 'limit' the channels in their devices to try to avoid noise, and therefor produce higher power outputs on more refined channels. Now, with that said. Even if they aren't doing that, what we have seen is the issues arise when Ameren runs Firmware updates to their devices, which apparently is fairly frequently. When this happens, ALL the radios are on at the same time instead small subsets of them for short pings of time. This is what causes your noise, is your town of 4,000 or however many just turned into 4,000 900Mhz. AP's all running at the same time, sending and receiving data. It really is sad these companies cut corners on cost by using the unlicensed band instead of something that's licensed, and hurt local companies like us. My advice is work with Ameren as much as you can. We were able to work with them and coordinate when their FW updates were to cut down on the amount of down time our customers experienced. I would say the 450 900Mhz. is definitely recommended in the simple fact of capacity and technology, if you're staying 900.
Thanks for the reply Ben, that is what I was afraid of. I like the idea of more bandwidth with the 450/450i equipment but if it can't negate the clients getting knocked off the network for 15 seconds every 30-45 seconds then it doesn't really do me any good. Everyone wants more speed but most people want reliability most of all and during a bad cycle I've seen these meter updates run from 5PM to 9AM or 16 hours. I don't know anyone who wants their Internet flakey for 16 hours. I was hoping the 450 stuff mitigated this reconnecting issue. My main reason for going with the 900Mhz equipment back in the day was for its NLOS functionality. I used the integrated dishes and overwhelming the single biggest factor in getting a functional installation was location. Its always amazed me that at one spot on a roof your signal could be -75db and two feet over it be -55db.
Do you have a contact person (email address or phone number) for someone at Ameren? My though would be if they limited their updates from 12AM to 5AM that would work great but how accommodating are they?
Any recommendations for NLOS equipment licensed or unlicensed? I'm pretty much writing off 3.65Ghz until the FCC figures out what they are ultimately going to do with this range. For me the axe in it was when they changed from Census tract auctions back to the carrier based auctions at T-Mobiles request but who knows maybe it will change again.
Thanks again for the feedback
We have been running 900MHz with Smart power meters for years. We had ubnt radios and have replaced most of them woth PMP450i 900. Both worked OK after the power company limited the data gathering to 3 times per day.
For the PMP450i radios the key is small channel size is better, we are runnig at 7 MHz.
Just to note that our distance is less than one mile, so we have strong signals.
Thanks for the feedback Dave, what channel did you go through to find the appropriate contact at your local power company?
Our 900 network is co-existing very nicely with smart meters. But it did take some coordination with our local utilities. We were able to set meetings up by directly reaching out to the utilities' engineering team or for water tanks, the GM and enlisting the support of our county commissioners who have an interest to see both sytems co-exist - they want their rural consituents to have reliable, high speed broadband.
Due to vast amount of trees, we have a large scale 900 network (60 towers sites) with both legacy 900 FSK and 900-450i. About 18 months ago our local utilities began deploying smart meters. At first, we would see bi-monthly "events" that would last up to 24 hour and ripple throughout our network. Very unpleasant causing slow speeds and dropped connections during these events.
Fortunately our local utilities are very conscious of "sharing" the spectrum and met with us. We determined that the customer meters did not cause us any problems since they report only once per day; late at night, at low powers and for a very brief time. But it was when their system pushed out new firmware, all their collectors would get very active and this created an "event". We also learned that since their system was new, these firmware and other updates were occurring frequently.
We shared with them the GPS locations of all our APs. In response, i) the utilities relocated a few of their collectors that were too close to our AP sites, ii) they re-scheduled their firmware updates to occur only 1am-5am which then takes more days for their sytem to complete and iii) they made a few other operational tweaks.
The results have been wonderful. We are operating clean. I believe that relocating a few of their collectors probably was the biggest impact. Also as their systems mature, utilities will have fewer firmware updates and even more control of operating paramaters to "quietly" udate their equipment.
Note: We do have a couple utilities using Satellite-based smart meters, so 900RF contention is a non-issue. I ask, why does a utility want to build and manage a 900MHz outside plant if they could've just used satellite?
This is great information mangophil!
Thanks for posting. An example of areas where 900 MHz can co-exist with Smart Meters and Smart Grid equipment. All is not lost when these move into an area.
Thanks mangophil, I appreciate the feedback. I'm trying to get a contact at my local power company ATM to see if I can't work something out.
This is a follow up to my inquiries as to whether the newer 450i 900Mhz equipment would be able to operate successfully in a electric/gas 900Mhz FHSS smart meter environment. After purchasing a 900Mhz 450i AP and CPE for testing I setup an AP using a 7Mhz channel band width, I was unable to get a CPE to connect to the tower at all during smart meter firmware update periods. Even outside the update periods the connection was spotty from general 900Mhz background noise. I would have to say if anyone is considering deploying 900Mhz in a Smart Meter environment be very VERY cautious before doing so.
Our electrical utility is slated to deploy AMI in 900 MHz next year, and has reached out to us so we can coordinate as much as possible. Which is nice. Their vendor is downplaying the risk of interference to/from rural broadband services. I'd like to counter that with specifics on the interactions between WISPs and utilities, both successful and not.
mangophil, Netcare, and anyone else with experience -- would you be able to share further details, if not in the thread, then via PM? I'd be interested in knowing which utilities you've dealt with, which equipment vendors the utilities used, and any information you have on the topology of their network.
If you've been successful in coexisting, I'm curious about which parts of the network actually clashed (collectors vs APs? SMs vs meters?), and what means were available to the utility to mitigate this -- like barring channels, modifying schedules, etc.
The reality is the only way you will have any chance of success is if they chose to do their collections and deployments during off peak hours like midnight to 5AM. During that time you will pretty much be down because these devices use FHSS and utilize every band in 900Mhz