We are seeing some great results from operators that have upgraded their software to the 24.2 Beta. There is a feature call Deeper Virtual Circuit in here that is proving to be a game changer for some folks. We’d love it if you can share your experiences using this software with as much detail as you care to share here.
Cambium will be discussing this software release in an upcoming webinar and discussing the benefits in a blog in the near future, but in the meantime, post here!
We ran multiple tests using a 450b SM on our PMP450m 3GHz AP’s using 40MHz/5ms and under ideal conditions we were able to reliably get 200-230mbps down and 30-60mbps up. We tested both NAT and bridge mode and it didn’t seem to make a difference. Under more challenging conditions and/or with the AP under load, we also noticed that speeds in general increased and it was much easier to hit the 100/20 mark. It’s also interesting to note that we pushed this firmware out across our entire network the other night, and during our after hours unlock window, some clients noticed the jump in speeds and asked if they could upgrade to our 200mbps plan!!
May I ask what kind of speeds you saw before upgrading, and are you running the same bandwiths as before upgrading?
And - are you running the “Max VC Depth Configuration” at the “default (256)”.
Do we need to upgrade both ap and sm’s, or is it enough to upgrade just the ap? I’am wondering because I really want to test this beta, and just upgrading the ap will make it much easier for us, as many of our customers are cabins where the sm’s isn’t online because the customers unplug the power when leaving.
Before we found it very difficult for 450b SM’s to reach speeds over 60-70mbps under any circumstances.
Yes, this mode is recommended for all channel widths and frame lengths for 450i, 450m, and 450v AP’s.
Yes. Both AP and SM’s need to be updated for the SM’s to see the speed increase. Also keep in mind that the best results will be with 30/40MHz @ 5ms frames.
Thanks for your reply! We’ll probably have to schedule a test for this beta - seems like it possibly will give us some benefits. We have found that running 30MHz @ 2,5ms frames gives us the most stable throughput at this point - being able to run 30/40MHz @ 5ms should give us a bit more “headroom”.
G’day all,
We have been running the feature for 20 days and have noted an immediate improvement across our network. We are very pleased with the results.
For your reference, the configuration we are using is as follows: PMP 450m, 5GHz band, 40MHz wide channel, 5ms frame period, a 75% downlink data setting, and the recommended default “Max VC Depth” of 256. Additionally, 95% of our SMs are PMP 450b, with the remaining being PMP 450i.
Yes we are also seeing great results here. We have 450m, 450i and 450v AP’s deployed and are using the 5Ghz, 40Mhz channel, 5ms frame. 75% download setting.
Download capacity is similar increase to all previous results posted here (seeing results from 150-220Mbps when testing the radio links). We do have some legacy 450 units in the network that we are now replacing with 450b units. Will also be doing some more testing with the 450v units carrier soon.
These are encouraging results. Would love to hear about the # of active SMs on AP and distances of SMs used for test.
We are running on 5 Ghz 450M with a mix of mostly 450br with some 450b & a few 450 sprinkled in. No connectorized. We have 17-21 active SMs on each AP. Most are LOS, few are nLOS. Reliably delivering a mix of 30, 50 & 100 Mbps with max 20 Mbps upload. We were at 2.5ms/40MHz but find that the 5ms gives better performance and customer satisfaction. AP reports 90% frame utilization with 98-100% success during peak traffic with mu-mimo from 2-12%. We now run at 80/20 split but will be testing 75/25 soon for the fun of it. VC deep at 256. We also locked the SMs to no greater than 6x/6x modulation.
Testing was performed on AP’s with as little as half a dozen SM’s to ones with over 100 SM’s, and short distances of under a mile, and distances of up to 20mi. From a testing standpoint, these things don’t matter as much as RSSI and modulation. That is what determines what throughput results you get. Here’s a small chart to help better understand what I’m talking about:
I noticed in your screen shots it looks like you’re using multiple priority channels for each SM. Why are you doing this? Unless you have a really good reason, most operators will only use the default low priority channel for every SM. Each additional channel used for each SM uses an additional virtual circuit (VC) and uses more resources across the AP… so unless you need it, it’s best not to use extra VC’s.
I think your results are on par with what I and others are seeing. It might even improve if you didn’t lock the modulation at 6x, and you only used low priority channel for every SM.
Before Deep VC, you would see an improvement in SM throughout by locking the downlink modulation of an SM that was capable of 7X or 8X down to 6X (with 30 or 40 MHz and 5ms frames). With Deep VC support in 24.2, you are better off letting the SM run at the highest possible modulation. You will see an improvement in SM throughput compared to locking it at 6X. The sector efficiency will also improve because more traffic is using higher modulations.
There’s going to be at least 1 more beta as there is a stuck session issue that they’re trying to track down and fix. If I had to guess as to a stable release, probably end of Feb.
Hi to all.
We are starting to deploy this beta initially in 2 AP 450m 5 GHz (one in an urban zone and the other in rural zone).
We are doing the preliminary documentation of the links to share the results in the next few weeks-
Hi
We are now testing version 24.2 on a 450m with 18 customers, and got an increase in download from 700Mbit/s to 844Mbit/s on AP, on the customer side we also see an increase from approx. 150-160Mbit/s to 190-210Mbit/s measured on speedtest.
But we see that all measurements on AP only show 1 on “Avrage group size” on upload ?