I might have heard, on this list or on the forums, that Cambium is developing an airtime-fairness feature to combat this very issue. Did I dream this or is this a confirmed feature? If Cambium is listening, I'd love to help in the beta if/when it becomes available.
On Fri, Jan 20, 2017 at 1:20 PM, Eric Muehleisen <[email protected]> wrote: > Sean, > > We are about half 5.4 and half 5.8. You need you to understand that in > this particular case we have about 20+ AP's crammed in a town about 2.5 > square miles along with two other wireless internet providers. The > frequency coordination and high noise floor is the primary reason for 1x > and 2x connections. On top of that we have around 3k subscribers and 30 > roaming combination techs installing CPE all across western Kansas. You may > have complete oversight of where and when you install customers... I simply > do not have that luxury. I preach the importance of quality installs > constantly, but between their supervisors, marketing teams and pressure > from customers, they install what they feel comfortable. The install tech > has full discretion, for better or for worse. In the face of all this, 2x > subs sometimes get installed. I understand fully that we made our own bed > here. It is what it is. This is why I say "in a perfect world". > > As I said before, we truck-rolled our entire customer base for those who > have poor/marginal signals. We made significant improvements but only > gained very little in capacity for our efforts. In hindsight, it cost us > more than we gained. Hence the reason I ask if 450i has any significant > improvements over vanilla 450. > > On Fri, Jan 20, 2017 at 12:48 PM, Sean Heskett <[email protected]> wrote: > >> eric are any of your APs using the 5.4 band? >> >> you could put up some additional APs in the 5.4 band and migrate the >> close-in clients to them to provide more capacity. >> >> -sean >> >> >> On Fri, Jan 20, 2017 at 10:52 AM, Eric Muehleisen <[email protected]> >> wrote: >> >>> In a perfect world. >>> >>> On Fri, Jan 20, 2017 at 11:32 AM, Kurt Fankhauser < >>> [email protected]> wrote: >>> >>>> You shouldn't be putting any clients on a 450AP with less than 4x >>>> connections, when they start pulling a lot of traffic they significantly >>>> choke the AP. Whoever engineered those links needs slapped. I know it >>>> probably looked like a good idea at the time to add a new customer but a >>>> couple low signal clients really affect total AP capacity. You could >>>> probably pull all the 1x and 2x clients and replace them with 10 times more >>>> customers running at 8x mnodulatio n and maintain the same utilization >>>> rates on that AP. >>>> >>>> On Thu, Jan 19, 2017 at 9:45 PM, Craig Schmaderer < >>>> [email protected]> wrote: >>>> >>>>> Remember that even a couple of sms runing at 2x or 4x will kill your >>>>> peak performance if they are the ones that are usually streaming. We are >>>>> very careful on what aps we put low signals on and we control what plans >>>>> we >>>>> offer as well based on signal strength. I just can not have a 2x customer >>>>> want a 10mb plan. I do think that a 450i will definitely help with uplink >>>>> interference like others have said. But i think your best bet is to drop >>>>> the cash and put some 450m up on that tower. Sounds like that is a cash >>>>> cow tower like my main tower is. >>>>> >>>>> >>>>> ------------------------------ >>>>> *From:* Af <[email protected]> on behalf of Jon Langeler < >>>>> [email protected]> >>>>> *Sent:* Thursday, January 19, 2017 4:18:22 PM >>>>> *To:* [email protected] >>>>> *Subject:* Re: [AFMUG] 450 vs 450i >>>>> >>>>> Put a Mikrotik behind an SM and speed test to the internet >>>>> >>>>> Jon Langeler >>>>> Michwave Technologies, Inc. >>>>> >>>>> >>>>> > On Jan 19, 2017, at 1:08 PM, Brian Sullivan < >>>>> [email protected]> wrote: >>>>> > >>>>> > I would check out page 9-51 in the PMP 450x Configuration and User >>>>> Guide 15.0.2 >>>>> > >>>>> >> On 1/19/2017 11:37 AM, Tushar Patel wrote: >>>>> >> So how can we tell when we are really saturating the connection? >>>>> > >>>>> >>>> >>>> >>> >> >
