Sriam Any work being done to lower the latency on the Epmp system,?
Gino A. Villarini President Aeronet Wireless Broadband Corp. www.aeronetpr.com @aeronetpr From: "[email protected]<mailto:[email protected]>" <[email protected]<mailto:[email protected]>> Reply-To: "[email protected]<mailto:[email protected]>" <[email protected]<mailto:[email protected]>> Date: Friday, September 26, 2014 at 10:10 AM To: "[email protected]<mailto:[email protected]>" <[email protected]<mailto:[email protected]>> Subject: Re: [AFMUG] Cambium Newbie Question Hi Jeremy, The doc uses an example terrain (much less the one west of Rolling Meadows ☺) and no way is it representative of all possible terrains out there. As a general rule of thumb, you don’t want two sectors on nearby towers facing each other operating on the same frequency. But if there is enough terrain between them providing plenty of isolation (attenuation), then it’s certainly worth a try. The four channel reuse on multiple towers is recommended, again as a rule of thumb, but you can certainly populate your network of POPs with two channels if the terrain allows it. Bumpy terrain, plenty of tree covers etc. will certainly help your cause in reusing two channels across towers. If you have four channels available, by all means throw them on your four sector tower. This mitigates the need for GPS sync. However, if the four channels are adjacent channels without much guard band, then GPS sync is still recommended. In an unsynchronized system using ABCD adjacent channels, you may need twice the amount of guard band as the channel size when using adjacent channels (read alternate adjacent channel since you need the guard band). With a GPS synchronized system, on ePMP, 5MHz guard band is recommended between adjacent channels. Regarding the “Front Sector” and “Back Sector” settings recommended in the doc (and User Guide), you will have to follow that. That is part of the magic sauce in ePMP that make GPS sync work on this platform. Please give link planner<https://support.cambiumnetworks.com/files/linkplanner> a try if you haven’t already. Even though it doesn’t predict or help with deploying a synchronized network, it does predict performance of your deployment (each POP) taking the terrain, antenna azimuth/elevation and many other key factors into account. And lastly, like Adam pointed out, you can switch the GPS source on the GUI. However, it doesn’t auto switch sources like PMP 450. This is something we still need to implement on ePMP. Thanks, Sriram From: Af [mailto:[email protected]] On Behalf Of Adam Moffett via Af Sent: Friday, September 26, 2014 8:41 AM To: [email protected]<mailto:[email protected]> Subject: Re: [AFMUG] Cambium Newbie Question I doubt anybody can make that perfect grid shown in the white papers. You just get as close as is reasonable. When ePMP first came out there was a document describing the two channel layout vs four channels and it spelled out that there are specific cases where you are guaranteed to get self interference in a two channel system. You can switch sync sources in the web GUI. According to the guys at the ePMP Tour in Albany, the "Front" and "back" sector designations are strictly informational. Like the "sectorID" in Canopy. The setting has no actual technical affect, so if you don't need it as a mnemonic device you can ignore it. Sriram, that brings up my next question. The channel planning model for reuse is great for Idealtown, located on a flat plain where one can permit and build POPs on a tidy rectilinear grid. (This may be just west of Rolling Meadows). I wonder about the utility of channel reuse in say, Realtown, where the topology is quite bumpy, forestation is patchy, and the operator takes what he can get in the way of locations for POPs. This is pretty much my situation, and probably plenty of other folks’ too. I’m trying to think of a broad rule set for channel planning in those conditions. For instance, I’m planning to expand into an area with existing structures (silos). In the attached image I’ve modeled coverage in Radio Mobile with an RSSI of –66dBm or better at the SM, assuming an ePMP AP/90°sector at power limit for max modulation and Force (25dBi) SMs (antenna pattern is just an omni for planning purposes). Max cell radius is 6km. This is over actual topology, of course, and using a publicly available ground cover (clutter) database, so it should be a pretty good prediction of which POP gets best signal to a given location. Each POP has its own color, with some reuse where it wouldn’t be confusing. (This is RM’s “combined cartesian” coverage, so there are plenty of locations where more than one POP can provide better than -66, but the POP with the strongest SS gets to put its color on the pixel.) Some of the POPs won’t want a full 4-sector deployment, but many, probably most, will. Am I better off, generally speaking, with the recommended 4-channel model, with two of the four channels on each POP (and the other two channels on the adjacent POP) than I am with the two channel model? And if so, would I just maintain the same azimuths for all of the POPs—e.g. channel A always at 0° and 180° and C at 90° and 270 ° on POPs 1,3, 5…, then channel B always at 0° and 180° and D at 90° and 270 ° on POPs 2,4,6…? Then maybe we could just leave out unnecessary AP quadrants on POPs where they weren’t going to do any good. Is there any reason to try the ABAB reuse model if four channels are available? Does the necessity of setting Frequency Reuse “Front” and “Back” go away in the ABCD model—and can anyone explain just what that’s doing? Whew. Oh, yeah—can you just software switch between the GPS timing signal on the (internal patch or) local GPS port and the signal on the Cat5/6 from a CMM, if you want that kind of redundancy? From: Af [mailto:[email protected]] On Behalf Of Sriram Chaturvedi via Af Sent: Thursday, September 25, 2014 5:45 PM To: That One Guy via Af Subject: Re: [AFMUG] Cambium Newbie Question Hi, Yes, the GPS chip comes with an internal patch antenna. The internal patch antenna is automatically disabled once you connect the external GPS antenna (and auto enables when you disconnect the external antenna). If you think the radio itself doesn't have clear LOS to the sky, then you can use the external antenna and place it elsewhere on the installation to get better LOS to the sky. There are a couple of documents on our support site (https://support.cambiumnetworks.com/files/epmp ) you can read through that will help answer questions about ABAB deployment using ePMP. Thanks, Sriram ________________________________ From: Af <[email protected]<mailto:[email protected]>> on behalf of That One Guy via Af <[email protected]<mailto:[email protected]>> Sent: Thursday, September 25, 2014 4:36 PM To: [email protected]<mailto:[email protected]> Subject: Re: [AFMUG] Cambium Newbie Question the APs come with an antenna for GPS, but its never been clear to me whether there is also an internal patch On Thu, Sep 25, 2014 at 4:31 PM, Jeremy Grip via Af <[email protected]<mailto:[email protected]>> wrote: So would you be able to switch over to the onboard sync remotely? Do you need an antenna for each AP for using it? Do you think it’s as precise as using an CMM4 (or SyncPipe Deluxe w/Gig Injector) if not as robust? If all POPs are sync’d with same Up/Dn ratio and max cell distance and they’re talking to the same birds, is it pretty much the same? From: Af [mailto:af-bounces+grip<mailto:af-bounces%2Bgrip>[email protected]<mailto:[email protected]>] On Behalf Of Adam Moffett via Af Sent: Thursday, September 25, 2014 2:55 PM To: [email protected]<mailto:[email protected]> Subject: Re: [AFMUG] Cambium Newbie Question .....also the PMP100 SyncInjector from Packetflux ought to work with ePMP. You might want the gigE version, but in the real world with a mix of subscribers at different MCS levels I'm not sure how likely you are to exceed 100x100. The CMM4 is a much more rugged beast. It is expensive, but you are not likely to go back and wish you'd bought the cheap one. My plan is to hook up the internal GPS and have it available, but also to provide sync over power. Once you are using GPS sync to re-use channels it becomes critical that it's always working, so better to have two timing sources available IMO. They have built in GPS if youre on a budget, not sure why alot of people are so die hard against using it On Thu, Sep 25, 2014 at 12:56 PM, Jeremy Grip via Af <[email protected]<mailto:[email protected]>> wrote: I’m looking at ePMP w/channel reuse from a cost-comparison standpoint. Trying to figure out how much I need to spend on GPS synch for a 4 AP/ 2 channel cluster. Does it need to be a CMM4? I will want to be synching multiple POPs… Jeremy Grip North Branch Networks,LLC -- All parts should go together without forcing. You must remember that the parts you are reassembling were disassembled by you. Therefore, if you can't get them together again, there must be a reason. By all means, do not use a hammer. -- IBM maintenance manual, 1925 -- All parts should go together without forcing. You must remember that the parts you are reassembling were disassembled by you. Therefore, if you can't get them together again, there must be a reason. By all means, do not use a hammer. -- IBM maintenance manual, 1925
