PMP450m LTE passes FCC CBRS approval

I see from the FCC authorizations page that QWP-30450M, the 3 GHz Medusa, has received an FCC approval (permissive change) with the following sentence: "This device supports LTE of 5, 20, and 40 MHz bandwidth modes for TDD LTE Band 48." And the power limits are quite reasonable (it's hard to get approved for anything near the theoretical +37 dBm/MHz maximum power), and the signal plots (which determine that) are exceptionally clean. We've always known Medusa to be a sector capacity leader, but the lower cost of LTE UEs makes it potentially cost-effective too, compared to using proprietary 450-family SMs. So while I haven't seen Cambium announce the product (they have talked about it coming this year), it sounds like it's right around the corner. Which could put some "alternative" purchases on hold, if the timing is right. So I have a few questions.

- When will it ship?

- Will it cost more than the proprietary Medusa?

- The FCC approval names 5, 20, and 40 MHz channels. Are 10 and 30 also supported? CBRS is based around 10 MHz channels so that would be very useful.

- Is this going to work by itself, like any other PMP, or will it emulate mobile LTE eNBs and require a Core?

- Will  LTE work under Part 90 rules too, prior to moving onto CBRS?

Thanks!

Thanks for the kickstart Fred!

I was going to post about this but you beat me to it... I will try to answer your questions here.

Q1. When will it ship?

A1. It's already shipping! 3 GHz PMP 450m has been shipping since last December (under Part 90 grant of course). 

Q2. Will it cost more than the proprietary Medusa?

A2. It is the same hardware that has now been issued a grant under Part 96 (in addition to Part 90).

Q3. The FCC approval names 5, 20, and 40 MHz channels. Are 10 and 30 also supported? CBRS is based around 10 MHz channels so that would be very useful.

A3. It lists those channel sizes, because those are the ones we obtained the original grant with, therefore continued and obtained the grant for Part 96 following a similar path. There is no issue with using a 10 MHz channel. We obtain explicit approval for the smallest and largest channel sizes (and one in the middle because it was easy to do), and anything in between can be considered a combination of those channels and is permissible. So, yes, feel free to use 10, 15 or 30 MHz channels under Part 96 as the SAS allows.

Q4. Is this going to work by itself, like any other PMP, or will it emulate mobile LTE eNBs and require a Core?

A4. This is PMP 450. The remaining hardware in the 450 platform will be approved shortly (we expect within a week or so), which will complete the proprietary system under Part 96. Again, this is Cambium Network's proprietary solution, which has a very large installed base in the United States. This means that ALL of these radios can someday be software upgraded and continue to operate under Part 96 rules! (It does not have ANYTHING at all to do with LTE or require any sort of core).

Q5. Will  LTE work under Part 90 rules too, prior to moving onto CBRS?

A5. I am not sure what this means... maybe your assumption is that 450m was some kind of LTE device.

Note that, to my knowledge (and you can confirm this for me), the only alternative protocol that is approved to use under Part 96 rules is "Cambium Technology", meaning our proprietary system (the other being, of course, E-UTRA or LTE).

All this being said, we are also in process of releasing the cnRanger PMP product line (which has its own product forum). This product will support 2.5 GHz (Band 41) first, then also be released under Part 96, hopefully by year end. cnRanger is our first fixed wireless solution utilizing LTE protocol.  Stay tuned for more information on this topic as we get closer to launch...

I'm happy to see the 450m on both Part 90 and 96. The question about LTE, though, stems from the fact that the Class III Permissive Change certificate doesn't just say Part 96, but also says "LTE" and "Band 48". And I have been told that the Medusa, but not the other 450-family devices, would be upgraded to LTE. The other devices don't have the hardware capability, but Medusa has so much DSP horsepower that it can do LTE as an SDR, even without an LTE chipset. To be sure I've seen little "toy" SDRs do it too, but not with beam forming.

So was the test lab wrong to include that LTE sentence in the Permissive Change?

You are correct that the only air interfaces currently registered with WInnForum for CBRS use are LTE and Cambium 450. But others can be added on request, and I expect "NR" (the 5G upgrade of LTE) to be added, if it isn't just treated as LTE. And some other vendors are expected to bring their own air interfaces too. But it will be easier to "coexist" (that is a loaded term on CBRS! don't get me started...) if LTE than if something else. BUT again kudos to Cambium for adding  LTE coexistence timing modes to 16.0.1, so the proprietary air interface can transmit and receive using LTE-compatible timing, and thus avoid that particular problem.

I happen to be in the middle of deciding what type of  AP/BTS to go with for a good-sized CBRS WISP project, which needs to be 25/3 (terms of the CAF grant), and I know the 450 family radios are good, but the price of the high-gain uncapped SM is just out of range. So a Cambium AP with commodity LTE SMs (CPE-CBSDs, now that the KDB has come out) would be a good combination.


@fgoldstein wrote:

I happen to be in the middle of deciding what type of  AP/BTS to go with for a good-sized CBRS WISP project, which needs to be 25/3 (terms of the CAF grant), and I know the 450 family radios are good, but the price of the high-gain uncapped SM is just out of range. So a Cambium AP with commodity LTE SMs (CPE-CBSDs, now that the KDB has come out) would be a good combination.


There's a 3GHz 450b variant that's coming soon that will be uncapped, along with having a much faster FPGA in it that can actually push 25/3 and beyond, at a much more reasonable price point. I would not spend any money on uncapped legacy SM's at this point, and bide your time till the 450b comes out.

The eventual goal of the 450m 3GHz AP is to be able to flash it with LTE code and turn it into an 8x8 RRH that can be paired with a Cambium cnRanger LTE BBU.

Here's a bit of rough math to determine the street cost of using 450m in native vs. LTE mode:

100 clients @ $175ea USD for LTE UE = $17500
100 clients @ $271ea USD (estimated) for 450b high gain = $27100 (NOTE there maybe a mid gain, slightly lower cost 450b variant released as well)

cnRanger BBU $6400 + 450m $7479 = $13,879

Complete LTE sector w/100 clients @ $31,379
Complete PMP450m sector w/100 clients @ $34,579

Thanks, Eric. If the 450b comes out at that <$300 price point soon, it would be quite competitive. If anyone needs to know about CPE-CBSD type approval (needed for a UE to have EIRP >+23 dBm/10 MHz), they can contact me off line and I'll put on my WISPA hat.

High gain LTE UEs are in that price range. Matching LTE timing really helps too. I will raise this with my team. The "plain old 450" (non-Medusa) AP is price competitive with decent-quality "cheap LTE" for the smaller sectors. I'm building a WISP in the woods. We're actually putting in ePMP on 5 GHz and using 3 GHz for the sites that doesn't reach, but that's most of them -- the ePMP skims the cream, and 450i on 900 is then available for the worst locations. (And 500 MHz TVWS in a few sites.) CBRS will be the biggest band. So CPE cost is important. I think we'll be experimenting with some existing 450 gear soon, on 3.65. And I think Cambium 450 is one of only two lines that is already type approved for both 3.65 (Part 90) and CBRS (Part 96).

It does however sound like there is an error in the Certificate of Equipment Authorization, which says LTE, unless they did get a prototype Ranger-Medusa hybrid (that sounds a little kinky :-) ) working in the lab.