I have always scratched my head over the idea that Ubiquiti and Cambium don’t have the volume for custom silicon. But if it’s true, why would fixed LTE have sufficient volume? Unless there’s a giant service provider planning to deploy it, like an AT&T, Verizon, Sprint, DISH, etc. for their fixed wireless service. Or unless there’s some huge volume in the international market, but supposedly that’s where the WiMAX market was. It would have to be high volume plus non standard for fixed wireless. Or someone would have to convince a standards based LTE chip maker to roll non standard hooks into their chip so fixed wireless could leverage the volumes of mobile wireless.
Typically the FPGA vendors like Altera have a push-the-button path to ASIC, but Cambium evidently doesn’t have the volume to justify that. Or so they claim. That approach does kind of freeze your feature set though, now the software defined part is only what you can change in the CPU firmware, not the FPGA design, even bug fixes and performance tweaks. From: Rory Conaway via Af Sent: Wednesday, November 05, 2014 7:50 AM To: [email protected] Subject: Re: [AFMUG] Telrad LTE vs 450 3.65GHZ The general path is either a chipset based CPE based on something like a handset chipset (not the best idea since it locks down a lot of things and the PHY layer really sucks for high-interference environments), FPGA or other hardware based SDR client and then move to a chipset to get the second generation cost down once the PHY layer is written the way you want it. I thought White Space vendors were going to pursue that path for example. LTE is a lot harder since there is a lot more work that has to be done on the PHY layer. LTE is the best option but proprietary LTE with various techniques, some of which are no-brainers, some of which are going to require serious out-the-box ideas, would give LTE the best chance when it moves from rural to urban. Rory From: Af [mailto:[email protected]] On Behalf Of Ken Hohhof via Af Sent: Wednesday, November 05, 2014 6:44 AM To: [email protected] Subject: Re: [AFMUG] Telrad LTE vs 450 3.65GHZ Exactly, can they do proprietary CPE? From: Jason McKemie via Af Sent: Wednesday, November 05, 2014 12:18 AM To: [email protected] Subject: Re: [AFMUG] Telrad LTE vs 450 3.65GHZ The latency is in the standard, so they would have to deviate from it. Being SDR, I suppose they theoretically could, I just haven't seen it happen yet. Presumably the CPE are built around a standards based chipset, so it would probably mean an increase in price for those as well. On Wed, Nov 5, 2014 at 12:08 AM, Rory Conaway via Af <[email protected]> wrote: Why do you say the latency would be significantly higher? That’s manufacturer dependent since they aren’t locked to the full cellular PHY. Rory From: Af [mailto:[email protected]] On Behalf Of Jason McKemie via Af Sent: Tuesday, November 04, 2014 8:55 PM To: [email protected] Subject: Re: [AFMUG] Telrad LTE vs 450 3.65GHZ The 450 is going to smoke the Telrad from a latency standpoint. This is a big downside to this gear IMO. If someone would optimize LTE for use in a fixed environment you could possibly get better latency, but I'm not seeing that happening. On Tue, Nov 4, 2014 at 9:40 AM, Matt via Af <[email protected]> wrote: Is anyone looking at both these options? Pros and cons of one vs other? Throughput, Sync, Interference, spectrum efficiency, cost etc? What I really wander is can you bond two non adjacent 10MHZ channels with Telrad to make a 20MHZ channel? I know you cannot with 450 and it might become very useful down the road. Right now with 450 and a perfect connection on a 20MHZ channel we can do around 80Mbps downstream per sector. With a 10MHZ channel and not so perfect connections that is really getting cut back. Also, right now we can do ABAB quite easily with 450 gear. With LTE can we do AAAA where spectrum is scarce?
