I don’t know how well MIMO would work through foliage, but what about a MIMO-A mode which maps data to different subcarriers on each polarity (perhaps that’s already the case, I don’t know) to counter FHSS interference? Also if the beacons, control slots, and retries could be made super robust via modulation and/or coding. Especially in the upstream direction, because the APs usually have to stick their heads above the foliage, which means they get clobbered with interference from everywhere, also if you try to obey the EIRP rules you can’t use SM side antenna gain to full advantage in the upstream direction (I think many people don’t worry about this when they’re trying to blast their way out of a forest).
Another thought regarding FHSS interference. Would it be acceptable to increase the latency to allow block coding? Or even a fast path and slow path like ADSL? I spent some time in the ADSL world back when the ANSI spec T1.413 was being developed, it is basically an OFDM system even though it uses the term DMT (discrete multitone) instead of OFDM. The FEC scheme in ADSL uses a Reed-Solomon block code with convolutional interleaving to correct impulse noise errors while introducing a minimum of additional latency. It strikes me that frequency hopping interference might look like impulse noise, although I haven’t checked into the hopping interval, it might be too long to fight with block coding. Just a thought. ADSL modems also dynamically vary the modulation (referred to as bit loading) on each subcarrier based on SNR. I don’t see that happening on an RF modem. ADSL operates over a twisted pair channel where crosstalk and attenuation can vary greatly over the channel width, due to things like bridge taps. RF channels are not usually this way, the attenuation and interference are usually about the same over the entire channel being used. From: Adam Moffett via Af Sent: Monday, October 06, 2014 1:10 PM To: [email protected] Subject: Re: [AFMUG] 900 MHz Survey - Request for Info Survey done....I'm going to address a question you didn't ask in the survey: There are two things I hate about 900mhz: First is the lower capacity, and a lot of the survey questions were pertinent to that. The second is there is a ton of interference and that makes it unreliable. I think it would be nice if a product could deliver higher capacity in 900mhz, but I also think it would be nice if we could get some rock solid IP connectivity without line of sight, even if it was at a low speed. I won't presume to tell Cambium how to do that, but maybe your next product could have an option for very small channels, or FHSS, or maybe tx and rx on different channels so I can avoid listening on a noisy channel at the tower but still transmit on it. I'd love to have more options in the toolbox to make a NLOS link keep on chugging along for telemetry, or remote desktop, or a single camera, or whatever. As some of you may already be aware, we are conducting some inquiries surrounding the 900 MHz band in order to properly address concerns in using this band, and help provide us the information needed to develop the product that you need to deliver service to your customers. The survey is just over 20 questions, and is located here: https://www.surveymonkey.com/s/XNS38W6 Please help us help you! Any information we gather will help us to make sure we’re developing the right product for your needs, and this info will not be used for any commercial or solicitation purposes. It’s optional to fill in the contact info at the end, but I encourage you to do so, in case further exploration of a few of the responses could help even more. The survey will stay open for about 2 weeks, so try to get to it soon. Let me know if you have any questions or problems accessing the survey. Thanks, Matt Mangriotis Senior Product Manager Cambium Networks 3800 Golf Road, Suite 360 Rolling Meadows, IL 60008 www.cambiumnetworks.com O: 847-439-6379 M: 630-308-9394 E: [email protected]
