Comments on: PalmSource Licenses Palm OS to PerComm

I saw HuneTec's presentation on ReFlex and Palm OS back at the May 2003 dev conference, and their method of integration was very, very complex. Of course, we've not seen their device show up yet, even after going through a spec change that boosted the screen to 320x320 and gave it an ARM processor.

ReFlex networks tend to charge high per-byte data rates, although WebLink has been building out pretty significantly. I think this may have some potential in embedded telemetry and the "blackberry" market, especially in areas where cell phones can't be used or have poor signals.

--
Ben Combee, CodeWarrior for Palm OS technical lead
Programming help at www.palmoswerks.com

Anyway, I'ld like to see how these ReFlex devices fair in the market when compared to other GSM/GPRS or CDMA2000 enabled devices...

Visit http://goodthatway.com/
-better living through better technology.

My "informal, non-scientific, not at all based in fact" market analysis predicts that PerComm would mae two models - a high end, always on communication device that uses CDMA/GSM (per the carrier), and ReFLEX technology, and a lowend "Zire 21" type clamshell model that is an intelligent 2way pager - OK, for IM, and limited email. Both have a target audience. Interesting times.

~ "Don't be too proud of this technological terror you've constructed." - DV ~

FLEX dosen't have signifigantly better penetration - It's just a lot easier/cheaper to deploy high-powered signal towers for pagers, and the FLEX protocol supports specific message identifiers - So, if the provider configures for it, the system will transparently resend the message, and if your pager has already recieved the message, it will ignore it.

If anybody's interested, POCSAG is an antique paging protocol that was buried at a crossroad with a stake through it's heart by FLEX in the mid 80's.

Personally, with RIM's market share eroding while their machines become more PDA-esque, I don't see a lot of room in the market for this company's offering, particularly with the Sidekick standing in the wings.

I'll be interested in seeing how they implemnt, though - Most of the PDA-based pagers I've seen so far have relied heavily on the PDA CPU - Keeping it from sleeping most of the time, and consequently slaughtering the battery life of the PDA in the process. (I had a FLEX Alphanum pager card for my PPPro that would actually eat a pair of AAA batteries in three days.)

If this was an announcement from Motorola(The inventors of the Re/FLEX protocol, IIRC), I'd be more interested, but this really looks like another Fossil-style abort-before-market thing.

If you don't think 45dB SNR gain means something for building penetration, then you're missing an important point. The same thing applies for ReFLEX on the reverse channel. In a typical scenario, each ReFLEX site will have a local 2path diversity receiver, and as many as 4-6 sites will all be able to pick up the mobile device's signal on a fragment by fragment basis. That's say 12 branch diversity.

And by the way, it's not thoroughly inexpensive to operate networks like this.

Looks like someone just got owned by engineer zang.

zang: "Bring it on Mo Fo! You want a piece of me!"

orb2069: "No good Sir! I'm sorry I opened my mouth. Please Beyotch Slap me again!"

Man, why don't you get away from the PC for a few months and lose all that weight? And wash that damned greasy body of yours too!

What do you think your mommy does every night? Ask her.
You're such a predictible little Beyotch, Michelle. Try to lighten up a little before you get hurt. Or someone straps you up and injects you full of Haldol for your illness.

If you like, I could give you one of my old Palms so you at least have a PDA to use. Let me know if you want one.

look for me, I will stand out. I'm the bigest asshole in town!

What part of the city is your newsstand in?

Apparently I was insufficiently clear - What I ment was, the frequency they use (Which IIRC is in the same band as the older POCSAG transmitters) dosen't have any better penetration than the previous systems.

You're right, of course, that the heavier deployment (...It's just a lot easier/cheaper to deploy high-powered signal towers...) and redundancy of FLEX systems (...FLEX protocol supports specific message identifiers...) does mean that more messages get through

But on a same-transmission-frequency and power basis, any given signal (What I would mean by 'Signal penetration') is not more likely to get through.

Sorry I wasen't clearer.

I'll allow that frequency is a factor in VHF/UHF propagation. But the marginal change in going from, say 840MHz for cellular to 940MHz for ReFLEX does not amount to a significant difference given the other effects in UHF props. There is more likely to be a significant difference in contrasting 900MHz and 1800MHz operation for NPCS and broadband PCS, with the advantage going to NPCS. This has more to do with path loss over open terrain than it does with propagation through building walls or internal scattering or absorption within the building structure and contents.

I return to my original points: the most significant differences between NPCS and digital cellular or broadband PCS vis a vis building penetration are as follows:
1. NPCS simulcasts at higher power than alternatives, yielding more aggregate power per bit.
2. NPCS bit times are longer, yielding more energy per bit than alternatives, energy being power x time.
3. NPCS bandwidths are lower, yielding less noise energy per bit, since noise power is proportional to noise bandwidth.
4. NPCS has a typical channel diversity of about 5-6 branches on the forward channel and twice that on the reverse channel.
The combination of the first 3 factors goes to the "contrast ratio" or expected Eb/No (signal energy per bit versus noise energy per bit), which like I said before may be 45dB or more in favor of NPCS. The fourth factor is harder to explain, but makes the NPCS channel look more like a non-fading wireline link as opposed to the Rayleigh fading link typical of digital cellular or broadband PCS.
This piece of the puzzle implies that NPCS has to give up far less of its total bandwidth to channel error coding and packet retransmissions than alternatives. It is not unusual for broadband channels to achieve less than 25% of user bandwidth to total bandwidth after rate 1/4 convolutional encoding, and block error check codes, and source and destination addressing, and retries, and ...

Problem is, people think paging is old technology. That's not true. The paging sector spends more of its money on networks than on marketing. I suppose that's boring.... ReFLEX is actually a packet data system optimized for device battery consumption.

Oh, did I mention how little battery consumption a ReFLEX mobile has relative to the alternatives?