[aprssig] 9600 Baud Packet Network?

[Ev Tupis]

250 bytes per packet at undefined data rate? This seems like a low throughput IoT device.

      From: Andrew Rich via aprssig <aprssig at tapr.org>
 To: Ross Whenmouth <ross at topwire.co.nz>; TAPR APRS Mailing List <aprssig at tapr.org> 
 Sent: Monday, June 13, 2016 6:55 AM
 Subject: Re: [aprssig] 9600 Baud Packet Network?
   
http://www.lprs.co.uk/assets/media/downloads/eRIC_USB_Dongle.pdf

-----------------------------Sent from my iPhoneAndrew Richvk4tec at tech-software.netwww.tech-software.net0419 738 223
On 13 Jun 2016, at 19:20, Ross Whenmouth via aprssig <aprssig at tapr.org> wrote:


 
Hi,
 
 Robert Bruninga wrote;
 
19k2 is there – the modules are $40
 Bob, could you please post a link to these modules?
 
 
 Re: faster than 9k6 without modifying radios
 At the penalty of increased sensitivity to noise, modulations more complex than K9NG/G3RUH can be fed through the linear audio channel of a pair of 9k6-ready ham transceivers. A 9k6 audio channel provides 5 kHz bandwidth, more than that of a standard telephone land-line - a quick back of the envelope with Shanon's Law indicates that we should be able to push bits much faster than 9k6 https://en.wikipedia.org/wiki/Shannon%E2%80%93Hartley_theorem
 Think DSP "TNC" software running in your PC, smartphone, Raspberry pi, or in a processor like a PIC32 or STM32, with automatic channel equalisation, GMSK/QPSK/n-QAM and Forward Error Correction. Note that about twenty years ago you could just go to the computer store and buy a consumer modem that would push 33k6 in 3.4 kHz BW down a (then) 40 year old telephone line...
 
 
 Re: Utilising smartphones, tablets, laptops, etc as user interfaces;
 Why not use Bluetooth to link your device to the TNC? - it is pretty much well universally supported by mobile devices and doesn't chew through as much battery power as WiFi
 eg http://www.mobilinkd.com/tnc2/
 
 
 Re: Transferring (larger) files;
 Send larger packets so there is less time consumed per unit payload on TXdelay, ACK, etc. eg a 1500 byte/12kbit packet takes less than 1.5 seconds to send at 9k6, and only incurs one TXdelay, where as, if you limit yourself to 256 byte AX.25 packets, you need to send 6 of them to move 1500 bytes... 
 Forward Error Correction seems like a good way to reduce packet re-transmissions (corruption of a single bit in standard AX.25 results in rejection of the packet, requiring a resend) http://www.stensat.org/docs/FX-25_01_06.pdf
 New data transfer protocols eg using a "Fountain Code" would allow most of the ACK/REJ packets to be eliminated and would be very efficient for simultaneously distributing a file to multiple recipients https://docs.switzernet.com/people/emin-gabrielyan/060112-capillary-references/ref/MacKay05.pdf
 
 
 73 ZL2WRW
 Ross Whenmouth
  
_______________________________________________
aprssig mailing list
aprssig at tapr.org
http://www.tapr.org/mailman/listinfo/aprssig


_______________________________________________
aprssig mailing list
aprssig at tapr.org
http://www.tapr.org/mailman/listinfo/aprssig


  
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.tapr.org/pipermail/aprssig_lists.tapr.org/attachments/20160613/3ab86baf/attachment.html>