Author Topic: Data Slicers and Fragments  (Read 21013 times)

nico

  • Member
  • *
  • Posts: 3
    • View Profile
    • CFSScan.com SA GRN Paging!
Data Slicers and Fragments
« on: 14 Apr 2016, 22:09:13 »
Hello,

We have started monitoring FLEX 1600 using PDW and a few Uniden UBC355XLT scanners. This is all working really well and actually hardly getting any garbled messages with the dxQuality consistent at 100%.

This brings me to some questions. What benefits would we get from using say a 4-level FSK (although we only use 1600 FLEX) and a discriminator tap?

The reason for asking the only issue we have is with fragmented messages. Now it is my understanding with the fragments is they are typically caused when the frame is filled and the next frame carries the remainder of the message. Now to combat this we have basically tried to delay in a 1-in-hold queue for PDW. Where as a message arrives it is held in queue for up to 2 seconds, after the 2 seconds the message is released from queue. If another message arrives while it is still in the queue it pushes the existing queue message out of the queue and places itself in queue. Now if a second message arrives that is +1 for the frequency the messages 'Message' part is appended to and the message is re-queued.

It would be interesting to find out thoughts on how to join up the fragments.

Finally if you believe using a data slicer and\or discriminator port will help at all can anyone let me know if they are able to ship to Australia?


Thank you for your hard work.


VidarParry

  • Top Member
  • ***
  • Posts: 126
    • View Profile
Re: Data Slicers and Fragments
« Reply #1 on: 15 Apr 2016, 05:30:00 »
Hi Nico and welcome!
I can't help you with the FLEX questions, but I do know that with POCSAG 100% quality doesn't mean 100% of all messages have been seen and processed by PDW. Mine displays 100% but I know misses a few (I estimate about 5% in my setup).

Having said that, a discriminator tap with the appropriate interface from Rene WILL improve things. And yes, Rene ships to Australia at a very reasonable price:-)

Cheers
...Steve

Discriminator

  • Administrator
  • Top Member
  • ***
  • Posts: 184
    • View Profile
Re: Data Slicers and Fragments
« Reply #2 on: 15 Apr 2016, 06:22:42 »
For FLEX1600 I would recommend the USB interface over the 4 level interface. You will only need a 4-level interface if you're monitoring 4 level modes like FLEX-6400 or ERMES. Further to that, the USB interface is powered from the USB port. For the 4 level interface you need a separate power supply. The USB interface will run on all Windows operating systems. The 4 level interface needs a real RS232 port and a legacy driver that can only be used up to Windows 7 (32 bit).

One of the great benefits of using an interface is that you don't have to tweak your PC's soundcard setting. It's plug and play. And it willl usually give you a better decoding rate than the soundcard, provided that you use a discriminator output on your receiver. If you use the earphone socket, you're better off using your soundcard.

On the fragmented messages: it's explained here under 'History'. PDW shows the messages as they are sent by the paging network. PDW is open source now, so feel free to develop a version that will combine fragmented messages  ;D

I ship to most civilized countries worldwide. This includes Australia  ;)

-Rene

nico

  • Member
  • *
  • Posts: 3
    • View Profile
    • CFSScan.com SA GRN Paging!
Re: Data Slicers and Fragments
« Reply #3 on: 15 Apr 2016, 07:26:37 »
Thank you both of you.

We have been developing the source of PDW to include pushing the decoded messages to a Message Queue and this is working great. I do understand why they are fragmented, so basically better equiptment wont actually help with fragments as this is a network "feature". That being said we have a low amount of fragments and consistantly coming up with better algorithms to join them together.

I appreciate the information about the interfaces. We are all on 64 bit systems, 1 Windows 2012 R2, 2x Windows 10 so It would seem the best bet is to go with the USB interface with the added benefit of being self powered.

Thank you again for the open source project, when we have managed to really crack the fragmented messages will push it back into an open source version.

Cheers, Nico