Batboard

I just received the Motorola cut sheets on their new XTL5000 line with LCD heads. In the specs under "Code Book Structure," it indicates "APCO Project 25 (IMBE): No code book."

This means that unlike digital cell phones that use code books in conjunction with their VOCODERs to recognize and then filter out "known noises," such as automobile air conditioners, various other HVAC sounds, wind noise, road noises, etc. there is no code book for P25 Phase I radios.

Question: Does anyone know if P25 Phase II plans to implement a code book, and if so, is there any reason why they couldn't leverage off the latest cellular code book(s)?

Thanks, in advance, for any insights on this.

Codebooks were used for VSELP. Compared to IMBE, codebooks are memory intensive, slower, & just don't sound as good. I don't expect any future digital standard to be codebook based.

Todd

http://www.mmtelcom.com/webdex/hard_voice_comp.html has a really good explanation of why linear prediction vocoders use codebooks but MBE vocoders don't. It's at the top of the page under "The Technology- MBE (Multi Band Excitation)".

And yet ALL digital cell phones, which typically employ leading edge vocoders, use code books. Cell phone outfits are going for 1) maximum talk paths per channel bandwidth (e.g., 4 talk paths per 25 kHz is typical), and 2) optimum voice quality.

My cell phone sounds a lot "better" than my XTS5000, and it uses one-half the bandwidth as the FDMA modulated IMBE P25 subscriber set.

So I wonder if Phase II is considering code books?

Keep in mind that when APCO et al selected IMBE to be the vocoder for Project 25 during the early nineties or thereabouts, IMBE really was cutting edge. None of the fancy codecs found in cellular phones nowadays existed then. GSM-EFR, from memory, didn't become common until at least 1998.

AMBE doesn't incorporate a codebook. MBE and LP codecs are intrinsically different, codebooks are a solution used by LPC to answer a problem that MBE handles completely differently.

While codebook based vocoders are more computationally intensive than MBE, extremely fast microprocessors are dirt cheap. For $4 you can get a processor that will run at 80 MIPS while drawing only a few mA, more than fast enough for such work.

For what it's worth, TETRA uses an ACELP (Algebraic Codebook Excitation Linear Prediction) vocoder, and IMO it sounds absolutely BRILLIANT compared to P25/IMBE. I have about a dozen speech samples of each floating about here somewhere, which I will try to relocate over the next couple of days.

Oh goodie - a VOCODER debate to further muddy the water.

P25 Pase-II was supposed to be a logical migration from C4FM 12.5 KHz channels to CQPSK 6.25 KHz channels. Then along came Ericsson & other TDMA interests and now we have EADS [airbus] bringing their considerable muscle to push for TETRA.

At best Pase-II will support multiple formats [which is the same as NO standard] - at worst this will drag the process on and on and on ....

Meanwhile Motorola is selling Phase 1 P25 trunking as fast as they can get it out the door - knowing that these public safety customers - unlike the computer marketplace - will not [and can not] disgard what they have invested in just because something new comes along.

By putting in the foundations of county-wide, regional, and state-wide Phase-1 systems - they have locked the majority of the market for the next decade or longer.

Is AMBE still the favorite for Phase II? If so, I guess that means they will not use a code book?

EADS is pushing TETRA? Surely you mean TETRAPOL?

mr.syntrx wrote:EADS is pushing TETRA? Surely you mean TETRAPOL?

After buying "Nokia Tetra25" they are able to give you both Tetra25 and TetraPOL.