ICOM & KENWOOD P25 IMBE RADIOS

[ASTROMODAT]

Anyone know what the status is re: ICOM & Kenwood P25 radios?

Larry

[Justin]

Vertex/Yaesu is slated to have some at the end of 2003. Icom and Kenwood should have similar to compete within a short time of that.

[USPSS]

Kenwood has a P25 trunking portable that works wonderful. I don't know about Icom or Vertex

[Hartley]

Hi,

What is the model number of the Kenwood P25 radio? I couldn't find any mention of such a thing on Kenwood's web site.

Hartley

[USGOVTECH]

Hello All,

The Kenwood model is the TK5400 (a portable) and very small. Currently the Kenwood 90 Series radios are PHASE I CAI P25 capable as they sit, just a firmware upgrade and you are off and rolling. The 5400 will be PHASE II 4 Slot TDMA standard and CAI compliant.

I hope that this helps.

[ASTROMODAT]

Does Kenwood plan to bring out an APCO P25 compliant mobile?

How about a Kenwood P25 portable for UHF conventional (non-trunked)?

BTW, I thought CAI IMBE was FDMA, as opposed to TDMA (like Tetra)?

Larry

[Twisted_Pear]

Does Kenwood plan to bring out an APCO P25 compliant mobile?

How about a Kenwood P25 portable for UHF conventional (non-trunked)?

BTW, I thought CAI IMBE was FDMA, as opposed to TDMA (like Tetra)?

Larry

Nope, TDMA. Usually TDMA is used to multiplex on a traffic channel when there isn't a large chunk to spread FDMA out over.

-Wayne

[ASTROMODAT]

Wayne, for the same throughput, FDMA uses the exact same amount of bandwidth as does TDMA. CDMA is the technique that is also known as "spread spectrum" that uses so much spectrum.

BTW, how come Motorola's technical sheets on ASTRO Saber and ASTRO Spectra both show FDMA as the digital modulation type if APCO 25 is TDMA?

Larry

[xmo]

APCO 25 Phase I is 12.5 KHz (2:1) FDMA

APCO 25 Phase II was originally supposed to be 4:1 FDMA (6.25 KHz) but is apparently being expanded to also include a TDMA solution as well as FDMA - a sort of double standard. (basically bending to European vendor pressure (Ericcson & Nokia) who claimed that FDMA was a strictly Motorola solution)

[ASTROMODAT]

When is Phase II likely to be implemented? Sounds like lots of current APCO 25 gear runs the risk of not being compatible if TDMA fans (Tetra) wins this battle!

Larry

[apco25]

pardon me, but TETRA BLOWS and should stay where it belong in europe!

[oh2lak]

[quote="apco25"]pardon me, but TETRA BLOWS and should stay where it belong in europe![/quote]

Is this because there's no monitoring gear available for TETRA? :)

I like TETRA. Works like a charm and audio quality is superb comparing to P25. Not saying that IMBE is bad but it gets bad if you listen to TETRA :)


..
Erik

[apco25]

Things I don't like about TETRA system. Typical lack of any off system use i.e conventional. No network, no radio.

I've also found it to be extremely infrastructure heavy due to the micro cell concept.

[Wowbagger]

APCO 25 Phase I is 12.5 KHz (2:1) FDMA

APCO 25 Phase II was originally supposed to be 4:1 FDMA (6.25 KHz) but is apparently being expanded to also include a TDMA solution as well as FDMA - a sort of double standard. (basically bending to European vendor pressure (Ericcson & Nokia) who claimed that FDMA was a strictly Motorola solution)

Actually, the QPSK solution (6.25 kHz) is only a 2 to 1 advantage over C4FM (12.5 kHz). The problem with it is that you have to have linear amplifiers for QPSK as opposed to the non-linear class C amps you can use for C4FM, which increases the cost of the system (and it increases the power consumption of the mobiles as the class AB amps you need for QPSK are much less efficient than the class C amps used in C4FM systems.)

Motorola is working on their 2 slot TDMA system as we speak, and that's about all I can say without violating my non-disclosure agreement.

As for the difference in voice quality - that's pretty much entirely due to the difference in vocoders - IMBE blows chunks (there's rather a nasty bug in it that causes an asymmetry in the recovered audio), AMBE is a bit better (they fixed that bug), and I'd still personally rather see them use speex.

[mr.syntrx]

Things I don't like about TETRA system. Typical lack of any off system use i.e conventional. No network, no radio.

You can use it off the network - TETRA calls that DMO (Direct Mode Operation). There's also support for vehicular repeaters etc too.

[wavetar]

Motorola is working on their 2 slot TDMA system as we speak, and that's about all I can say without violating my non-disclosure agreement.

Would that not just be "MotoTRBO"? In which case, what's the secret, since it's now available?

As for the difference in voice quality - that's pretty much entirely due to the difference in vocoders - IMBE blows chunks (there's rather a nasty bug in it that causes an asymmetry in the recovered audio), AMBE is a bit better (they fixed that bug), and I'd still personally rather see them use speex.

We're talking vocoders that were obsolete when they were implemented...IMBE was introduced in the early 90's and was the sh!t at the time...certainly quantum leaps above the Motorola SecureNet algorithm. Soon after, AMBE was a clear level beyond that...but technology has advanced exponentionaly since then...so of course other vocoding methods would stand head & shoulders above both of those...but there has to be a standard which is set upon for initial standardization.

If the RF world decided to base the future on Speex, another algorithm would blow it out of the water 3 years from now & we would be dealing with the pundits saying..."we should have done such & such...". We can't please everybody.

I for one think the recovered audio with IMBE is satisfactory 99% of the time...I believe most people's poor experiences with it are in regards to poorly constructed/implemented systems...surely with the SZOL system here in my area, the problems with IMBE are very few & far between.

Todd

[batdude]

firmware firmware firmware



DSP8 is miles different from even DSP7.


the problem with M is that new firmware = new revenue stream....


ho hum



doug

[wavetar]

firmware firmware firmware



DSP8 is miles different from even DSP7.


the problem with M is that new firmware = new revenue stream....


ho hum



doug

Motorola doesn't introduce the new Firmware to improve IMBE...DVSI does. Not only that, but intrinsically IMBE will never reach the recovered voice quality that algorithms created afterwards with newer technology can. However, I agree with you that Motorola sees advancing the recovered audio with IMBE as a cash cow.

Todd

[mancow]

As for the difference in voice quality - that's pretty much entirely due to the difference in vocoders - IMBE blows chunks (there's rather a nasty bug in it that causes an asymmetry in the recovered audio), AMBE is a bit better (they fixed that bug), and I'd still personally rather see them use speex.

I agree, I listened to examples of the speex vocoder and it's incredible what it can do.

Also, what about 0-QPSK? I was reading some stuff about Inmarsat M and it seems they use that along with AMBE. From what I read the O-QPSK helps with the amplifier issue by "smoothing" out the overall modulation.

[oh2lak]

Things I don't like about TETRA system. Typical lack of any off system use i.e conventional. No network, no radio.

I've also found it to be extremely infrastructure heavy due to the micro cell concept.

TETRA is a cell-concept infrastructure, if you build it like having conventional base stations dropped here and there, it won't work. It has to be designed and build like a cell network. Cell density is not very big due to low frequency, one cell can easily handle 20km radius in non-urban environment, microcells do not exist. Indoor coverage is made usually with smart repeaters or extending the cell providing service outdoors also to indoor premises. TETRA signalling has the similar features than cellulars to make handovers between cells etc.

TETRA spec has DMO (Direct Mode) for 'simplex' communication, the spec also defines DMO repeaters and gateways to provide direct mode - infrastructure communications. Most modern mobile units can do both of these functions.

I use both P25 and TETRA, so this is not religion thing, but TETRA has gone further IMHO. It has more advanced PSTN features, full-duplex terminals, advanced multi-layer encryption and data services like SMS, GPRS etc.

Just my 2 eurocents :)

..
Erik

[Wowbagger]

I agree, I listened to examples of the speex vocoder and it's incredible what it can do.

Also, what about 0-QPSK? I was reading some stuff about Inmarsat M and it seems they use that along with AMBE. From what I read the O-QPSK helps with the amplifier issue by "smoothing" out the overall modulation.

Orthogonal QPSK will reduce the AM a bit, but at the expense of widening the spectrum a bit. Also, OQPSK is *not* decoder compatible with C4FM - that's the whole reason the parameters of CQPSK are what they are - so that at symbol time, the instantaneous deviation is either +/- 600 or +/- 1800 Hz just like C4FM.

What scares me is the next logical step in P25 - combine the TDMA system with the CQPSK system to get 2 conversations in a single 6.25 kHz channel.

As for TETRA vs. APCO-25 - personally, I don't care: either way you are going to be buying your test gear from Aeroflex, so I get paid the same. From my standpoint, the fact that the TETRA standard is more rigorously defined (protocol state machine defined in SDL rather than just implied) makes it nicer.

[compuman81]

As far as the audio quality improving, well , it has. Why?

Motorola has made advancements in the radio's ability to tear the bits off the air and provide an overall cleaner more accurate bitstream into the DVSI chip = more clarity. IMBE will always sound like IMBE.... the codec hasn't changed ( the code running in the embedded DVSI chip ), however the application of IMBE in the field being used as a replacement for analog audio will become more plausible due to these improvements.

[Wowbagger]

As far as the audio quality improving, well , it has. Why?

Motorola has made advancements in the radio's ability to tear the bits off the air and provide an overall cleaner more accurate bitstream into the DVSI chip = more clarity. IMBE will always sound like IMBE.... the codec hasn't changed ( the code running in the embedded DVSI chip ), however the application of IMBE in the field being used as a replacement for analog audio will become more plausible due to these improvements.

Sorry, but you are wrong. The bit error rate into the IMBE vocoder on a normal P25 signal is close enough to zero that you will get no errors in many minutes of audio.

Only when the signal quality is REALLY bad - as in, you are at the fringe of the signal - do you get any bit errors going into the vocoder. Hell, you really don't even get that many bits of error going into the forward error correction codes in front of the vocoder - my experience is usually about a bit error a minute in normal use, and that bit error is corrected by the FEC. Remember, there's about as many bits of forward error correction as there are actual vocoder bits.

And there is a BIG difference between the IMBE vocoder and the AMBE vocoder - like I said, there was a bug in the IMBE vocoder that they fixed in the AMBE vocoder, so that the AMBE vocoder, even when handling IMBE data, is a more faithful representation of the input audio.

[mr.syntrx]

The software upgrades affect how the radio processes the decoded audio after it comes out of the vocoder.

[Wowbagger]

The software upgrades affect how the radio processes the decoded audio after it comes out of the vocoder.

What kind of processing are you thinking of? Once you have the 8kSample/second PCM data out of the vocoder, there is precious little you need to do other than stuff it into the DAC driving the speaker.

[compuman81]

I'm obviously not talking about normal operation.................. the system works as designed to provided there's a decent signal.


My post was regarding performance of my radios on the county wide trunked system talkgroups that run digital and as well as the conversations i have on simplex. Have you ever brought your astro into a big steel skyscraper and tried to listen to a conversation with that much blockage, even when the transmitter is a few blocks down? Compare performance between a radio manufactured in 1996 versus a recently purchased XTS 5000. Tell me if you think they behave the same. I sure don't. Obviously DVSI hasn't changed anything, so where's the difference on motorola's part? The radios certainly couldn't have got more sensitive?

[Wowbagger]

I'm obviously not talking about normal operation.................. the system works as designed to provided there's a decent signal.

OK, that I can agree with, but that caveat was not clear before.

The radios certainly couldn't have got more sensitive?

Well, there's two possibilities for that. First, yes they most certainly CAN have become more sensitive - there's been some good improvements between 1996 and now - especially if they have gone from "demodulate the FM down to baseband and then try to detect the C4FM" to "digitize the IF and then process it from there."

Then there can be improvements in the data slicer and recovery routines - being better able to decode the bits before the rest of the system can help in high received modulation fidelity error situations.

Then there can be improvements in how they apply their forward error correction - they can be more fully using the FEC to correct more bits of errors.

[compuman81]

I'm obviously not talking about normal operation.................. the system works as designed to provided there's a decent signal.

OK, that I can agree with, but that caveat was not clear before.

The radios certainly couldn't have got more sensitive?

Well, there's two possibilities for that. First, yes they most certainly CAN have become more sensitive - there's been some good improvements between 1996 and now - especially if they have gone from "demodulate the FM down to baseband and then try to detect the C4FM" to "digitize the IF and then process it from there."

Then there can be improvements in the data slicer and recovery routines - being better able to decode the bits before the rest of the system can help in high received modulation fidelity error situations.

Then there can be improvements in how they apply their forward error correction - they can be more fully using the FEC to correct more bits of errors.

These are the technical improvements that I was trying to refer to antecedently. I'm not familiar with the details of all the components in the radio, but I have a good understanding of the whole audio on one side to audio on the otherside process.

I guess with the astros there is not only one "sensitivity" anymore. It used to refer to just the radio receiver portion, but now I suppose one could look at the digital front end system in its entirety and starting comparing numbers of the system's overall ability to produce good sounding audio in the end.