TETRA phase adjustment bits

Hi!

I know this is slightly off-topic, as it is not really related to
gnuradio
(despite the USRP and gnuradio being used in my current experiments
regarding
TETRA):

Does anyone know how the phase alignment bits in TETRA exactly work?

I’m reading ETSI EN 300 392-2 V3.2.1 and can find the definition at
Section 9.4.4.3.6.

So the phase adjustment bits need to be set to the negated sum of the
phase shift of two other bits in the burst.

Let’s assume I’m using hc1, hc2 and thus look at the symbols 8 (bit
15+16) and
108 (bit 215+216).

If all of those four bits are 1 (like in the case of the synchronization
continuous down-link burst), there is a total phase shift of -6pi/4 (two
times
-3pi/4).

But how do you correct such a phase shift in only a single adjustment
symbol?
Using the two bits in the adjustment symbol, I can only apply +1pi/4,
-1pi/4,
+3pi/4 or -3pi/4 ?!?

Any help is appreciated.

Regards,
Harald

  • Harald W. [email protected]
    http://laforge.gnumonks.org/
    ============================================================================
    “Privacy in residential applications is a desirable marketing option.”
    (ETSI EN 300 175-7 Ch.
    A6)

On Sun, Jan 02, 2011 at 09:02:09PM +0100, Harald W. wrote:

So the phase adjustment bits need to be set to the negated sum of the
phase shift of two other bits in the burst.

sorry, my mistake. This should read: “… the negated sum of the phase
shift
of two other symbols in the burst.”

  • Harald W. [email protected]
    http://laforge.gnumonks.org/
    ============================================================================
    “Privacy in residential applications is a desirable marketing option.”
    (ETSI EN 300 175-7 Ch.
    A6)

On Sun, Jan 2, 2011 at 3:02 PM, Harald W. [email protected]
wrote:

Hi!

I know this is slightly off-topic, as it is not really related to gnuradio
(despite the USRP and gnuradio being used in my current experiments regarding
TETRA):

I think this list is a fine place to ask these kinds of questions.

Regards,
Harald

Harald,
Have you looked at this paper?
http://wsl.stanford.edu/~ee359/doppler2.pdf

It provides a very nice description of pi/4 DQPSK and a receiver
architecture for it. It might provide the answers you’re looking for.

Tom

Dear Tom (and others),

On Mon, Jan 03, 2011 at 11:22:11AM -0500, Tom R. wrote:

So the phase adjustment bits need to be set to the negated sum of the
phase shift of two other bits in the burst.

Let’s assume I’m using hc1, hc2 and thus look at the symbols 8 (bit 15+16) and
108 (bit 215+216).

A friend has meanwhile (privately) solved the problem: It is not the sum
of the
phase shift of two symbols, but the sum of the phase shift of all
symbols in
the range n1…n2. And as that range is always an odd number, we can
always
compensate the cumulative phase shift in a single additional symbol.

Have you looked at this paper?
http://wsl.stanford.edu/~ee359/doppler2.pdf

I had seen this before, but thanks anyway.

I’ll follow-up to this group once I’m getting anywhere with my code.

  • Harald W. [email protected]
    http://laforge.gnumonks.org/
    ============================================================================
    “Privacy in residential applications is a desirable marketing option.”
    (ETSI EN 300 175-7 Ch.
    A6)