Walsh codes in CDMA

hey....I would like to know something about length 64 Walsh codes that
are applied in CDMA 2000 1xEVDO....plz help....what is the difference
between Walsh codes and PN sequences.

Thanks
TD

koolguyuf wrote:

> hey....I would like to know something about length 64 Walsh codes that
> are applied in CDMA 2000 1xEVDO....plz help....what is the difference
> between Walsh codes and PN sequences.
> 
> Thanks
> TD
> 
The codes in a given set of Walsh codes are mutually orthogonal.

The codes in a given set of 5-bit PN sequences are guaranteed to _not_ 
be mutually orthogonal -- at least if you're talking about PN sequences 
generated by linear shift-and-xor.

A _long_ PN sequence can be _almost_ orthogonal with other PN sequences, 
and with shifted values of itself, but it's never going to be _exactly_ 
orthogonal.  PN sequences are cool for spread spectrum because they're 
going to be orthogonal to everything but another PN sequence of the same 
length, and almost orthogonal to that unless you have the secret decoder 
ring combination.  They're a pain because it takes a long time to 
synchronize up to them, and they have to be generated, etc., etc.

-- 

Tim Wescott
Wescott Design Services
http://www.wescottdesign.com

Posting from Google?  See http://cfaj.freeshell.org/google/

"Applied Control Theory for Embedded Systems" came out in April.
See details at http://www.wescottdesign.com/actfes/actfes.html

koolguyuf wrote:

> hey....I would like to know something about length 64 Walsh codes that
> are applied in CDMA 2000 1xEVDO....plz help....

Not sure about CDMA2000 however the Walsh system of the length 64 was 
used in the deceased IS-95.

> what is the difference
> between Walsh codes and PN sequences.

There is nothing in common. Those are the absolutely different things 
used for absolutely different purposes.

VLV

koolguyuf wrote:
> hey....I would like to know something about length 64 Walsh codes that
> are applied in CDMA 2000 1xEVDO....plz help....what is the difference
> between Walsh codes and PN sequences.
>
> Thanks
> TD

The transmitter modulates multiplies the channels (pilot, traffic, etc)
with different Walsh codes. The codes are perfectly orthogonal to each
other. The length is variable (not fixed at 64).

The entire transmission, containing a sum of multiple channels
separated by Walsh codes, is multiplied by a 32768-chip PN sequence.
Each tower has a unique PN code offset so that multiple towers can use
the same frequency.

John

"Vladimir Vassilevsky" <antispam_bogus@hotmail.com> wrote in message 
news:8LGQg.7092$Ij.4415@newssvr14.news.prodigy.com...

>> what is the difference
>> between Walsh codes and PN sequences.
>
> There is nothing in common. Those are the absolutely different things used 
> for absolutely different purposes.


Well, I wouldn't like to make such strong statements as nothing in
common.  One bit is the same in each Walsh code, and if we drop
that, there is a fixed permutation that will map the remaining bits
of each Walsh code (except the +++....+ code) into a different
cyclic shift of a PN sequence. Conversely, there is a fixed
permutation that can be applied to each cyclic shift of a PN
sequence to get the different Walsh codes (except for the
extra bit that is the same for all Walsh codes, and of course, we
don't get the ++++....+ sequence).  In particular, if we wish to
compute the all the cross-correlation values of a sequence x with
a PN sequence, the easiest way to do this is to apply the fixed
permutation to x and compute the Hadamard transform of the
permuted x.  A fast Hadamard transform algorithm is even more
efficient than a fast Fourier transform algorithm since it needs
O(N log N) additions and subtractions and NO multiplications,
and was, in fact, used for telemetry data from the Mariner 9
mission to Mars about 35 years ago. See, for example,
Chapter 14 of the MacWilliams and Sloane book "Theory of
Error-Correcting Codes"

--Dilip Sarwate 

Dilip V. Sarwate wrote:
> 
> Well, I wouldn't like to make such strong statements as nothing in
> common.  One bit is the same in each Walsh code, and if we drop
> that, there is a fixed permutation that will map the remaining bits
> of each Walsh code (except the +++....+ code) into a different
> cyclic shift of a PN sequence.

Very interesting. I didn't know about this property. Walsh and PN seem 
to belong to the different classes...

> See, for example,
> Chapter 14 of the MacWilliams and Sloane book "Theory of
> Error-Correcting Codes"

Certainly. I need to read this book.

VLV

johns@3db-labs.com wrote:
> koolguyuf wrote:
> > hey....I would like to know something about length 64 Walsh codes that
> > are applied in CDMA 2000 1xEVDO....plz help....what is the difference
> > between Walsh codes and PN sequences.
> >
> > Thanks
> > TD
>
> The transmitter modulates multiplies the channels (pilot, traffic, etc)
> with different Walsh codes. The codes are perfectly orthogonal to each
> other. The length is variable (not fixed at 64).
>
> The entire transmission, containing a sum of multiple channels
> separated by Walsh codes, is multiplied by a 32768-chip PN sequence.
> Each tower has a unique PN code offset so that multiple towers can use
> the same frequency.
>
> John


Hey John,

Thanks for the reply. I assume that you are talking about the forward
link, i.e from the Base Station to the mobiles ---- what happens on the
reverse link then --- if every mobile uses the same frequency (as in
CDMA) do each of the mobiles in a particular sector have a unique PN
sequence ?

Thanks,
TD

koolguyuf wrote:
> johns@3db-labs.com wrote:
> > koolguyuf wrote:
> > > hey....I would like to know something about length 64 Walsh codes that
> > > are applied in CDMA 2000 1xEVDO....plz help....what is the difference
> > > between Walsh codes and PN sequences.
> > >
> > > Thanks
> > > TD
> >
> > The transmitter modulates multiplies the channels (pilot, traffic, etc)
> > with different Walsh codes. The codes are perfectly orthogonal to each
> > other. The length is variable (not fixed at 64).
> >
> > The entire transmission, containing a sum of multiple channels
> > separated by Walsh codes, is multiplied by a 32768-chip PN sequence.
> > Each tower has a unique PN code offset so that multiple towers can use
> > the same frequency.
> >
> > John
>
>
> Hey John,
>
> Thanks for the reply. I assume that you are talking about the forward
> link, i.e from the Base Station to the mobiles ---- what happens on the
> reverse link then --- if every mobile uses the same frequency (as in
> CDMA) do each of the mobiles in a particular sector have a unique PN
> sequence ?
>
> Thanks,
> TD

Affirmative. The mobiles operate on different Walsh channels. 

John

good replies by greats indeed,


walsh codes are used in synchronous CDMA or synchronous MC-CDMA or
synchronous TFL-CDMA or synchronous MC-DS CDMA,

and when synchronous walsh codes offer perfect orthogonality, in the
sense complete zero Cross correlation.

and its design is quite easy, its based on quadrant division method (i
mean to say you can construct any walsh for that matter on paper within
seconds)

PN sequences are having near to zero Cross correlation.

as told above these two are used for completely different purpose.

regards
particle (filter) reddy

johns@3db-labs.com wrote:
> koolguyuf wrote:
> > johns@3db-labs.com wrote:
> > > koolguyuf wrote:
> > > > hey....I would like to know something about length 64 Walsh codes that
> > > > are applied in CDMA 2000 1xEVDO....plz help....what is the difference
> > > > between Walsh codes and PN sequences.
> > > >
> > > > Thanks
> > > > TD
> > >
> > > The transmitter modulates multiplies the channels (pilot, traffic, etc)
> > > with different Walsh codes. The codes are perfectly orthogonal to each
> > > other. The length is variable (not fixed at 64).
> > >
> > > The entire transmission, containing a sum of multiple channels
> > > separated by Walsh codes, is multiplied by a 32768-chip PN sequence.
> > > Each tower has a unique PN code offset so that multiple towers can use
> > > the same frequency.
> > >
> > > John
> >
> >
> > Hey John,
> >
> > Thanks for the reply. I assume that you are talking about the forward
> > link, i.e from the Base Station to the mobiles ---- what happens on the
> > reverse link then --- if every mobile uses the same frequency (as in
> > CDMA) do each of the mobiles in a particular sector have a unique PN
> > sequence ?
> >
> > Thanks,
> > TD
>
> Affirmative. The mobiles operate on different Walsh channels.
>
> John

Hi John,

Thanks for the help .I have one more question.
I would like to know the mathematical/physical/intuitive significance
of
an erasure threshold in the decoder of a communication system. I am
working in the simulation of a reverse link channel in the cdma 2000
system.

Thanks,