Scrambling in UMTS.
Prathap
1) What exactly is scrambling in UMTS. I understand by chanelization,
but what is scramling??
for ex:
if 10101011 is my bit after chanelization, and 11001100 is my
scrambling code,
what might happen to my spread code( 10101011) after scrambling.
Is scramling some kind of code that I insert as header before my
data, or does scrambling alter my data?
I have read that in the downlink the scrambling code is 38400 chips,
which constitutes 10 ms, and this scrambling is used to seperte data
from one cell with another. So, if my data frame is 10ms, then the 10ms
scrambling code is applied to this and it is transmitted.. am I
correct? ( By scrabmling I assume here that I some how alter actual
signals transmitted) .
2) In UMTS,
Downlink: chanelisation is used to seperate connections of
different users within one cell ( I assume it also seperates different
connections/channels of the same user), and scrabmling is used to
seperate cells.
Uplink: Chanelization is used to seperate different channels of the
same UE and scrambling is used to seperate UE's.
What I don't understand is, why can't uplink be the same way as
downlink?
app
can model it with 1's and -1's and then it is multiplication) with the
channelization code before modulating the (channellized and now also
scrambled) chipstream to QPSK symbols. In reception, the QPSK symbols
are de-modulated in a clever CDMA receiver and then "de-xored" with the
scrambling code, to get the chipstream back.
(NOTE ! I use the word "chip" here as "one element of a CDMA code",
irrelevant of the possible modulation frequency and technique. This is
a source of endless confusion, because modulation frequency = chip rate
= number of CDMA elements per second in narrowband cdmaOne with BPSK,
but in QPSK the modulation frequency (still often called "chip rate")
is half of the number of CDMA "elements" (often called "chips"!!!) sent
per second. You can also say, that the baud rate in QPSK is half of the
bit rate, and you can not call them both "chip rate"!!!)
Scrambling codes are long, like you said. They have a cycle, which
consist of pieces of the same length as the CDMA code (=spreading
factor). In your example the scrambling code 11001100 is just an 8-bit
part of the 38400 long sequence, which will cycle again after 38400/8 =
4800 CDMA codes.
In UMTS, the uplink and downlink scrambling are really similar, it is
just how you want to paraphrase it. The idea is, that even if two
sources are using the same CDMA code, a diffrent scrambling code makes
their signals unique: In DOWNLINK , the sources are tranceivers, one in
a cell. So cells sectored in the same location or in neighbouring base
stations can re-use the official CDMA code tree without coordination.
UPLINK the sources are handsets. So when each handset has a unique
scrambling code, they can again re-use the UMTS official CDMA code
tree.
BTW, here is something that I don't understand: why do not the
scrambling codes destroy the orthogonality? After scrambling, we use
what is effectively random codes, how can it work? What is the
difference, if we actually picked the codes at random, with no regard
to the official UMTS code tree?