WCDMA Basics

DS-CDMA means that user data is spread over a much wider bandwidth
through multiplication by a sequence of pseudo-random bits called chips.
Figure 4-3 provides a conceptual depiction of this spreading. One can see
that the user data, at a relatively low rate compared to the rate of the
spreading code, is spread over a signal that has a higher bit rate. We can
also see that the signal that is transmitted has pseudo-random characteristics.
When transmitted over a radio interface, the spread signal looks like
noise.

If multiple users transmit simultaneously on the same frequency, then
the stream of data from each user needs to be spread according to a different
pseudo-random sequence. In other words, each user data stream needs
to be spread according to a different spreading code. At the receiving end,
the stream of data from a given user is recovered by despreading the set of
received signals with the appropriate spreading code. Of course, what is
being despread is the complete set of signals received from all users that
are transmitting.

Imagine, for example, two users (A and B) that are transmitting on the
same frequency, but with two different spreading codes. If, at the receiving
end, the received signal is despread with the spreading code applicable to
user A, then the original data stream from user A is recovered. The data
stream that is recovered does have some noise created by the fact that the
received signal also contains user data from user B. The noise, however, is
small.
Similarly, if the received signal is despread according the spreading code
used by user B, then the original data stream from user B is recovered, with
a little noise generated by the presence of user A’s data within the spread
signal. Provided that the rate of the spreading signal (the chip rate) is far
larger than the user data rate, then the noise (that is, the interference) generated
by the presence of other users will be sufficiently small to not inhibit
the recovery of the data steam from a given user. Of course, as the number
of simultaneous users increases, so does the interference and it eventually
becomes impossible to recover a specific user’s data with any confidence.
In other words, for a given bit of recovered user data, the signal-to-noise
ratio must be sufficiently high. In CDMA, we refer to Eb/No, where Eb is the
power density per bit of recovered user data and No is the noise power density.
Provided that Eb/No is sufficiently large, then the user data can be recovered.
The ratio of the chip rate to the user data symbol rate is known as the
spreading factor. The capability to recover a given user’s signal is directly
influenced by the spreading factor. The higher the spreading factor, the
greater the capability to recover a given user’s signal. In terms of transmission
and reception, a higher spreading factor has an equivalent effect as
transmitting at a higher power. Thus, the magnitude of the spreading factor
can be considered a type of gain and is known as the processing gain. In
dB, the processing gain is given by 10
10Log10 (spreading rate/user rate).
In some cases, this can be quite a large number and can help to overcome
the effect of interference generated by the presence of other users.
If, for example, the processing gain for a given CDMA service were 20 dB
and if an Eb/No value of 5 dB were needed, then for a given user, the signalto-interference ratio can be as low as -15 dB and the user’s signal can still
be recovered. This is because the despreading benefits from the processing
gain of 20 dB. Note that, for a given chip rate, the processing gain for lowbit-
rate user applications is greater than for high-bit-rate applications,
which often means that lower-bit-rate applications can tolerate more interference
than high-bit-rate applications.

The WCDMA air interface of UMTS (hereafter simply WCDMA) has a
nominal bandwidth of 5 MHz. While 5 MHz is the nominal carrier spacing,
it is possible to have a carrier spacing of 4.4 MHz to 5 MHz in steps of 200
kHz. This enables spacing that might be needed to avoid interference, particularly
if the next 5-MHz block is allocated to another carrier.

The chip rate in WCDMA is 3.84
106 chips/second (3.84 Mcps). In theory,
for a speech service at 12.2 Kbps (and, for now, assuming no extra bandwidth
for error correction), the spreading factor would be 3.84
106/12.2

103  314.75. This would equate to a processing gain of 25 dB. In reality,
however, WCDMA does include extra coding for error correction. Consequently,
a spreading factor as high as 314.75 is not supported, at least not
in the uplink. The supported uplink spreading factors are 4, 8, 16, 32, 64,
128, and 256. The highest spreading factor (256) is used mostly by the various
control channels. Some control channels can also use lower spreading
factors, while user services generally use lower spreading factors.
Table 4-3 provides a summary of the spreading factors and the corresponding
data rates on the uplink.

At first glance, it appears that the lowest spreading factor (4) provides a
gross rate of only 960 Kbps and a usable rate of only 480 Kbps. This does
not meet the requirements of IMT-2000, which states that a user should be
able to achieve speeds of 2 Mbps. In order to meet that requirement, UMTS
supports the capability for a given user to transmit up to six simultaneous
data channels. Thus, if a user wants to transmit user data at a user rate
greater than 480 Kbps, then multiple channels are used, each with a
spreading factor of four.With six parallel channels, each at a spreading factor
of four, a single user can obtain speeds of over 2 Mbps.
In the downlink, the same spreading factors are available, with a spreading
factor of 512 also possible. One difference between the uplink and downlink,
however, is the number of bits per symbol. As will be described in
Chapter 6, “Universal Mobile Telecommunications Service (UMTS),” the
uplink effectively uses one bit per user symbol, while the downlink effectively
uses two bits per user symbol. Consequently, for a given spreading
factor, the user bit rate in the downlink is greater than the corresponding
bit rate in the uplink. The user rate in the downlink is not quite twice that
in the uplink, however, due to differences in the way that control channels
and traffic channels are multiplexed on the air interface. The details of
uplink and downlink transmissions are provided in Chapter 6. Table 4-4
provides a summary of the spreading factors and the corresponding data
rates on the downlink.

As is the case for the uplink, WCDMA supports multiple simultaneous
user data channels in the downlink, so that a single user can achieve rates
of over 2 Mbps. It should be noted, however, that Table 4-4 does not tell the
whole story of possible data rates on the downlink. WCDMA supports a concept
known as compressed mode, whereby gaps exist in downlink transmission
so that the terminal can take measurements on other frequencies.
When compressed mode is used, a reduction will take place in the data rate
compared to that shown in Table 4-4.
An important capability of WCDMA is that user data rates do not need
to be fixed. In WCDMA,channels are transmitted with a 10-ms frame structure.
It is possible to change the spreading factor on a frame-by-frame basis.
Thus, within one frame, the user data rate is fixed, but the user data rate
can change from frame to frame. This capability means that WCDMA can
offer bandwidth on demand. Note that rate changes every 10 ms do not
apply to AMR speech as each speech packet is 20 ms in duration, so that the
speech rate can change every 20 ms if needed, but not every 10 ms.