. 60
( 132 .)


these mechanisms will be dealt with in turn in the following subsections. The key goals
of power control are to:

• provide each UE with suf¬cient quality regardless of the link condition or distance
from the BTS;
• compensate for channel degradation such as fast fading and attenuation;
• optimize power consumption and hence battery life in the UE. Open loop
This is used when a handset ¬rst enters the cell. The handset will monitor the CPICH of
the cell and will take received power level measurements of this channel. This information
will be used when setting its own power level. The power radiated by the base station will
reduce as the distance from the tower increases. Simply having a received power level
is not enough information for the mobile device to set its own transmit power. This is
because it does not know at this stage at what power level the CPICH was transmitted. The
base station sends out power control information on the broadcast channel, which includes
an indication of the power level at which the CPICH is being transmitted. The UE can now
determine how much power has been lost over the air interface and thus has an indication
of its own distance from the base station. The handset can now calculate what power to
transmit at; if it has received a weak signal it will transmit a strong signal since it assumes
that it must be a long distance from the base station. Conversely, if it receives a strong
signal it assumes that it is near to the base station and can thus send a weaker signal.

The actual mechanism is that the UE will listen to the broadcast channel. From this, it
will ¬nd out the following parameters:

• CPICH downlink transmit power
• uplink interference
• constant value.

The UE will then measure the received power of the pilot (CPICH RSCP). The initial
power used is then:
Initial power = CPICH downlink transmit power ’ CPICH RSCP
+ UL interference + Constant value
The constant value really provides a correction ¬gure for better approximation of an
appropriate start power. It should be noted that this is a rather approximate mechanism
since the base station will be transmitting in a different frequency band to what the
handset will use and the power loss over the air interface may be signi¬cantly different
for each band.
Figure 6.17 gives an example of how two different mobile devices, depending on their
distances from the base station, will send an access request on the RACH at different
power levels. This will reduce interference to other mobile devices. The diagram shows
separate BCCH and CPICH for the two different mobile devices; in reality only one of
each of these is broadcast and all the mobile devices receive it. Inner loop
Inner loop power control feedback is a form of closed loop power control. Information is
sent in every slot, i.e. 1500 times per second. This can be compared with IS-95, which
has feedback 800 times a second, and GSM, where it is only carried out approximately

Station -1 BCCH: Transmit Power of CPICH is 2 watt
CPICH: (power reference)

RACH: connect request (higher power)
er o
received Mobile sm tt )
ran 2 wa rence
power 0.5 Watt : T is
Station -2 H fe
BC PICH wer re t
C (po
nec r)
CP con powe
CH r
RA (lowe
power 1 Watt

Figure 6.17 Example of open loop power control

twice per second. This type of power control is required because the open loop system
can only give a rough estimate and is not accurate enough to deal with problems such as
fast fading. To control the power level on the uplink, the base station performs frequent
signal to interference ratio (SIR) measurements on the received signals from each of the
mobiles. This value is compared to a target SIR value and if the power from the mobile
station is deemed to be too high or low it will be told to decrease or increase the power
accordingly. Since this task is executed 1500 times per second, it is much faster than
power control problems, such as fast fading, that may occur, and hence can compensate
for these. This fast power control is very effective for slow to moderate movement speeds
of the mobile device. However, bene¬ts decrease as the speed of the mobile increases.
This also deals with the near“far problem, where signals from mobile devices which are
far from the base station will suffer greater attenuation. The object of the fast power
control is that the signal from each mobile should arrive at the BTS at its target SIR
value. The same type of power control is used on the downlink. When communicating
with mobile devices that are on the edge of the cell the base station may marginally
increase the power it sends. This is required since these particular mobiles may suffer
from increased other-cell interference. Outer loop
As noted, inner loop power control measures the power from the mobile device and
compares it to a set SIR target. This target value is set and adjusted by the outer loop
power control within the RNC. This value will change over time but does not need to be
adjusted at the same high frequency. The target value is actually derived from a target
BER or BLER that the service is expected to meet. Some errors with the data received
from the mobile device are expected. If there are no errors, then the UE is assumed to
be transmitting at too high a power, with the consequences of causing interference and
reducing the battery life of the device. To implement this method of power control the
mobile device will compute a checksum before sending any data. Once received, a new
checksum is computed on the data and this is compared to the one sent by the mobile
device. The BTS will also measure the quality of the received data in terms of BER. If
too many frames are being received with errors, or frames have too high a BER, then the
power can be increased. The target set-point is not static: it does change over time. This
is required so that the cell can be more ef¬ciently utilized.

The physical layer is structured into radio frames, each of 10 ms duration, and a radio
frame is divided up into 15 slots, as shown in Figure 6.18. Blocks of data are transferred
across the air interface in each radio frame, and the data rate at which information is sent
may change with radio frame granularity. Within a slot, control functions such as power
control take place.

10 ms 10 ms

Radio Frame Radio Frame

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14

15 slots: 2/3ms each

Figure 6.18 UMTS radio frame structure

6.8.1 Physical layer procedures
Figure 6.19 shows a synopsis of the procedures that must be performed at the physical
layer for transfer of data across the Uu air interface. In this diagram, two transport channels
are processed and then combined into a single channel, referred to as a coded composite
transport channel (CCTrCH). Since this multiplexing of channels is done at the physical
layer, this allows one user to have a number of simultaneous transport channels, each with
their own QoS pro¬le. This is used, for example, when a subscriber is making a phone call
(logical DTCH) and has a control signalling channel at the same time (logical DCCH).
The physical layer must ensure that data is transferred reliably across the air, so the
processes of cyclic redundancy check (CRC) attachment, coding and interleaving are
designed to provide this. The composite channel is spread with the appropriate codes and
then modulated for transmission. Each of these steps is now explained in further detail.

6.8.2 Data protection
UMTS provides for three basic layers of protection on the data: the CRC to detect for
errors, convolution coding for forward error correction (FEC) and interleaving to dis-
tribute burst errors throughout the data. All of these general principles were described in
Chapter 2. The CRC size can be 0, 8, 12, 16 or 24 bits, and the actual size to be used for
a particular transport channel is de¬ned when the channel is established by higher-layer
signalling. At the receiver the CRC result is passed upward to higher layers and erro-
neous data is not always discarded since this corrupted data may still be of use in some
For FEC, the options are to provide no coding, convolution coding or turbo coding.
Again, the decision of which to use is made at channel establishment. Before passing
through the coder, all of the transport blocks that have been sent in the de¬ned transmission
time interval (TTI, explained shortly) are concatenated together serially and coding is
performed on this block. Should the resulting block be too large for the coder, it is
segmented into appropriate sizes. The choice of which coding scheme to use for the

TrCh 1 TrCh 2
Transport Transport
Block Block

CRC Attachment CRC Attachment

Convolution/ Convolution/
Turbo Coding Turbo Coding

Interleaving Interleaving
Physical Layer Procedures

Radio Frame Radio Frame
Segmentation Segmentation

Rate Matching Rate Matching

Channel Mux
Coding and
PhCh 1
Modulation and

Figure 6.19 UMTS physical layer procedures

Table 6.7 Transport channel coding schemes
Transport channel Coding Rate
PCH Convolution coding 2
DCH, FACH, DSCH Convolution coding ,
or CPCH Turbo coding 3
No coding n/a

channel is dependent on the level of error protection required. Table 6.7 shows the coding
schemes available for different transport channels.
The resultant block of data then undergoes an interleaving process, as was described
in Chapter 2.

6.8.3 Radio frame segmentation and rate matching
Depending on the data to be transmitted it will be split or concatenated into a num-
ber of blocks, which may consist of more than one 10 ms radio frame. These blocks
are passed between the physical and the upper layers at certain time intervals. These
blocks are referred to as transport blocks (TB). For example, a speech call may pass a
transport block to the physical layer every 20 ms. This timing is de¬ned when a con-
nection is established and is known as the transmission time interval (TTI). For the
previous example, the TTI is 20 ms. Where the TTI for the transport channel is more
than the 10 ms radio frame size, the block must now be segmented into 10 ms radio
frames. Rate matching is implemented to ensure that the coded transport block size is
then brought to the same size as is expected by the spreading process. That is, the bit
rate after the rate matching should equate to the channel bit rate. For example, at a
spreading factor (SF) of 128, in the uplink the bit rate is 30 kbps. The rate matching
process must bring the number of bits in each radio frame to 300. The rate matching
achieves this by either bit repetition, if less than the required number, or by bit punc-
turing of some of the coding, should the size be larger. In Figure 6.20, it is assumed
that a transport block of 244 bits is passed to the physical layer every 20 ms. For this
block, the required error protection is an 8-bit CRC plus 1/2 rate convolution coding. At
the physical layer, the CRC is attached and this block passed through the convolution
coder. This results in a block of 504 bits. This is now rate matched by bit repetition to
become 600 bits, and is then segmented into two 300-bit blocks. These are then spread


. 60
( 132 .)