Power Control in WCDMA

In any CDMA system, power control is of critical importance. Because all
users share the same frequency at the same time, it is important that one
user not transmit at such a high power that other users are drowned out. If,
for example, a user near the base station were to transmit at the same power
level as a user at the cell edge, then at the base station, the signal from the
nearby user would be so great that it would completely overpower the signal
from the far-away user. The result is that the signal from the far-away user
would be impossible to recover. This is known as the near-far problem.

To avoid this problem, mechanisms are required whereby the UE can be
instructed to adjust its transmit power up or down so that all transmissions
from all users in the cell arrive at the base station with the same power level.
Not only is power control required to combat the near-far problem, it is also
required to combat the effects of Raleigh fading, where the received signal
can suddenly drop by many decibels as a result of multi-path propagation,
which results in multiple copies of a signal arriving at the receiver out of
phase. Thus, power control is deployed both in the uplink and downlink.
In general, power control in WDCMA uses two main techniques�"openloop
power control and closed-loop power control.With open-loop power control,
the terminal estimates the required transmission power based upon
the signal power received from the base station and information broadcast
from the base station regarding the transmit power from the base station.
Specifically, the base station broadcasts the transmit power used on the
CPICH, and the terminal uses this information in conjunction with the
received power level to determine the power that should be used on the
uplink. In general, however fading in the uplink and fading in the downlink
are unrelated. Consequently, open-loop power control provides only a very
rough estimate of the ideal power that the terminal should use. For this reason,
open-loop power control is used only when the UE is making initial
access on the PRACH or PCPCH. In all other situations, closed-loop power
control is used.

Closed-loop power control means that the receiving entity (the base station
or UE) measures the received Signal-to-Interference Ratio (SIR) and
compares it with a target SIR value. The base station or UE then instructs
the far end to increase the transmitted power if the SIR is too low or
decrease the power if the SIR is too high. Closed-loop power control is also
known as fast power control since it operates at a rate of 1,500 Hz. In other
words, power control commands and changes happen at a rate of 1,500
times per second. This rate is sufficiently fast to overcome path loss changes
and Rayleigh fading effects for all situations except where the UE is travelling
at high speed.

Closed-loop power control commands are sent on physical control channels
that are associated with physical data channels. Recall, for example,
that in the uplink, the DPDCH has an associated DPCCH. Among other
pieces of information, the DPCCH carries transmit power control commands
back to the base station. A power control command is sent in every
slot. Because 15 slots are available for each 10 ms, we have a rate of 1,500
power control commands per second. Each power control command can
instruct the sender to leave the transmitted power unchanged or to
increase or decrease the transmitted power in steps of 1dB, 2dB, or 3dB.
Similarly, for the downlink DPDCH, an associated DPCCH sends power
control instructions to the UE, along with other functions.

There is also another form of power control known as outer-loop power
control, with the primary objective being maintaining the service quality at
the optimum level. In general, the objective of power control is to maintain
the SIR at the receiver at the optimum level�"not too high and not too low.
The target SIR value, however, is a function of the required quality for the service to be supported. If we measure service quality in terms of Frame
Error Rate (FER) on the air interface (as determined by a cyclic redundancy
check (CRC), then the SIR can be considered a function of FER.

The acceptable FER can vary from service to service. Speech service
using the Adaptive Multirate (AMR) coder at 12.2 Kbps, for example, could
support a FER of one percent without noticeable service degradation. A
non-real time data service could support much higher FER rates before
retransmission, allowing retransmission to correct errors. The impact to
such a service is greater delay and a lower overall throughput, but such
impact can be perfectly acceptable for a non-real time service.

A real-time data service, however, may have a far more stringent FER
requirement, perhaps 1
103 or better. Consequently, depending on the
service requirements, the FER may need to vary, which means that the
required SIR may need to vary. This variation in the required SIR is known
as outer loop power control. It uses closed-loop power control to instruct the
sender to vary the transmit power.With outer-loop power control, however,
the reason for the change is due to a new SIR requirement.