Uplink Budget Analysis

In previous cellular systems (1G and 2G) the link budget has been calculated on factors
that are noninteractive and clearly definable (quantifiable) cell by cell. The differences
between theoretical modeling results and practice have been mainly due to the inaccurate
characterization of the propagation terrain and clutter factors. These problems still
exist in 3G network planning but are added to by the interactive factors that we discussed
earlier in this chapter. Link budget calculations will therefore have to take
account of these additional factors.
We may consider some example assumptions:
Case 1
A 12.2 kbps voice service traveling at 120 kmph.
Assumptions:
 A 3 dB intracell interference rise (a cell loading of 50 percent).
 No intercell interference rise.
 Being voice, a soft handover is anticipated; hence, there will be some gain, let’s
assume 3 dB.
 A processing gain of 25 dB (10log3840/12.2).
 An Eb/No target of 5 dB.
 A fast fading margin of 0 dB—fast fading is ineffective at 120 kmph.
Case 2
144 kbps real-time data service traveling at 3 kmph.
Assumptions:
 Again, a 3 dB intracell interference rise (50 percent cell loading).
 A high transmit power is available as the mobile is away from the body. Hence,
there are no body losses.
 An Eb/No target of 1.5 dB—as fast fading power control is effective at 3 kmph.
 Fast Fading Margin of 4 dB (the Eb/No target rises by 4 dB as the mobile moves
to the cell edge).
It remains to be seen how true-to-life these considerations are (and become!).

Noise rise (as seen by the Node B) will be a function of the offered traffic measured
as throughput. The amount, distribution, and burstiness of offered traffic all contribute
to the achieved sensitivity in the Node B transceiver.
On the downlink, the capacity constraint may well be OVSF code-limited or orthogonality
constraints (interrelated code domain effects). An orthogonality factor of 1
means the code streams are perfectly orthogonal. You will see orthogonality figures
quoted in load factor calculations.
The downlink is more load-sensitive than the uplink. For example, one 10 W transmitter
is shared amongst all users. On the uplink, each user has his or her own PA. The
load limitation effectively limits downlink capacity. Increasing Node B power (downlink
power) increases coverage and capacity but only at the cost of reducing the capacity
and coverage available from adjacent cells.
Capacity calculations also need to take into account soft handover overheads. For a
given number of cells, assumptions have to be made on uplink and downlink throughput
and the number of users in soft handover.
Having established some capacity parameters, we need to establish the coverage
available to users defined in terms of the service level being delivered to them. The
coverage probability will be influenced by the mobility of the user—whether they are
walking or driving or riding in a train.
Overall, the lessons learnt from 2G implementation 10 years ago were that early
simulations based on (in GSM’s case) relatively simple assumptions were overoptimistic.
We might expect 3G simulations to be even wider off the mark, given the additional
number of variables introduced into the simulation.
In practice, there will be a number of performance limitations in early 3G deployment
that will make early link budget simulations hard to achieve.
Handset sensitivity generally improves as a technology matures partly because of
device and design optimization and partly because volume product produces performance
benefits (better control of component tolerance).
A similar pattern will probably emerge with IMT2000 handsets. Performance
degrades if the air interface is required to do something it was not designed to do. This
suggests that IMT2000 network performance will begin to settle down and provide
good, consistent video and voice quality by 2005 to 2006.
A much more in-depth analysis of 3G system planning can be found in Radio Network
Planning and Optimization for UMTS, published by Wiley and authored by Jaana
Laiho, Achim Wacker, and Tomas Novosad, (ISBN 0-471-48653-1). Our thanks also to
Mason Communications for their advice on system planning issues, some of which are
referred to in the preceding text. Information on Mason Communications 3G system
planning service is available at www.masoncom.com.