Radio Planning

With existing TDMA systems it has been relatively simple to derive base station and
handset sensitivity. The interference is effectively steady-state. Coverage and capacity
constraints can be described in terms of grade of service and are the product of network
density.

In IMT2000, planning has to take into account noise rise within a (shared) 5 MHz
channel, an allowance for fast power control headroom at the edge of the cell and soft
handover gain. The interference margin is typically 1 to 3 dB for coverage-limited conditions
and more for capacity-limited networks. Fast power control headroom is typically
between 2 and 5 dB for slow-moving handsets. Soft handover gain—effectively
uplink and downlink diversity gain—is typically between 2 and 3 dB. There are four
power classes. Class 1 and 2 are for mobiles. Class 3 and 4 are for handsets (mobiles
would, for example, be vehicle-mounted). These are shown in Table 11.11. Typical maximum
power available at a Node B would be 5, 10, 15, 20, or 40 Watts.
In 3GPP, Eb/No targets are set that are intended to equate with the required service
level. (As mentioned in earlier chapters, Eb/No is the energy per bit over the noise floor.
It takes into account the channel coding predetermined by the service to be provided.)
The Eb/No for 144 kbps real-time data is 1.5 dB. The Eb/No for 12.2 kbps voice is 5 dB.
Why does Eb/No reduce as bit rate increases? Well, as bit rate increases, the control
overhead (a fixed 15 kbps overhead) reduces as a percentage of the overall channel
rate. In addition, because more power is allocated to the DPCCH (the physical control
channel), the channel estimation improves. However, as the bit rate increases, the
spreading gain reduces.
IMT2000 planning is sensitive to both the volume of offered traffic and the required
service properties of the traffic—the data rate, the bit error rate, the latency, and servicedependent
processing gain (expressed as the required Eb/No).
System performance is also dependent on system implementation—how well the
RAKE receiver adapts to highly variable delay spreads on the channel, how well fast
fading power control is implemented, how well soft/softer handover is configured,
and interleaving gain.
Downlink capacity can also be determined by OVSF code limitations (including
nonorthogonality) and downlink code power. The power of the transmitter is effectively
distributed among users in the code domain. 10 W, for example, gets distributed
among a certain amount of code channels—the number of code channels available
determines the number of users that can be supported. On the uplink, each user has his
or her own PA, so this limitation does not apply.
A Node B will be exposed to intracell and intercell interference. Intracell interference
is the interference created by the mobiles within the cell and is shown.

Interference from adjacent cells initially exhibits a first-order increase. At a certain
point, handsets start increasing their power to combat noise rise, which in turn
increases noise rise! A first-order effect becomes a second-order effect; the cell has
reached pole capacity. The rule of thumb is that a 50 percent cell load (50 percent pole
capacity) will result in 3 dB of intracell interference. A75 percent cell load implies 6 dB
of intracell interference.
In other words, say you have a microcell with one transceiver. It will have a higher
data rate handling capability than a macrocell with one transceiver because the microcell
will not see so much interference as the macrocell.