Distributed Antennas for In-Building Coverage

As we move from 2G to 3G technologies, base station form factor (at least temporarily)
increases, the need to deliver more linearity increases base station Node B hardware
footprint. At the same time, the minimum bandwidth available from a Node B transceiver
is 5 MHz, compared to the minimum bandwidth of 200 kHz (an eight-slot, eightchannel
single RF carrier mini GSM base station).

For in-building coverage we need small base stations and, often, not a lot of bandwidth.
Abase station in a small hotel foyer does not need 5 MHz of RF bandwidth. This
makes distributed antennas quite attractive, certainly in the early stages of network
deployment.
The idea of distributed antennas is to have a donor base station, say, in the basement
of a large building. The RF signal is then distributed to a number of antennas mounted
throughout the building. The problem with distributed antenna solutions is that losses
in copper cable can be quite substantial.
One option is to use RF over fiber. The RF signal is converted to an optical signal
using a linear laser and is then delivered down a fiber-optic cable. We cover RF over
fiber in Chapter 13 (“Network Hardware Optimization”).
Summary
In this chapter we reviewed some of the important system design considerations
implicit in implementing a 3G network with a 3G radio physical layer. We have said
that the radio physical layer directly influences network performance, and we address
this in more detail in future chapters.
We discussed some of the design and performance parameters of the Node B. We
said that physical size (form factor) is driven by the ever-decreasing size and volume
of 2G base stations and that a particular design challenge is to deliver the additional
linearity needed in 3G hardware within a sufficiently compact, lightweight product
footprint.
Node B hardware determines how much offered traffic can be supported and how
the offered traffic will be accommodated in terms of cell sectorization. We introduced
some of the radio layer enhancements that are available, such as downtilt antennas,
and highlighted the differences between handset RF design and Node B design and
some of the options for implementing Node B hardware (RF/IF and baseband processing).
We emphasized that the RF performance of the Node B (code orthogonality
on the downlink and receive sensitivity on the uplink) directly influences radio system
planning.
In addition, we reviewed some of the lessons learned from system planning in 1G
and 2G cellular networks and pointed out that initial coverage and capacity simulations
are often overoptimistic. The additional number of variables in CDMA planning
make it harder to pin down likely system performance.
We also reviewed some of the present simulations reviewed in the present planning
literature and advised some caution in how the present figures should be interpreted.
We pointed out that not only Node B RF performance but also handset RF performance
is a major component of the RF link budget and that both Node B and handset RF performance
increase as the network technology matures (particularly if market volume is
achieved—the performance advantage of volume). A1 dB improvement in Node B on
handset sensitivity translates into a 10 percent decrease in network density. 281