Antennas

Most present deployment objectives can be met by using conventional passive antennas
with a modest amount of reconfigurability—some electrical downtilt, for example.
Smart adaptive antenna schemes may become more economically attractive in the
future because of the need to selectively null out interference from in-band users.
Many operators will need antennas that can service the 1900/2100 MHz band, the
1800 band, and the 900 band. The half-wave/quarter-wave relationship between 900 and
1800 MHz makes this feasible although rather suboptimum in terms of antenna performance.
As offered traffic moves increasingly toward a more complex rich media product
mix, then radio link/radio bandwidth quality will become increasingly important.
Present networks are being designed to support a bit error rate of either 1 in 103 or 1
in 106. Given that wireline bit error rates are typically 1 in 1010, it could be argued that
radio bandwidth quality should be equivalent. Reducing the bit error rate from 1 in 103
to 1in 106 with the same delay parameters requires a 3 dB increase in the radio link
budget. Decreasing from 1 in 106 to 1 in 109 requires an additional 3 dB.
Smart antennas provide one way in which the link budget can be improved. (An
alternative, of course, is just to increase the existing network density using conventional
antennas.) Flexible antenna configurations—that is, antennas that can adapt themselves
as interference moves in the cell—are a logical way to improve capacity and coverage.
The cost economics for widespread deployment, however, are as yet unproven.