Power control in wireless FDMA systems

As the near-far problem can be so dominant in wireless FDMA operation, it is worth looking briefly at techniques for alleviating the problem. The most effective technique over and above maximizing the filtering and spectral shaping within the modem and improving linearity in the TX/RX subsystems is to attempt to level out the signal power from each user at the receiver site. If each user were able to fully control the output power from its own transmitter, and assuming that it knew the path loss to the receiver, then it could adjust its own power to ensure a fixed minimum (yet sufficient) level from all users at the receiving site.
Determining the path loss is the key problem. Certainly it is possible in a duplex link for the remote user to measure the power received from the base-station site and hence calculate the path loss in the 'downlink'; however, unless the user is stationary and operating on the same frequency for transmit and receive, this does not necessarily translate to the path loss in the 'uplink' direction. For example, the remote user could be receiving in a frequency selective fade, in which case the unit would   overestimate the path loss involved. Alternatively the uplink could be subject to a frequency selective fade and the unit not generate sufficient transmit power. A solution to this problem is to operate a closed loop power control system whereby the base-station unit monitors the signal power from each remote unit and issues commands to increase or decrease TX power accordingly. This can, however, incur a high signalling overhead in the system. It turns out that power control is very critical in CDMA and to a lesser extent TDMA operation, and closed loop power control is common in CDMA applications.