A MIMO WLAN based on linear channel inversion

Multiple-input multiple-output (MIMO) techniques may increase the capacity of wireless local area networks (WLAN) since they profit from rich scattering of radio waves in indoor environments. On the other hand, much signal processing must be performed in a very short time to reach practical data rates. This paper reports on elementary steps towards MIMO transmission when the channel is known to the transmitter. In most cases, channel information can be obtained at the transmitter by means of reciprocity. Linear channel inversion (LCI) may then be employed allowing simple processing suitable for real-time applications. In contrast to signal processing at the receiver, LCI offers isotropic noise gain which simplifies coding. On the other hand, power distribution at the transmitter depends on the statistical properties of the MIMO channel. Results in the case of Rayleigh fading indicate that at least one additional antenna is needed at the transmitter for reliable operation. A simple formula for the mean transmitter power is reported and confirmed by simulation. Performance can now be expressed in terms of transmitter power and a fair comparison with transmission based on signal processing at the receiver is possible.