**Zeroforcing algorithm** or Interference nulling algorithm, refers to a form of linear processing algorithm used in communication systems which nulls out interference signal in the time domain or inverts the frequency response of the channel in the frequency domain. An earlier work for this form of signal processing is described in ^{[1]}.

## Algorithm Edit

If the channel response for a particular channel is $ \mathbf{h} $, the interference nulling algorithm multiplies the reciprocal of $ \mathbf{h} $ to the input signal which removes the effect of channel from the output signal. If zeroforcing weight vector is $ \mathbf{w} $, the relationship with the channel is

- $ \mathbf{w}^H \mathbf{h} = 0 $

The effect of channel can be the interference signal, the intersymbol interference (ISI) or etc.

The interference nulling algorithm removes all interference and is ideal when the channel is noiseless. However, when the channel is noisy, the zero-forcing algorithm will amplify the noise depending on the channel maganitude at the point of frequencies $ f $. A more balanced linear processing in this case is the minimum mean-square error processing, which does not usually eliminate interference signal completely but instead minimizes the total power of the noise and interference components in the output.

## External Links Edit

## References Edit

- ↑ Winters, J.H, "The Diversity Gain of Transmit Diversity in Wireless Systems with Rayleigh Fading," IEEE Trans. on Vehicular Technology, Feb. 1998.