A generalised error model
The previous chapters have treated the influence of different radiofrequency (RF) impairments in multiple-antenna OFDM systems, independently. In this chapter a first step is made towards the modelling of the composite severeness and impact of different RF impairments. Such a generalised impairment model has the major advantage that not all effects have to be investigated separately, but that the combined impact of all impairments can be captured. This combined influence of all impairments is also the measure that is of interest to system designers.
One commonly used measure for the aggregate severity of imperfections in system design is the error vector magnitude(EVM) , which basically measures the root mean square (rms) error in the estimated symbols. In order to unambiguously relate this measure to the final system performance measure, i.e., the error rate of the detection, the distributionof the error term resulting from all implementation deficiencies has to be known. Generally, though, it is assumed to be a zero-mean complex Gaussian process. It was shown in the previous chapters, however, that when the impairments induced by the RF part of the system prevail, this will not be a good approximation, making a performance estimation using EVM far from accurate. Furthermore, the traditional EVM measure does not allow to take the nonstationarity of the impairments into account, which will be present when systems apply power-save modes and switch between transmitter/receiver (TX/RX) modes.
KeywordsMicrowave Attenuation Covariance Settling
Unable to display preview. Download preview PDF.