The residual BER - error floor, though useful and widely used metrics for the end-to-end digital radio transmission performance, provides no insight into the error-generating analog impairments (e.g. modulation inaccuracy, power amplifier compression, carrier recovery phase error, or I-Q cross-talk), which, however, are easy to identify by VSA tools such as constellation analysis. As the EVM analysis has become very popular figure of merit in this regard, in this paper, after reviewing crucial advantages of EVM analysis with respect to what we get from BER, we consider estimating EVM from the residual BER assuming that the latter's generally non-linear and non-additive “constituents” are substituted by the equivalent AWGN source producing the same BER (and EVM) degradation. The resulting EVM(BER) curves were verified in the LTE lab to be a very good first approximation of EVM from available residual BER, when (expensive) VSA tools are not available.
[1]
Hüseyin Arslan,et al.
Error vector magnitude to SNR conversion for nondata-aided receivers
,
2009,
IEEE Transactions on Wireless Communications.
[2]
Adriana Lipovac.
OFDM Error Floor Prediction in a Small-Time-Dispersion Channel
,
2014,
Wirel. Pers. Commun..
[3]
Adriana Lipovac,et al.
Verification of OFDM Error Floor Prediction in Time-Dispersive LTE FDD DL Channel
,
2017,
Wirel. Pers. Commun..
[4]
Lajos Hanzo,et al.
Single- and Multi-carrier Quadrature Amplitude Modulation : Principles and Applications for Personal Communications, WLANs and Broadcasting
,
2000
.