EXFO PSO-200 Optical Modulation Analyzer - Manual (Page 261)

Measurement Definitions
Optical Modulation Analyzer 253
Measurements for Constellation Charts
Error Vector Magnitude
The deviation of an acquired signal sample and its ideal position
1
in the
constellation chart according to the modulation format can be computed
as the magnitude of the corresponding error vector (red), and referred to
as the error vector magnitude (EVM):
where n is the sample index of the N symbol center samples from the
acquired waveform. Consequently, there is one EVM value associated to
each sample (this is what is displayed in the EVM graphs, as explained in
Error Vector Magnitude (EVM) Diagram on page 46).
To provide a compact single number figure of merit, the root mean square
(rms) value of these EVMs is computed:
given in percent to the magnitude of the longest reference vector.
Magnitude Error
The magnitude error (EM) is the deviation in magnitude between the
sampled signal and the ideal reference constellation, as illustrated in the
graph on page 252. It is calculated as:
given in percent to the magnitude of the longest reference vector.
1. For input pattern types 'Random' and 'Repetitive pattern' the ideal position refers to the closest symbol of the ideal
reference constellation. For 'Standard PRBS' and 'User-defined pattern' the ideal position is the expected constellation
symbol associated with each sample, based on your input signal.
EVM n ( ) S
meas
n ( ) S
ref
r n ( )
– =
EVM
rms
1
N
---- EVM n ( )
2
n 1 =
N 
100  %rms   =
EM
rms
1
N
---- S
meas
n ( ) S
ref
r n ( )
–  
2
n 1 =
N 
100  %rms   =