In the first simulation scenario the quality of the voice
traffic is evaluated (FER) as a function of Circuit
Switched (CS, i.e. voice) traffic. In Figure 2, it is
shown that when the voice traffic increases, the
quality of the speech gets worse, especially when the
Circuit Switched (CS) traffic is very high. When
assuming FER * 3% as a limit for bad quality for a
cOnnection and when allowing no more than 2% of
bad quality for all the C~M~~~~OIIS, then the maximum
CS trdlic that can be offered to the system is
approximately 26 Erl. When the amount of bad quality
samples of FER >= 1% is set to 2% then the
maximum CS traffic load becomes 16.5 Erl.