why only 10khz-10Mhz is used for CE102 test ? why not other frequency range is used for CE102 test ?
CE tests measure absorbed currents from the grid due to pulse or nonlinear loads with harmonic content, using local AC lines.
This can then radiate this noise to nearby radios causing interference in reception, (e.g. AM, SW bands) Also, with many similar loads on the grid, this may create excessive Neutral currentsresulting in winding thermal issues in Distribution Transformers (DT) designed for 50 or 60 Hz. Thus the CE (Conducted Emissions) test limits these interference issues with limits.
To test this you need a line voltage rated differential coil inductance and grid Cap( called a LISN for Line Impedance Stabilizer network to look like a "Pi" filter. This can raise grid source impedance and shunt ingress noise from grid, so egress (emissions) of current are converted to Voltage (V=IR with R=50) measured in uV but in dBm on Spectrum Analyzer (SA) converted from power using 50 Ohms.
So you need; LISN, preamp(maybe) and SA then compute dBm to uV (@50 Ohms) to chart results. The LISN raises the ource impedance well above 50 Ohms so this is accurate. There are other methods too.
Ref (not exactly as shown and resonant f of LC is well below 10k shown on graph)
This test procedure is used to verify that electromagnetic emissions from the EUT (equip. under test) do not exceed the specified requirements for power input leads, including returns.