5. Answer the questions:
What provides the general features for frequency translation?
What is “carrier compression”?
What is carrier & signal suppression?
When is it desirable that the “virtual” carrier difference over a network is maintained to within 2 Hz?
When is correction applied?
PART 3 (20.7 )
20.7 Noise contributions
Noise is the largest source of degradation to analogue systems.
Two contributions have to be considered, thermal and unlinearity noise. To evaluate the noise performance of an equipment both noise source contributions are calculated in pWOp and summated.
20.7.1 Definitions
There are three power levels in common use, dBr, dBm and dBmO.
The OdBr point is the level at a reference point within the system that all the transmission levels refer. This used to be the 2 wire audio point or the 2 wire point of origin, as it was sometimes called.
This physical entity within a network has all but disappeared and a point of reference is now normally taken as the virtual outgoing switch point and is set at -4dBr. From the transmissions point of view the audio output from the channel translating equipment is adjusted to suit the required stated dBr level at that point and this becomes the level to which all the transmission levels within the system refer. Transmission levels are referred to in dBr.
The power at a various points in the transmission are referred to in dBm. If a test signal is injected into a OdBr point at OdBm then the test signal level throughout the transmission can be referred to in dBm.
It is sometimes the case that for noise calculations where the transmission level is required in dBm rather than dBr it is assumed that the power at the OdBr point is OdBm.
This is the power of a signal in dBm referred to a point where the transmission level is OdBm.
For instance if a signal power is -80dBm at a -30dBm transmission level point, then the signal is defined as -50dBmO.
20.7.2 Psophometric weighting
When noise, either thermal or unlinear noise, is added to a telephone conversation, the degree of annoyance or the effect on the intelligibility of the conversation is not the same for all frequencies of added noise.
A weighting curve has been constructed to characterise this effect. Various curves exist (CCITT , Bell System C message etc.) all giving weighting factors of 2 to 2.5dB. The UK adopt the CCITT curve in Figure 20.9 providing a weighting factor of 2.5dB. This factor allows that for a 4kHz channel, considered over the speech bandwidth of 3.1kHz, 2.5dB more Gaussian type noise can be tolerated from the system when weighting is applied.
When this correction is applied the suffix 'p' is added to the noise power figure (i.e. dBmOp). An additional weighting factor is included when the full channel width of 4kHz is considered.
The weighting factor for a 4kHz channel considered over the full 4kHz is 2.5 + 10 log(4/3.1) or 3.6dB.
20.7.3 Thermal noise
This is sometimes called Browian or Gaussian noise and is noise that is associated with the random movement of electrons at temperatures above absolute zero.
For the FDM range of frequencies the noise power that is generated is given by Equation 20.3, where P is noise power in watts, k is Boltzmann's constant, T is the absolute temperature in degrees Kelvin, and B is bandwidth in Hz.
If T = 310°K and B = 4000Hz then the available noise power is -137.5dBm.
The absolute thermal noise power at the output of a circuit element of gain GdB and noise figure FdB within a 4kHz channel is given by Equation 20.4.
If the absolute power at the transmission level point T is known (TL in dBm) then the thermal noise contribution to the noise power in a 4kHz channel can be calculated in dBmOp (or pWattOp) as in expressions 20.5 and 20.6, where 3.6dB is the weighting factor for a 4kHz bandwidth. Definitions of the noise factor F are found in Connor, 1973.
