Analogue Radio vs. Digital Radio
Traditional radio systems such as AM/FM/LW/SW systems have
always used analogue technology. DAB is the first (to my knowledge) major
digital system designed specifically for digital broadcast radio.
Analogue Radio
Analogue radio transmission consists of transmitting the
actual audio signal modulated onto the RF carrier. Analogue basically means that
the signal can take on any value (within the limits set by the transmitter).
The problem with transmitting analogue audio signals is that because any
noise, interference or self-interference (multipath effect) is added to the
signal at any point then this cannot be removed from the audio signal and this
degrades the audio quality of the signal or causes hiss.
Digital Radio
Digital radio systems such as DAB or digital radio that is
delivered via digital satellite (DSat) or Freeview, addresses the disadvantages that hamper the analogue
transmission systems (although these can usually be overcome by improving your
FM reception by purchasing a better FM aerial and/or relocating the FM aerial). Digital radio systems consist of transmitting
digital waveforms on
the carrier to the receiver. These waveforms are then decoded to binary format
to make up the digital words
that carry the amplitude values of the audio waveform (this is expanded
upon on the MPEG Coding page).
At the radio receiver, the carrier part of the signal is
removed by ‘downconverting’ from radio frequency (RF) to low frequency
(termed 'baseband') which just leaves the digital
signal along with the noise and interference that has been added to the signal
at the transmitter, in the air, and at the receiver itself. The receiver then
decides which symbols were transmitted. In the simple binary case, the digital 0s
and 1s will be transmitted as equal amplitude but opposite in sign. The receiver
will then just look at whether the signal is above or below zero volts and for
example if the voltage is above zero volts then it will decide the bit
transmitted was a 1 and if it is below zero volts then it will decide that a 0
was transmitted. Because of the noise and interference the voltage value at the
decision instant might not be the same sign as the bit that was transmitted and a ‘bit error’ will
be made. The proportion of the bits that are received in error (the bit error
rate, or BER, calculated by dividing the total number of bit errors by the total
number of bits transmitted) is directly related to the received signal power, which is why
installing an external aerial pays off because an external aerial will
invariably receive a higher signal power than an internal aerial.
Unfortunately there are strict limitations to the power level at which
broadcasters are allowed to transmit at. This is a necessary limitation because
otherwise the transmitted signals would cause too much interference in adjacent
frequency bands.
DAB and digital terrestrial TV (DTT) use a more advanced form of digital
modulation than is explained above. Both of these systems use COFDM, which stands
for Coded Orthogonal Frequency Multiplexing.
Error Correction
The main advantage of transmitting radio digitally is
that the bit errors that are made when the receiver chooses the incorrect symbol
can in most cases be corrected. This is achieved by using forward error
correction coding (FEC coding). The FEC encoder adds redundant bits to the
original bitstream so that the receiver
can ‘decode’ the signal and in the vast majority of cases correct any
errors. An example of a very simple error correction code (this is not a
practical code because it is very inefficient and is not very effective at
correcting errors) uses the rule that if a 1 is to be transmitted then it is
repeated 3 times. Then the receiver takes a majority decision so that it decides
that a 1 was transmitted if 2 or 3 ones are received. Therefore if one error is
made out of the three bits transmitted that error can be corrected. This is an
example of a block error correction code because a group of input bits (in this
case one input bit) are transmitted as a larger block of coded bits.
A measure of the redundancy and of the power of an error correction code
is its code rate which is given by the number of input bits to the forward error
correction encoder at the transmitter divided by the number of transmitted bits.
For the example above, one input bit goes into the FEC encoder at a time and three come
out. Therefore its coding rate equals 1/3. The lower the value of the coding
rate the more powerful the code will be at correcting errors and vice versa. For
wireless transmission, block coding is not the primary form of FEC coding,
although it may be used as well. The preferred type of FEC coding for wireless
systems is called convolutional coding and is more powerful than block coding.
Typical code rate values used are 1/3, 1/2, 2/3 and 3/4.
The type of FEC coding used on DAB is a form of convolutional coding where
different parts of the audio bitstream use different code rates. For example,
scale factors are shared between a lot of samples and therefore it is more
important that these scale factors should not be received in error compared to
the individual samples, so a lower code rate (higher protection level) is used
to protect the scale factors than the individual samples.
A similar variable code rate will be applied to the video stream on DTT.
|