Digital Signal Processing Reference
In-Depth Information
Apart from its much higher level of redundancy, there is in fact little difference
between dsss and the traditional fec-coding techniques. In case of dsss, the
sequences are controlled by the spreading code. A subset of only two comple-
mentary sequences is allowed at a certain moment, controlled by the alignment
of the low-rate information stream with respect to the high-rate spreading code.
Also a very interesting observation here is that, at receiver side, the decoding
of the received signal is performed before the demodulation. The reason for
this is very simple, but has quite far-reaching consequences: It was already
mentioned that the level of redundancy in the gps system is higher than in
most other wireless links. The power of the gps signal seen at the antenna
connector of a typical gps receiver is in the order of
−
130 dBm. However,
the total integrated noise power that is present in the same frequency band is
no less than
111 dBm, considerably higher than the signal of interest. As
a consequence, the gps signal cannot be 'seen' by the receiver before it is
despread, and it is thus also not possible to use the standard clock recovery
strategies to create a stable time base for the demapper in the receiver.
Also, the coding threshold of most traditional decoder algorithms requires a
bit-energy to noise power density (
E
b
/N
0
) which is larger than unity. They
will generally not be able to operate under the aforementioned circumstances.
Because of this, the decoder of a dsss receiver is commonly implemented as a
despreading correlator. A correlating receiver, however, fails to function for as
long as the correct alignment between the received signal and the despreading
sequence has not been achieved. In fact, finding the correct timing offset of the
despreading sequence is quite similar to achieving synchronization between
the internal states of encoder and decoder in a traditional fec error coding
scheme. The difference is that synchronization in a dsss system is usually not
achieved in a single time frame. In practice, the receiver finds the correct offset
of the spreading code using a more or less intelligent trial-and-error mecha-
nism. The more processing power that is available, the more possible offsets
the receiver can check out in a single time frame. This is of course at the cost
of processing power and energy consumption. As a consequence, it can take
a considerable amount of time before the receiver in a dsss communication
link is able to synchronize to the transmitter. From the moment the receiver is
able to lock on the transmitted signal, the advantages of correlation processing
become visible. In the example of the gps system, the receiver won't even be
able to detect the signal until correct synchronization has been achieved. Only
after successful synchronization, the snr is boosted by more than 43 dB as a
result of which clock recovery and tracking becomes possible.
For moderate-length spreading sequences, the considerable synchronization
delay for the receiver to lock on the transmitter makes the general dsss sys-
tem unsuitable for duplex communication links. This issue becomes especially
−
