| Summary: | Error-correcting codes are used in communication systems to provide reliable transmission
on practical communication channels and thus reduce the number of retransmissions.
The ability of a particular code to be able to detect and correct errors at the
receiver depends on the individual code structure as well as on the decoding algorithm.
Codes that provide strong error correcting capabilities usually involve a number of
operations on the information bits and hence are not easily decodable at the receiver.
Rather than relying on the use of a single high complexity code, a concatenation of
two or more simple codes can be used. Serially concatenated encoding structures are
very popular means to achieving close to capacity performance. Good performance of
concatenated codes with practically viable decoding times is attributed to a technique
known as iterative decoding. Iterative decoding can only be used in a situation where
the component decoders are capable of generating soft information for the transmitted
data. Soft information for symbol-by-symbol estimation is usually obtained by using
the Bahl Cocke Jelinek Raviv (BCJR) algorithm or some sub-optimal version of it.
Differential encoding of the data at the transmitter is regarded as an effective approach
to combat phase ambiguities at the receiver, when using phase shift keying (PSK)
schemes. Since the information is transmitted in difference of phases rather than the
absolute phase of the transmitted symbol, there is more protection against an unknown
channel phase at the receiver. The serial concatenation of an error correcting code with
a differential encoder has been found to provide very good performance for various kinds
of channel conditions.
In this work we propose and analyse the design of a serially concatenated structure
which is very simple to implement and is particularly favourable for a low power scenario
such as deep space applications. The proposed system comprises of a concatenation
of simple parity check codes and a differential encoder (DE). Individually, these
codes are very weak codes as they provide minimal error-correcting capabilities. We
optimise system design parameters through extensive analysis of the system structure
using extrinsic information transfer (EXIT) charts. It is shown through simulations
and analytical results that the proposed concatenated codes provide performance very
close to capacity. Comparison of these simple parity check codes with certain other
very powerful outer codes such as Low Density Parity Check (LDPC) codes show the
superior performance of the proposed codes inspite of much lower decoding complexity.
For the case, where channel phase is unknown or perfect synchronisation is not attainable
at the receiver, several estimation algorithms have been proposed in the literature
to combat the effects of channel phase. These algorithms can usually be divided into
two categories. The first is those that use pilot symbols, which increases the transmission
overhead. The second is to not have any explicit channel estimation mechanism.
Here we adopt the second one and consider two approaches. One is based on the quantization
of the unknown phase and is computationally intensive with the complexity
depending on number of levels of quantization. The other is to do a blind estimation
based on the information derived from received symbols. On the lines of the second
approach we propose a simple method for noncoherent decoding that uses a posteriori
information to estimate the channel phase. It is shown that the method works well for
the cases of low to moderate variations in channel phase. === Applied Science, Faculty of === Electrical and Computer Engineering, Department of === Graduate
|
|---|
