ERROR CONTROL IN DIGITAL SPACE COMMUNICATION SYSTEMS

In this paper we introduce the use of replication decoding with linear block codes as a mean for improving the performance and reducing the probability of error in digital space communication systems. Replication decoding is a symbol-by-symbol maximum likelihood decoding based on soft decision. The space communication channels are among the few physical communication channels that can be modeled by an additive Gaussian noise channel model.
Replication decoding will be described, and an upper bound on the decoding error probability for Gaussian noise channels will be derived. The performance of the proposed replication technique, when sequentially implemented, over additive Gaussian noise channel is evaluated by computer simulation. We will show that -che replication decoding scheme improves the performance consi-derably even for poor channels, i.e., low signal to noise ratio channels. Cases of contiuous a priori algebraic values (which corresponds to soft decision) will be considered for performance evaluation.
The Hamming (15,11) code, which is a single error correcting linear cyclic block code, is the code used in the simulation. Probability curves and tables of the simulation results will be presented to demonestrate the performance of the proposed method quantitatively.