Identification of T-codes with minimal average synchronisation delay

When T-codes are used to encode an information source most efficiently, the ir tendency towards self-synchronisation is among the strongest. The best T -codes achieve self-synchronisation within 1.5 characters following a lock loss. Attempts have been made to build databases of the best T-codes, These rely on the comparison of average synchronisation delay (ASD) values that have to be determined using complex algorithms. Criteria are proposed for s electing T-code sets that have intrinsically low ASD values. These criteria are derived from the underlying synchronisation process of T-codes based o n the T-code construction algorithm. This latest contribution means that T- codes that are optimal, in terms of both coding efficiency and synchronisat ion performance, can be found for any given information source more easily compared to previous methods.