When the idea of chaos synchronization is applied to the communication,a major challenge lies in the fact that even very weak channel noise added to the driving signal may corrupt the synchronization state,hence destroy the communication.To overcome this difficulty,chaotic synchronization systems that are robust to the additive channel noise have been designed [1].In these systems,synchronization errors of lower frequency part are taken as the indication of the synchronization degree.In this work we perform symbolic analysis of these systems.The information theoretic way [2,3] for quantifying synchronization is compared with the synchronization errors with emphasis on describing the noise resistance effect.It is found that the former is more helpful for understanding the information transition process.The potential application to chaos-based communication is discussed.