A study of speech scramblers using the transmultiplexer technique
Mitsuhiro Azuma, Takayuki Hasebe, Naoya Torii, Ryota Akiyama · Electronics and Communications in Japan (Part III Fundamental Electronic Science) · 1991
Abstract This paper considers a scrambling technique for an analog voice signal, and proposes a band‐split scrambling method applying the transmultiplexer technique, which can effectively divide the voice band using a small amount of signal processing. The principles of signal processing and the filter design are presented. By computer simulation, the quality of the decoded speech, the effect of the channel characteristics and security are evaluated. As a result of quantitative evaluation of the quality of the decoded speech, the relation between the quality of the decoded speech and the number of taps is quantitatively evaluated for fixed and variable scramblings. It is seen that the quality of the voice signal is little affected by the channel characteristics, such as group delay distortion, synchronization error and frequency off‐set. Thus, the proposed method is seen to be economical, without requiring additional circuitry as used by various kinds of cancellers. As for security, an intelligibility test using numbers was executed, and high security, comparable to or better than the conventional method, is obtained. Thus, the proposed secure communication method is shown to be effective, with excellent performance in the quality of the scrambled decoded speech economy and security.