送受同形エンファイスの時間域スクランブラへの適用

Abstract

The isomorphic emphasis has previously reported on spectrum inversion to prevent frequency resources from exhaustion. In this paper, we discuss about that the isomorphic emphasis has also succeeded in preventing speech quality from degradation during time domain scrambling without any expansion of frequency occupancy bandwidth over radio security speech channels. The isomorphic emphasis is again realized with the canonical form, which consists of cascaded a differentiator, a time domain scrambler with arbitrary sequential pattern, and a integrator. As well known, PM transmission systems being featured of f^<-2> fading noise shape, this time domain scrambler degrades consequentially speech quality by changing power ratio of speech signal to fading noise over subjective frequency domain. However, the isomorphic emphasis is shown to prevent time scrambled signal from both expansion and degradation over all of the subjective frequency domain, even if there exists discontinuity among input data according to time scrambling. These facilities of the isomorphic emphasis in the time domain scrambling for speech security are verified from theoretical analysis and computer simulations over poor fading channels.