- 期刊
- OpenAccess
Chinese Two-Part Codebook Generation System
中文來去密語本創生系統
摘要
由於中文書寫結構不屬拼音系統，而常用之中文字亦超過五千字以上，故中文秘密通訊系統有賴密語本之大量使用。又基於安全之理由，密語本須定期予于以更換，故發展電腦化之中文密語本創生系統有其必要性。在本文中，吾人運用隨機變量之加倍權順序統計量組計─具有彈性、快速、便利、經濟特性之密語本創生系統，俾滿足軍事、企業、貿易、行政單位秘密通訊之需求。首先，吾人運用加權之均等分佈順序統計量來解決同音賽佛（Homophonic Cipher）之問題，並經評估後選用經過改良之指數創生法來解決密碼碰撞（code-collision）的問題以降低其複雜度並增進其成功之機率。最後，則運用不歸還隨機抽樣法作密碼之指派。
關鍵字
無資料並列摘要
It is true that the popular Chinese system of writing is not a phonetic system and the number of Chinese characters. Tzu, most commonly used is over 5000. This is the main reason why Chinese cryptosystem heavily depends on the use of codebooks.Since the codebook should be updated periodically for the sake of security, it is worthwhile to develop a computerized Chinese codebook generation system. In this study, the idea of weighted order statistics of random variates is adopted to design a codebook generation system which can be used flexibly, securely, rapidly, friendly, and economically to satisfy the requirements of the classified information communications among military, business, trading and administration units.In this study, the weighted order statistics of the uniformly distributed random variates is adopted for Homophonic Cipher purpose. The refined exponential method is chosen for eliminating the code collision with small complexity and high probability of success. Then the random sampling without replacement is used as the method of code-assignment.