金融交易密码多模态化方法

针对金融交易系统目前单一密码键盘模态客户身份认证存在的安全性、可靠性问题，提出了密码多模态化方法。多模态感知器获取密码码字信息，格式归一化传送所获原始模态信息，按属性前置处理、分类当前模态码字信息，公共单元融合多模态密码。对于确认的M位密码，每位分别采用N种模态输入，密码窃取率为1/(10MCN×MM)。以密码键盘与黑箱子指书为缺省模态的多模态密码输入系统的实用效果表明，所提方法实现了多模态密码码字无序交融输入，M=6时系统的密码窃取率为1/(106C2×66),而密码破解技术门类数及难度则随着模态数量增加而增加。密码输入的安全性及可靠性等多方面性能较现有单模态身份认证工作状态得到大幅提升。

Password multimodality method in financial transactions

In financial transactions, some safety and reliability problems exist in the client authentication system with single keyboard password mode. To solve these problems, the password multimodality method was proposed. Modal sensors accessed the password codeword information and transmitted its normalized result. The formatted code information was pre-processed in property and then classified by its attributes. After that, the multimode passwords were fused through sharing public units. For a certain M bits password, if each bit had N kinds of possible modals, the password theft rate was 1/(10MCN×MM). In this multimode input system, the keyboard password and black box handwriting are the two default modals, and the application results demonstrate that the proposed method realizes the disordered blending input of multimodal password codeword. At M=6, the password theft rate is 1/(106C2×66), and the method types and difficulty of cracking those passwords increase with the number of modals. The password input system's safety, reliability and other performances are significantly better than those with only one modal.