摘要：Interacting neu ral n etworks hav e been calculated an aly tically. At each trainin g step two networks receiv ea commo n rand om inp ut v ector an d learn their mutu al ou tp ut bits. In 200 5, Ch en-C ai propo sed a no velremote u ser au thentication scheme b as ed on interacting neural network learn in g. The current paper, h owev er,d emons trates that Chen-Cai's scheme is vu lnerable to a replay attack an d an in sider attack, an d th en anim proved sch eme was presen ted in o rd er to reso lve such problems
关键词：In teractin g n eu ral n etwork; Authentication; Password; Smart card; Cryptanalysis
