基于二进制补丁比对的软件输入数据自动构造

Automatic software input data construction based on evolve algorithm

在软件二进制补丁比对结果的基础上,提出一种基于目标的适应值函数模型。该模型对同一种群中所有个体评估,寻找优秀个体以遗传到子代中,产生能够到达或接近到达可疑函数的个体数据。该算法不需要源码。实验结果表明,该算法简化了逆向分析人员构造软件畸形数据的过程,为其快速触发软件脆弱点创造了条件。

Based on the result of software patch difference,a fitness function model based on the target is put forward. The function model is used to evaluate all of the individuals in the same population,to find and pass excellent individuals down to next generation and to generate individuals which can or nearly can get to suspicious function. This algorithm does not require source code. Experimental result shows that the algorithm simplifies the process that reverses analyzer constructs abnormal software input data and offers the opportunity to quickly trigger the software vulnerability.