摘要
传统软件单元安全性测试方法忽略了对威胁的分类,导致软件异常提取精度较低。为保证软件单元运行的稳定性,提出基于改进遗传算法的软件单元安全性测试方法。划分系统软件与应用软件威胁种类,提取异常特征。通过二进制编码完成软件安全性度量编码。利用有限个体表示空间特征优化,将空间内最优个体作为初始种群。在传统交叉变异算法基础上添加活力种群,规定一种新的自适应变化规则,抑制种群变坏。提出并行机理,确定个体取值范围的上、下约束条件,提高算法局部寻优能力。利用四叉树表示软件特征,构建软件单元安全测试模型。实验结果验证了上述方法对软件安全性可疑点的探测更加全面,且探测错误率较低。
The traditional software unit security testing method ignores the classification of threat level,resulting in low accuracy of software anomaly extraction.A software unit security testing method based on improved genetic algorithm is presented for improving the stability of software unit operation.The threat types of system software and application software were divided,and the abnormal features were extracted.Based on binary coding,software security metrics coding was achieved.Meanwhile,the finite individuals were utilized to represent the spatial feature optimization,and the optimal individuals in the space were taken as the initial population.Based on the traditional crossover mutation algorithm,the active population was added,a new adaptive change rule was provided to restrain the population deterioration.The parallel mechanism was put forward to determine the upper and lower constraints of individual value range,thus the local optimization ability of the algorithm was improved.The quad-tree was utilized to represent software characteristics and build a software unit security test model.The simulation results show that this method has low detection error rate and comprehensive detectability for detecting suspicious points of software security.
作者
王超
庞先伟
WANG Chao;PANG Xian-wei(Southwestern University of Finance and Economics,Sichuan Chengdu 610074,China)
出处
《计算机仿真》
北大核心
2021年第1期265-268,330,共5页
Computer Simulation
关键词
改进遗传算法
并行激励
活力种群
软件单元
安全性测试
Improved genetic algorithm
Parallel incentive
Dynamic population
Software unit
Security testing
