DNS Fuzz Testing Technique Based on Multi-Indicator Fusion

With the continuous expansion of the Internet scale, the domain name system, as the key infrastructure of the Internet, is of great significance in terms of its operational stability and security. Existing fuzz testing techniques have a single evaluation index for test cases, and there is a bottleneck in the efficiency of test case generation, which is difficult to meet the needs of the increasingly complex DNS security testing. In view of this, we designed a fuzz testing method based on the fusion of multiple indicators for domain name system software testing. On the basis of AFLGO, we first introduce update region, vulnerability update region, new region, three new test case evaluation criteria, and then combined with the multi-indicator fusion algorithm to construct an energy allocation model, which accurately allocates the energy to different test cases, so that the test cases are more targeted to cover the key basic blocks of the domain name system. Experiments show that our designed method outperforms traditional methods in generating use cases efficiently and provides a more efficient fuzz testing technique for software testing of domain name system.