Co-EDR: Cooperative Event-Driven Fuzzing for Efficient Vulnerability Detection in Firmware

With the Internet of Things (IoT) becoming increasingly integrated into everyday life and industry, ensuring efficient and secure data processing in large-scale IoT networks is critical. Typically, vulnerabilities in wireless firmware are patched via over-the-air updates. However, timely patching remains a challenge. Efficient fuzzing techniques have been proposed to address this problem, but conventional fuzzing techniques often rely on random input generation, resulting in poor code coverage, high overhead, and limited detection performance because of the exhaustive injection of all possible input combinations. To address these limitations, this study proposes a cooperative event data recording (Co-EDR) technique to efficiently verify firmware stability in wireless communication systems. Co-EDR improves fuzzing efficiency by recording event data that previously triggered crashes and reusing these data as inputs. In addition, it uses cooperative fuzzing across multiple devices to improve code coverage and vulnerability detection. Experimental results indicate that Co-EDR results in an 8.52-fold reduction in the time required to achieve the same vulnerability detection rate as conventional random fuzzing, using approximately 2.47 times less memory and enhances the fuzzing efficiency by up to 32.5 times. These improvements enable expedited vulnerability detection and timely firmware patching with reduced resource expenditure compared with conventional techniques, rendering the proposed method a practical solution for improving the security of IoT systems.