Android Application Reinforcement Method Based on 5G Terminal Device
Yaqiong Guo, Jingnan Chen, Zichao Xu · 2024
This paper presents a novel Android application reinforcement method designed for 5G smartphones, addressing the security vulnerabilities inherent in Android's open-source na-ture. The method involves extracting key functions from Android applications, decompiling them into Smali code, and then trans-forming them into C++ code. This code is virtualized at compile time and repackaged into a new APK file, offering a more efficient reinforcement approach compared to traditional methods. It uses less computational power and storage while meeting the high safety standards of 5G technology. This approach is particularly effective against automated, low-threshold reverse engineering tools, which have challenged previous hardening methods. In our experiments, we analyzed six common Android shell tools on enhanced applications, demonstrating the ineffectiveness of these tools against our method. Our reinforcement overcomes the easy recovery issue of traditional methods and shows strong resistance to general reverse-engineering tools. We evaluated the performance pre- and post-hardening in terms of CPU utilization, size, and runtime memory usage. For instance, CPU utilization post-hardening increased by 0.9% for Gaud Map, 4.8% for WPS, and 1.3 % for public comments. The use of native Layer code and function localization enhances program efficiency, offsetting the performance overhead of fusion, thus maintaining overall performance stability. Our method's robustness against common reverse-engineering tools, with minimal impact on CPU usage, application size, and memory usage, confirms its practicality for deployment in the 5G ecosystem.