Development of a Secure and Licensed Mobile Framework for Real-Time Physiological Data Analysis

For companies active in the domain of developing data-driven and proprietary algorithms, a big challenge is always how to share or sell their solution to other companies without exposing the internal logic. This problem is more highlighted, especially in the use cases where the data processing should be done on the edge and in real time, even in remote areas, eliminating the possibility of using a server-side logic. The solution presented in this work provides a reliable way to securely share the physiological prediction algorithms for third-party mobile applications, in the domain of sleep, fatigue, and Alcohol misuse. It consists of a Swift XCFramework for iOS and an Android AAR library, each providing a clear and consistent public interface while keeping the logic secure and unreachable. The solution works independently of the source of health data and provides the result in a defined manner without exposing the core logic. To enhance the security and prevent misuse of the library or framework, a comprehensive offline licensing system is implemented to generate licenses to be shared with clients. Each license indicates a list of features available in it, the bundle ID or package name of the app it can run on, and the expiry date. To further increase the security of the solution, security checks for jailbroken or rooted devices are added, in addition to the detection of reverse engineering attempts.