COMPARATIVE ANALYSIS OF STATE MANAGEMENT MODELS IN MOBILE APPLICATIONS DEVELOPED USING FLUTTER

Abstract

The objective of this study is to formalize state management models in Flutter applications and evaluate them against criteria of transition determinism, performance, scalability, and architectural complexity. State management approaches are interpreted as extended transition systems with asynchronous events; safety and liveness properties are formalized using temporal logic. A formalized Flutter state management model is proposed as a class of reactive computational systems with defined safety and liveness properties. In addition, the study incorporates benchmark-based analysis of frame rendering time and CPU utilization to provide an empirical comparison of the considered approaches. The obtained results are further interpreted in the context of real-world development scenarios, enabling the formulation of practical recommendations for selecting appropriate state management architectures depending on application scale and complexity. The study revealed significant differences among approaches in determinism, performance, and architectural complexity, demonstrating that Riverpod provides the most balanced performance characteristics due to efficient dependency tracking, while Bloc and Cubit ensure higher structural clarity at the cost of increased computational overhead. MobX shows moderate performance with advantages in managing complex interdependent state, whereas Provider remains suitable for small-scale applications due to its simplicity. These findings enable informed architecture selection based on production application requirements, including system scale, performance constraints, and maintainability needs.