A Latency-Aware Secure Data Access Model for Cloud-Native Web and Mobile Platforms

Abstract

Cloud-native web and mobile applications require highly secure yet ultra-low-latency data access to ensure optimal user experience (UX) and transaction speed. Traditional security models often introduce significant latency overheads due to complex, synchronous authentication and authorization checks.¹ This paper proposes a Latency-Aware Secure Data Access Model (LASDAM) designed to decouple robust security enforcement from the critical path of data retrieval and modification. The model employs a hybrid authorization strategy, combining lightweight, decentralized token-based authorization for high-volume read operations with stricter, synchronous Policy-as-Code (PaC) enforcement for sensitive write operations. Key architectural elements include a security proxy layer for request interception and a tiered caching mechanism for policy and access tokens. The empirical evaluation, conducted under varying network conditions and high-throughput scenarios, demonstrated that LASDAM achieved a 95th percentile (P95) read latency reduction of up to 45% compared to a fully synchronous security model, while maintaining a $100\%$ security efficacy rate against common data access violations. This demonstrates a viable path to integrating high-security standards without sacrificing the performance required by modern consumer platforms.