International Journal of Distributed and Cloud Computing

1. Heet Nagoriya – Department of Information Technology, G H Patel College of Engineering and Technology (GCET), Charutar Vidya Mandal University, Anand, India.

2. Kush Patel – Department of Information Technology, G H Patel College of Engineering and Technology (GCET), Charutar Vidya Mandal University, Anand, India.

Received
18-Jul-2026
Accepted
04-Aug-2026
Published
15-Oct-2026
Abstract
Choosing the right communication and computation paradigm for distributed systems—including Remote Procedure Calls (RPC), Serverless Functions-as-a-Service (FaaS), Event-Driven Architecture (EDA), WebAssembly at the edge (Wasm) and Web APIs (REST/GraphQL)—is often based on experience and qualitative guidelines rather than a structured evaluation. This paper presents the Workload-Aware Scoring Framework (WASF), a multi-criteria decision model that compares these five paradigms across seven architectural dimensions: latency, scalability, reliability, cost efficiency, consistency, operational complexity and edge affinity. We define five representative workload profiles and calculate weighted composite scores to identify the most suitable architectural choice for each workload. To examine the stability of the framework, we perform a robustness analysis by perturbing the weight assigned to each evaluation dimension. The results show that the recommended architecture remains unchanged for four of the five workload profiles, while the Hybrid Enterprise profile reveals a close trade-off among FaaS, Wasm Edge, and REST/GraphQL. We also introduce analytical cost–latency models together with an effective-throughput-under-failure equation that captures the relationship between reliability and system coupling. Finally, a workload-aware decision tree translates the proposed framework into a practical guide for architects and practitioners. Overall, the framework provides a systematic approach for comparing modern distributed system paradigms and supports more informed architectural decision-making across a wide range of application scenarios.
Locked
Subscribed
Open Access
Locked Content