A Systematic Review of Cybersecurity Quality Metrics for Decentralized Computational Ecosystems

Rapid advances in blockchain technology, distributed computing, federated learning, and privacy-enhancing methods have collectively given rise to what researchers now call the Decentralized Computational Ecosystem (DCE) an umbrella term covering digital infrastructures in which computation, trust establishment, and data governance are distributed across heterogeneous node networks rather than concentrated within centralised authorities. Despite a decade of accelerating output, the literature on DCE remains markedly fragmented: individual components are studied in isolation, and no prior systematic review has mapped cybersecurity concerns holistically across all five recognised DCE layers. This study attempts to fill that gap through a PRISMA 2020-compliant systematic review, applied with explicit attention to methodological rigour and reproducibility. A structured database search executed across IEEE Xplore, Elsevier ScienceDirect, SpringerLink, MDPI, and the ACM Digital Library between January and March 2025, covering publications from 2015 onward retrieved 147 unique records after de-duplication. Following title/abstract screening and full-text eligibility assessment, 101 peer-reviewed articles were retained for qualitative synthesis. These were organised into five DCE component categories: Distributed Ledger Protocols (DLP), Decentralized Compute Infrastructure (DCI), Privacy-Preserving Computation Protocols (PPCP), Decentralized AI and Data Marketplaces (DAIDM), and Decentralized Threat Detection Systems (DTDS). Thematic analysis of the included studies yielded six cross-cutting cybersecurity quality metrics and one system-level governance metric: threat detection accuracy, privacy protection strength, trust and verification transparency, attack resilience, scalability and performance stability, interoperability and composability, and governance, ethics and compliance. Coverage of these metrics was assessed systematically across all five DCE component layers. Persistent deficits are apparent in governance integration, cross-layer incentive alignment, and unified evaluation frameworks gaps that are characterised with specific design directions in the final sections of this paper.