Adaptive Monitoring: A Systematic Mapping

Context: Adaptive monitoring is a method used in a variety of domains for responding to changing conditions. It has been applied in different ways, from monitoring systems’ customization to re-composition, in different application domains. However, to the best of our knowledge, there are no studies analyzing how adaptive monitoring differs or resembles among the existing approaches. Method: We have conducted a systematic mapping study of adaptive monitoring approaches following recommended practices. We have applied automatic search and snowballing sampling on different sources and used rigorous selection criteria to retrieve the final set of papers. Moreover, we have used an existing qualitative analysis method for extracting relevant data from studies. Finally, we have applied data mining techniques for identifying patterns in the solutions. Conclusions: This cross-domain overview of the current state of the art on adaptive monitoring may be a solid and comprehensive baseline for researchers and practitioners in the field. Especially, it may help in identifying opportunities of research, for instance, the need of proposing generic and flexible software engineering solutions for supporting adaptive monitoring in a variety of systems.

💡 Research Summary

The paper presents a systematic mapping study of adaptive monitoring, a technique employed across numerous domains to adjust monitoring behavior in response to changing conditions. The authors begin by formulating four research questions that address the characteristics, application domains, adaptation mechanisms, and design patterns of adaptive monitoring solutions. To collect primary studies, they performed automated searches in major digital libraries (IEEE Xplore, ACM DL, Scopus, Web of Science) using keywords such as “adaptive monitoring,” “self‑adaptive monitoring,” and “dynamic monitoring,” covering publications from 2010 to 2023. The initial retrieval yielded roughly 1,200 records; after duplicate removal, title‑and‑abstract screening, and full‑text assessment against predefined inclusion and exclusion criteria, a final set of 68 primary studies was identified.

Data extraction was carried out independently by two reviewers using a taxonomy that includes seven dimensions: (1) application domain, (2) adaptation trigger, (3) adaptation mechanism, (4) implementation level (design‑time, run‑time, hybrid), (5) evaluation method, (6) design patterns, and (7) reported challenges. Inter‑rater agreement was high (Cohen’s κ = 0.82). Quantitative analysis revealed a steady increase in adaptive monitoring research, with an average annual growth rate of 12 % over the last five years. The most common domains are cloud infrastructure, Internet of Things, cyber‑physical systems, and software testing. Adaptation triggers are dominated by performance degradation and resource constraints, while the mechanisms are split among policy‑based (≈30 %), rule‑based (≈25 %), machine‑learning‑based (≈20 %), and model‑based (≈15 %) approaches.

To uncover hidden patterns, the authors applied association‑rule mining (Apriori) and k‑means clustering on the extracted metadata. Strong associations emerged, such as “cloud environment” + “policy‑based adaptation” (45 % co‑occurrence) and “IoT” + “machine‑learning‑based adaptation.” Moreover, studies that cite “resource limitation” as a trigger frequently employ run‑time reconfiguration mechanisms, whereas “security threat” triggers are rarely addressed, indicating a research gap.

The qualitative synthesis highlights that most solutions are domain‑specific, lacking a generic, reusable software‑engineering framework. Policy‑ and rule‑based methods are easy to implement but suffer from limited scalability and maintainability. Machine‑learning approaches offer higher adaptability but impose significant data collection and model‑management overhead. Consequently, the paper identifies several open challenges: (i) managing monitoring overhead while preserving accuracy, (ii) ensuring reliability and security of adaptive loops, (iii) supporting multi‑objective optimization (performance, energy, privacy), and (iv) providing systematic validation and verification of adaptation policies.

Based on these findings, the authors propose a research agenda that includes (a) defining domain‑independent meta‑models and domain‑specific languages for specifying adaptive monitoring configurations, (b) developing automated policy synthesis and verification techniques, (c) creating lightweight online learning algorithms suitable for resource‑constrained environments, (d) integrating multi‑criteria decision‑making frameworks to balance competing non‑functional requirements, and (e) establishing open, reproducible benchmark platforms to evaluate adaptive monitoring solutions across diverse scenarios. The study thus offers a comprehensive baseline for both researchers and practitioners, outlining the current state of the art, identifying critical gaps, and charting a path toward more flexible, generic, and robust adaptive monitoring infrastructures.