Review on Requirements Modeling and Analysis for Self-Adaptive Systems: A Ten-Year Perspective

Context: Over the last decade, software researchers and engineers have developed a vast body of methodologies and technologies in requirements engineering for self-adaptive systems. Although existing studies have explored various aspects of this field, no systematic study has been performed on summarizing modeling methods and corresponding requirements activities. Objective: This study summarizes the state-of-the-art research trends, details the modeling methods and corresponding requirements activities, identifies relevant quality attributes and application domains and assesses the quality of each study. Method: We perform a systematic literature review underpinned by a rigorously established and reviewed protocol. To ensure the quality of the study, we choose 21 highly regarded publication venues and 8 popular digital libraries. In addition, we apply text mining to derive search strings and use Kappa coefficient to mitigate disagreements of researchers. Results: We selected 109 papers during the period of 2003-2013 and presented the research distributions over various kinds of factors. We extracted 29 modeling methods which are classified into 8 categories and identified 14 requirements activities which are classified into 4 requirements timelines. We captured 8 concerned software quality attributes based on the ISO 9126 standard and 12 application domains. Conclusion: The frequency of application of modeling methods varies greatly. Enterprise models were more widely used while behavior models were more rigorously evaluated. Requirements-driven runtime adaptation was the most frequently studied requirements activity. Activities at runtime were conveyed with more details. Finally, we draw other conclusions by discussing how well modeling dimensions were considered in these modeling methods and how well assurance dimensions were conveyed in requirements activities.

💡 Research Summary

This paper presents a systematic literature review (SLR) of requirements modeling and analysis for self‑adaptive systems (SAS) covering the decade from 2003 to 2013. The authors followed a rigorously defined protocol: they selected 21 high‑impact publication venues and eight major digital libraries, applied text‑mining techniques to generate comprehensive search strings, and used the Kappa coefficient to resolve disagreements between reviewers, thereby ensuring reproducibility and methodological soundness. After applying inclusion and exclusion criteria, 109 primary studies were retained for detailed analysis.

The review first maps the temporal and venue distribution of the selected papers, revealing a steady increase in publications after 2006 and a concentration in both general software‑engineering journals (e.g., IEEE Transactions on Software Engineering) and SAS‑specific outlets (e.g., ACM Transactions on Autonomous and Adaptive Systems). This trend indicates growing interest in integrating requirements engineering with runtime adaptation.

The core contribution lies in the taxonomy of modeling methods and requirements activities. The authors extracted 29 distinct modeling techniques and grouped them into eight categories: enterprise models, goal models, requirement models, structural models, behavioral models, environmental models, integrated models, and miscellaneous approaches. Enterprise models, which capture business processes and organizational objectives, were the most frequently employed, whereas behavioral models—formal representations of state transitions, events, and execution traces—received the most rigorous empirical evaluation.

Correspondingly, 14 requirements activities were identified and organized along four chronological timelines: (1) requirements elicitation and analysis, (2) requirements design and implementation, (3) requirements verification and validation, and (4) runtime requirements management. The most studied activity is “requirements‑driven runtime adaptation,” reflecting the central SAS promise of autonomously adjusting behavior in response to evolving requirements at execution time. Activities at the runtime stage were described with the greatest level of detail, underscoring the community’s focus on operational assurance.

Quality attributes were mapped to the ISO 9126 standard, yielding eight attributes (functionality, reliability, usability, efficiency, maintainability, portability, security, and compatibility). The majority of the primary studies emphasized functionality and reliability, while security, portability, and compatibility received comparatively little attention. The authors also cataloged twelve application domains—ranging from smart homes and mobile services to cloud infrastructures and Internet‑of‑Things (IoT) environments—highlighting that cloud and IoT contexts are especially fertile grounds for SAS research.

To assess the methodological rigor of each primary study, the authors scored papers based on research design, reproducibility of experiments, and validation techniques. Papers employing behavioral models and conducting controlled experiments achieved the highest quality scores, suggesting that formal modeling facilitates more robust empirical validation.

From the synthesis, several key insights emerge:

1. Modeling Method Adoption Is Uneven – Enterprise models dominate in frequency, but behavioral models dominate in evaluation depth. This suggests a gap between high‑level business modeling and low‑level formal verification that future work should bridge.

2. Runtime Focus – The prominence of requirements‑driven runtime adaptation confirms that the field has moved beyond static requirement specification toward dynamic, closed‑loop adaptation mechanisms.

3. Quality Attribute Imbalance – Non‑functional attributes such as security and portability are under‑represented, indicating an opportunity for research that integrates security‑by‑design or portability concerns into adaptive requirement models.

4. Domain Concentration – Cloud computing and IoT are the primary application arenas, reflecting the need for elasticity and context awareness in these environments.

5. Assurance Dimensions – While many studies address verification at design time, fewer provide systematic assurance for runtime adaptation (e.g., runtime verification, monitoring, and fault tolerance).

The authors conclude by recommending that future research should (a) develop integrated modeling frameworks that simultaneously address multiple dimensions (goal, structural, behavioral, environmental), (b) incorporate under‑explored quality attributes, especially security, into the adaptation loop, and (c) devise standardized evaluation protocols to improve comparability across studies. By offering a comprehensive map of the state‑of‑the‑art, this SLR serves as a valuable reference for researchers and practitioners aiming to design, analyze, and validate self‑adaptive systems with robust, requirement‑centric foundations.