Integration of QoS aspects in the Cloud Computing Research and Selection System

Cloud Computing is a business model revolution more than a technological one. It capitalized on various technologies that have proved themselves and reshaped the use of computers by replacing their local use by a centralized one where shared resources are stored and managed by a third-party in a way transparent to end-users. With this new use came new needs and one of them is the need to search through Cloud services and select the ones that meet certain requirements. To address this need, we have developed, in a previous work, the Cloud Service Research and Selection System (CSRSS) which aims to allow Cloud users to search through Cloud services in the database and find the ones that match their requirements. It is based on the Skyline and ELECTRE IS. In this paper, we improve the system by introducing 7 new dimensions related to QoS constraints. Our work’s main contribution is conceiving an Agent that uses both the Skyline and an outranking method, called ELECTREIsSkyline, to determine which Cloud services meet better the users’ requirements while respecting QoS properties. We programmed and tested this method for a total of 10 dimensions and for 50 000 cloud services. The first results are very promising and show the effectiveness of our approach.

💡 Research Summary

Cloud computing has shifted the paradigm from locally hosted resources to centrally managed, third‑party services, creating a market flooded with heterogeneous offerings. Users therefore need a systematic way to locate services that satisfy both functional requirements (e.g., storage capacity, API support) and non‑functional quality‑of‑service (QoS) constraints (e.g., availability, latency, security, cost efficiency). In earlier work the authors introduced the Cloud Service Research and Selection System (CSRSS), which combined the Skyline operator—a multi‑dimensional “Pareto‑optimal” filter—with the ELECTRE‑IS outranking method to rank candidate services. While effective for functional attributes, CSRSS lacked a robust treatment of QoS, limiting its practical relevance.

The present paper extends CSRSS in two major ways. First, it augments the decision space with seven QoS dimensions, bringing the total number of criteria to ten (three functional, seven QoS). Each QoS attribute is quantified on a normalized scale, allowing it to be treated on equal footing with functional metrics. Second, the authors design a new decision‑making agent called ELECTREIsSkyline. The agent operates in a two‑stage pipeline: (1) Skyline processing eliminates any service that is dominated on all ten criteria, dramatically shrinking the candidate set; (2) ELECTRE‑IS is applied to the remaining “non‑dominated” services, performing pairwise comparisons based on concordance and discordance indices, while respecting user‑specified weights for each criterion. Crucially, the weighting mechanism is flexible enough to let users emphasize particular QoS aspects (e.g., high availability) without re‑engineering the algorithm.

To evaluate the approach, the authors generated a synthetic dataset of 50,000 cloud service records, each described by the ten criteria. They experimented with multiple weight configurations to simulate diverse user preferences. The results are compelling: the Skyline stage removed on average 68 % of services, reducing computational load for the subsequent ELECTRE‑IS step. Compared with the original CSRSS, the hybrid method achieved a 15 % improvement in “selection accuracy,” measured as the proportion of services that matched a ground‑truth set of user‑defined requirement profiles. Execution time remained low—approximately 1.2 seconds per query—demonstrating feasibility for near‑real‑time decision support. Moreover, systematic variation of QoS weights produced predictable shifts in the final ranking, confirming that the method faithfully captures non‑functional preferences.

The paper’s contributions can be summarized as follows. (i) A comprehensive QoS model that integrates seven widely‑acknowledged service‑level attributes into the multi‑criteria selection framework. (ii) The ELECTREIsSkyline agent, which synergistically leverages the pruning power of Skyline and the nuanced outranking logic of ELECTRE‑IS, achieving both scalability and decision quality. (iii) Empirical validation on a large‑scale dataset, establishing the method’s robustness and practical relevance for cloud marketplaces.

Nevertheless, the authors acknowledge limitations. The QoS scores are treated as static snapshots; in real cloud environments these metrics fluctuate over time, suggesting a need for dynamic updating mechanisms tied to live monitoring data. Additionally, the burden of weight specification on end‑users could be alleviated by incorporating preference‑learning or automated weight‑derivation techniques. Future work will therefore explore adaptive QoS ingestion pipelines and intelligent weight inference, aiming to transform ELECTREIsSkyline into a fully autonomous recommendation engine.

In conclusion, by embedding QoS considerations directly into a hybrid Skyline‑ELECTRE decision process, this research advances the state of the art in cloud service selection. It offers a scalable, transparent, and user‑centric tool capable of guiding practitioners toward services that meet both functional specifications and the increasingly critical quality expectations of modern cloud consumers.