A novel scalable representative-based forecasting approach of service quality

Yahyaoui, Hamdi; Own, Hala S.; Agwa, Ahmed M.; Maamar, Zakaria

doi:10.1007/s00607-020-00802-z

Cited by 5 publications

(2 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In [36], researchers presented a Markov Chain-based monitoring model that took into account memory, CPU, and storage. Other researchers developed a novel and flexible representation-based forecasting method for predicting QoS response time and throughput based on ANN [37]. For an evaluation of the Weight Service Rank Approach (W-SR), paper [38] evaluated accountability, agility, cost, performance, assurance, security, and usability.…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

A Survey on Improving QoS in Service Level Agreement for Cloud Computing Environment

Edinat

Al-Sayyed

Hudaib

2021

Int. J. Interact. Mob. Technol.

View full text Add to dashboard Cite

Cloud computing is considered one of the most important techniques in the field of distributed computing which contributes to maintain increased scalability and flexibility in computer processing. This is achieved because it, using the Internet, provides different resources and shared services with minimum costs. Cloud service providers (CSPs) offer many different services to their customers, where the customers’ needs are met seeking the highest levels of quality at the lowest considerate prices. The relationship between CSPs and customers must be determined in a formal agreement, and to ensure how the QoS between them will be fulfilled, a clear Service Level Agreement (SLA) must be called for. Several previously-proposed models used in the literature to improve the QoS in the SLA for cloud computing and to face many of the challenges in the SLA are reviewed in this paper. We also addressed the challenges that are related to the violations of SLAs, and how overcoming them will enhance customers’ satisfaction. Furthermore, we proposed a model based on Deep Reinforcement Learning (DRL) and an enhanced DRL agent (EDRLA). In this model, and by optimizing the learning process in EDRLA, proposed agents would be able to have optimal CSPs by improving the learning process in EDRLA. This improvement will be reflected in the agent's performance and considerably affect it, especially in identifying cloud computing requirements based on the QoS metrics.

show abstract

Section: Related Workmentioning

confidence: 99%