Adaptive and large-scale service composition based on deep reinforcement learning

Wang, Hongbing; Gu, Mingzhu; Yu, Qi; Tao, Yong; Li, Jiajie; Fei, Huanhuan; Jia, Yan; Zhao, Wei; Hong, Tianjing

doi:10.1016/j.knosys.2019.05.020

Cited by 43 publications

(16 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Wang et al [16] proposed a novel service composition system depending upon Deep Reinforcement Learning (DRL) for adaptive and large-scale service composition. The projected method is preferable for partial observable service platforms, which makes it better work for real time scenarios.…”

Section: Related Workmentioning

confidence: 99%

Hybrid Metaheuristics Web Service Composition Model for QoS Aware Services

Rajeswari¹,

Jayashree²

2022

Computer Systems Science and Engineering

View full text Add to dashboard Cite

Recent advancements in cloud computing (CC) technologies signified that several distinct web services are presently developed and exist at the cloud data centre. Currently, web service composition gains maximum attention among researchers due to its significance in real-time applications. Quality of Service (QoS) aware service composition concerned regarding the election of candidate services with the maximization of the whole QoS. But these models have failed to handle the uncertainties of QoS. The resulting QoS of composite service identified by the clients become unstable and subject to risks of failing composition by end-users. On the other hand, trip planning is an essential technique in supporting digital map services. It aims to determine a set of location based services (LBS) which cover all client intended activities quantified in the query. But the available web service composition solutions do not consider the complicated spatio-temporal features. For resolving this issue, this study develops a new hybridization of the firefly optimization algorithm with fuzzy logic based web service composition model (F3L-WSCM) in a cloud environment for location awareness. The presented F3L-WSCM model involves a discovery module which enables the client to provide a query related to trip planning such as flight booking, hotels, car rentals, etc. At the next stage, the firefly algorithm is applied to generate composition plans to minimize the number of composition plans. Followed by, the fuzzy subtractive clustering (FSC) will select the best composition plan from the available composite plans. Besides, the presented F3L-WSCM model involves four input QoS parameters namely service cost, service availability, service response time, and user rating. An extensive experimental analysis takes place on CloudSim tool and exhibit the superior performance of the presented F3L-WSCM model in terms of accuracy, execution time, and efficiency.

show abstract

Section: Related Workmentioning

confidence: 99%

Hybrid Metaheuristics Web Service Composition Model for QoS Aware Services

Rajeswari¹,

Jayashree²

2022

Computer Systems Science and Engineering

View full text Add to dashboard Cite

show abstract

“…The K-means clustering technique is used for grouping the web services with similar Quality of Service (QoS). To finding the best answer to the QoS-based service composition problem, some methods can be used, such as Integer linear programming (ILP) techniques, called an optimization algorithm (Liang et al, 2019;Lodi & Nagarajan, 2019;Wang et al, 2019). But some heuristic methods like a genetic algorithm (GA) can be used, known as an approximation algorithm (Dwiardhika & Tachibana, 2019).…”

Section: Related Workmentioning

confidence: 99%

“…An integer linear programming (ILP) technique can be applied to answer the QoS-based service composition problem, called the restriction optimization problem (Liang et al, 2019;Lodi & Nagarajan, 2019;Wang et al, 2019). Then, each ILP solver can be appropriated for this plan; however, several variables in this model belong to multiple service competitors, it can only be carried out efficiently for meager examples.…”

Section: The Maximum Generation Maxgenerationmentioning

confidence: 99%

QoS-aware service composition based on context-free grammar and skyline in service function chaining using genetic algorithm

Khosravian

Emadi

Mirjalily

et al. 2021

PeerJ Computer Science

View full text Add to dashboard Cite

Service function chaining (SFC) is a mechanism that allows service providers to combine various service functions and exploit the available virtual infrastructure. The best selection of virtual services in the network is essential for meeting user requirements and constraints. This paper proposes a novel approach to generate the optimal composition of the service functions. To this end, a genetic algorithm based on context-free grammar (CFG) that adheres to the Internet Engineering Task Force (IETF) standard and Skyline was developed to use in SFC. The IETF uses cases of the data center, security, and mobile network filtered out the invalid service chains, which resulted in reduced search space. The proposed genetic algorithm found the Skyline service chain instance with the highest quality. The genetic operations were defined to ensure that the service function chains generated in the algorithm process were standard. The experimental results showed that the proposed service composition method outperformed the other methods regarding the quality of service (QoS), running time, and time complexity metrics. Ultimately, the proposed CFG could be generalized to other SFC use cases.

show abstract

“…Moreover, machine learning can not only be combined with heuristic algorithms but can also be used to construct neural network models to solve VRPs. Wang et al [30] built neural network models for TSP and then trained neural network models and adjusted the parameters iteratively. Experiments verified that the solution quality of the well-trained neural network was better than that of state-of-the-art results of learning algorithms.…”

Section: Related Workmentioning

confidence: 99%

A Pointer Neural Network for the Vehicle Routing Problem with Task Priority and Limited Resources

Sheng

Wei

2020

ITC

View full text Add to dashboard Cite

The vehicle routing problem with task priority and limited resources (VRPTPLR) is a generalized version of the vehicle routing problem (VRP) with multiple task priorities and insufficient vehicle capacities. The objective of this problem is to maximize the total benefits. Compared to the traditional mathematical analysis methods, the pointer neural network proposed in this paper continuously learns the mapping relationship between input nodes and output decision schemes based on the actual distribution conditions. In addition, a global attention mechanism is adopted in the neural network to improve the convergence rate and results. To verify the effectiveness of the method, we model the VRPTPLR and compare the results with those of a genetic algorithm. The parameter sensitivity of each algorithm is assessed using different datasets. Then, comparison experiments with the two algorithms employing optimal parameter configurations are performed for the validation sets, which are generated at different instance scales. It is found that the solution time of the pointer neural network is much shorter than that of the genetic algorithm and that the proposed method provides better solutions for large-scale instances.

show abstract

Adaptive and large-scale service composition based on deep reinforcement learning

Cited by 43 publications

References 17 publications

Hybrid Metaheuristics Web Service Composition Model for QoS Aware Services

Hybrid Metaheuristics Web Service Composition Model for QoS Aware Services

QoS-aware service composition based on context-free grammar and skyline in service function chaining using genetic algorithm

A Pointer Neural Network for the Vehicle Routing Problem with Task Priority and Limited Resources

Contact Info

Product

Resources

About