A resilient and distributed near real-time traffic forecasting application for Fog computing environments

Pérez, Juan L.; Gutierrez-Torre, Alberto; Berral, Josep Lluís; Carrera, David

doi:10.1016/j.future.2018.05.013

Cited by 36 publications

(28 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…There are several parameters to evaluate these three aspects; for example, network quality is one of the reliability assessment parameters. Pérez et al 111 proposed that by establishing a strong connection between two FNs, the exchange of data packets and QoS can be guaranteed. Increasing the density of nodes in areas where there are many IoT devices creates a strong connection between FNs, which can be achieved with good latency and low traffic overhead, and the probability of reconnection decreases.…”

Section: Discussion and Comparisonmentioning

confidence: 99%

Task scheduling approaches in fog computing: A systematic review

Alizadeh

Khajehvand

Rahmani

et al. 2020

Int J Communication

View full text Add to dashboard Cite

SummaryThe Internet of Things (IoT) interconnects billions of physical objects to collect and exchange information and makes available various applications. Despite all the advantages of the IoT, some of its applications are not feasible because of the existing restrictions in the IoT sensors. To overcome these restrictions, cloud computing has been developed in recent years. Although cloud computing has removed some of these restrictions, the cloud itself has faced some other challenges. One of these challenges is the distance between the cloud and the end‐devices, which is not ideal for delay‐sensitive applications and also leads to high communication costs and security problems. To handle these challenges, the fog computing has been introduced in literature, which places the resources and services at the edge of the network, close to the end‐devices. On the other hand, fog nodes suffer from heterogeneity, uncertainty, and dispersion of resources, and also processing, storage, and memory limitations. Thus, there is need for a proper task scheduling approach to use these resources optimally. In this study, the task scheduling algorithms proposed by different researchers for the cloud‐fog environment, their advantages and disadvantages, and also various tools and issues regarding the scheduling methods and their restrictions were discussed. The results indicated that about 58% of the scheduling algorithms use static scheduling and that the other 42% use dynamic scheduling, and the delay metric is the most noteworthy parameter considered in most studies with a share of nearly 17%. Finally, we identified the open issues related to the task scheduling based on cloud‐fog computing and provided some directions for future works.

show abstract

Section: Discussion and Comparisonmentioning

confidence: 99%

Task scheduling approaches in fog computing: A systematic review

Alizadeh

Khajehvand

Rahmani

et al. 2020

Int J Communication

View full text Add to dashboard Cite

show abstract

“…To identify the algorithm of big data analytics in smart city, some researchers [7], [8], [9], [10], [11], [22] have expressed their opinions. The algorithm of big data analytics in smart city are :…”

Section: To Verify the Performance Of Classification Trained Algorithmentioning

confidence: 99%

“…5. Data distribution algorithm has result fog nodes are resistant to problems of back-haul connectivity and could transmit data to the cloud location event with problems of severe connectivity [11].…”

Section: Artificial Neural Network (Ann) Andmentioning

confidence: 99%

“…To identify tools of big data analytics in smart city, some researchers have expressed their opinion. The tools of big data analytics in smart city presented in Table V. Among the leading platforms for big data storage, processing, and management include IBM infosphere, Hadoop map-reduce, spark, stratosphere and NoSQL database system management [11], [12], [13], [15], [17], and [18]. Based on Table V, Hadoop is the most widely used tools on big data analytics.…”

Section: Tools Of Big Data Analytics In Smart Citymentioning

confidence: 99%

See 1 more Smart Citation

Big Data Analytics Algorithm, Data Type and Tools in Smart City: A Systematic Literature Review

Putra

Kurniawan

2018

2018 International Conference on Information Technology Systems and Innovation (ICITSI)

View full text Add to dashboard Cite

The smart city generated rapidly huge of data. Data can analyze with big data analytics technology to give solution from past data in smart city problem and help better solution in decision making. This paper summarizes the existing condition of big data analytics in the smart city in term of the algorithm, data type and tools were built using systematic literature review (SLR) as the standard methodology used to solve any problem by tracing the result from the previous research. The problem in SLR called as research question (RQ). To achieve that goal, we define some RQs related to that scope and clarify each question by tracing previous research. The research paper from reputable journal databases such as EEEE Xplore, Scopus, ScienceDirect, and Springerlink. After synthesizing 15 articles, the results are: algorithm data mining like ANN, Markov, graph mining, etc. needs to improve. That algorithm not enough to handle high data volume, high variety and high velocity to store and processes data; main data type have big chance to give the solution in the smart city is social media. That data has the potential to help in decision making in the smart city problem; Hadoop is the top tool to store and analyze data with highperformance, stable, reliable computing for the different type of data. Combination Hadoop with spark give less overhead to handle the high velocity and volume of data.

show abstract

“…In-field retraining procedures are explored in Section IV, where starting from initially trained ML models, they are retrained with augmented datasets that include not-yet-considered patterns, added as soon as they are detected. Strategies for its practical implementation in optical networks are discussed: from typical individual learning, where each agent detects new patterns from their local sources and uses them for retraining, to collaborative learning, where agents spread knowledge among themselves to speed up the learning curve [22]. Because ML training is a hard task and requires large computation capabilities, analysis of distributed and centralized options reveals their pros and cons.…”

Section: Introductionmentioning

confidence: 99%

Learning Life Cycle to Speed Up Autonomic Optical Transmission and Networking Adoption

Velasco¹,

Shariati²,

Boitier

et al. 2019

J. Opt. Commun. Netw.

View full text Add to dashboard Cite

Autonomic optical transmission and networking requires machine learning (ML) models to be trained with large datasets. However, the availability of enough real data to produce accurate ML models is rarely ensured since new optical equipment and techniques are continuously being deployed in the network. One option is to generate data from simulations and lab experiments, but such data could not cover the whole features space and would translate into inaccuracies in the ML models. In this paper, we propose an ML-based algorithm life cycle to facilitate ML deployment in real operator networks. The dataset for ML training can be initially populated based on the results from simulations and lab experiments. Once ML models are generated, ML retraining can be performed after inaccuracies are detected to improve their precision. Illustrative numerical results show the benefits of the proposed learning cycle for general use cases. In addition, two specific use cases are proposed and demonstrated that implement different learning strategies: (i) a two-phase strategy performing out-of-field training using data from simulations and lab experiments with generic equipment, followed by an in-field adaptation to support heterogeneous equipment (the accuracy of this strategy is shown for a use case of failure detection and identification), and (ii) in-field retraining, where ML models are retrained after detecting model inaccuracies. Different approaches are analyzed and evaluated for a use case of autonomic transmission, where results show the significant benefits of collective learning.

show abstract

A resilient and distributed near real-time traffic forecasting application for Fog computing environments

Cited by 36 publications

References 14 publications

Task scheduling approaches in fog computing: A systematic review

Task scheduling approaches in fog computing: A systematic review

Big Data Analytics Algorithm, Data Type and Tools in Smart City: A Systematic Literature Review

Learning Life Cycle to Speed Up Autonomic Optical Transmission and Networking Adoption

Contact Info

Product

Resources

About