A fog-enabled smart home solution for decision-making using smart objects

https://doi.org/10.1016/j.future.2019.09.045Get rights and content

Abstract

The development of new smart objects for the sensing and actuation of a given place or environment led both the academia and industry to research and propose new protocols and intelligent systems to support such objects. One of the systems that has been gaining prominence is the smart residential environments. In this context, homes are equipped with smart objects to manage the living resources. However, managing such objects in residential environments requires data contextualization, i.e. collecting data from heterogeneous devices and actuate on the environment through context information generated from such data. To solve this problem, we propose an intelligent decision system based on the fog computing paradigm, which provides an efficient management of residential applications. The proposed solution is evaluated both in simulated and real environments. When compared with other studies from the literature in a simulated environment, the proposed solution shows a higher success rate with a lower delay in the decision-making process, higher efficiency in information dissemination with a lower overhead in the communication infrastructure, and increased robustness in processing with a lower power consumption. These results are also observed when considering a real environment evaluation.

Introduction

Smart cities are becoming increasingly present in the administration and governance of large metropolises. The advances in the development of smart objects [1] and its services [2] play an important role in the realization of smart cities. In this scenario, the goal is to merge information from objects and intelligent services to efficiently manage the resources and services from a metropolis [3], [4]. Many researchers are interested in investigating smart cities in different application domains, such as smart home, domotics, e-health, energy efficiency, and assisted environment.

Given this context, Home Automation Systems (HAS) stands out as a smart home equipped with smart and/or specialized objects to manage the living resources of the residence [5], [6], [7]. HAS is one of the application domains gaining prominence in smart cities studies for being a promising option to solve one of the major global challenges: energy efficiency. Smart objects in HAS, such as smart televisions, set-top boxes, routers, smart meters, and smartphones, can be used to monitor and control the environment in which they are deployed. Such objects can process data and connect to the Internet, creating a fog computing environment. Moreover, they can intercommunicate to improve the information quality in the decision-making process [8].

The fog computing paradigm [9], [10] is a promising alternative to assist smart objects with restricted computational resources, such as the ones employed in HAS. Furthermore, the fog computing paradigm can meet requirements that are not fulfilled by a centralized model. Such paradigm extends the computational resources available in the cloud infrastructure to the network edge, providing mobility, scalability, low latency, and robustness in the services provided to the users  [9], [10], [11]. Also, it enables real time information analyses through the distribution of the decision-making process to the network edges [9], [12].

However, information contextualization in HAS environments still is a great challenge. In particular, establishing a correlation between the qualitative and quantitative data obtained from devices to perform the decision-making process (i.e., generating viable information) within the communication infrastructure itself creates new research questions, among which two stand out: (i) How to use the data disseminated in an implicit, raw and incomprehensible (i.e. environment data without context) from smart objects manner to improve the decision-making process?; and (ii) How to overcome the absence of communication interoperability and robustness when performing the decision-making process on devices with fewer resources? In addition, the lack of a flexible infrastructure to deal with the roles of each smart objects, with a low overhead in processing and that meets the needs of residents (for example, maximize their energy efficiency) are some of the problems addressed in the literature, which this study also investigates. However, it is worth noticing that, the solutions found on the literature comprise specific, local problems and fragmented investigation, indicating a promising and unconsolidated field of research.

With this in mind, this article goes further than the solutions presented by [13], [14] and proposes ImPeRIum, an intelligent decision-making system that creates a fog computing environment to manage home applications. The decision-making process in ImPeRIum is performed through computational intelligence techniques. In our previous work [15], a preliminary study on the use of ensemble classifiers to enhance precision in the decision-making process was made. However, the ensemble classifiers were not integrated into the fog computational environment, hence the results obtained were just exploratory. The fog computing paradigm was used to integrate a greater intelligence into the control infrastructure of the house. Also, it is possible to process content locally on the infrastructure and reduce time in decision making. Such infrastructure is responsible for collecting and disseminating data on the environment and then detecting and acting on the desired applications. To disseminate the data in ImPeRIum, we implemented a communication module based on the Publish/Subscribe (Pub/Sub) paradigm, which also allows the connection of new devices and deals with the interoperability problem among them. To reduce the processing overhead of the infrastructure, the features of ImPeRIum are distributed among the network nodes. The results of an extensive evaluation, considering different scenarios and parameters, show the viability and efficiency of ImPeRIum when compared with other solutions from the literature.

The remainder of this article is organized as follows. Section 2 presents related work, discussing the major challenges for this research. Section 3 describes the development of our solution, and Section 4 shows how the solution was validated. Finally, Section 5 presents the conclusions and future works.

Section snippets

Related work

This section presents the main challenges and open problems in the HAS domain that will be investigated in this article, and also some existing solutions that attempt to solve these problems. For instance, until the time of this research, it was not found a solution that uses the smart devices of a residence to create a fog computing environment that manages the decision-making process of the resident’s applications.

Currently, the potential use of wireless network technologies as an internal

A neural-fog control system for residential infrastructures

This section introduces ImPeRIum, a smart decision system with distributed computational resources that creates a fog computing environment to manage residential applications. The development of ImPeRIuM was based on computational intelligence techniques to perform the decision-making process, as well as on a number of smart devices found in a residence (such as set-top boxes, smart televisions, smart meters and routers) to create a fog computing environment. To handle the data heterogeneity

Performance evaluation and methodology

This section presents the results of the performance assessment and the methodology used to generate the results. The validation of ImPeRIum was divided into two stages. The first stage comprised a validation of ImPeRIum (using ANN-MLP and the Ensemble) by comparing it with two approaches from the literature, CONDE and ResiDI. These approaches were chosen due to their resemblance with this research, as already discussed in Section 2. The second stage evaluated the resource management of

Conclusion and future research

This article proposed ImPeRIum, a smart approach to manage residential environments. By considering the profile of residents, ImPeRIum monitors and acts on the residential environment through the intercommunication of smart objects. In case of an event that is out of the expected, ImPeRIum can detect it and act on it accordingly to keep the environment in a pre-established state. An extensive assessment considering different scenarios showed the viability and efficiency of ImPeRIum in smart

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Geraldo P. Rocha Filho ( [email protected]) is an Assistant Professor at the Department of Computer Science (CIC) at University of Brasília (UnB). He received his Ph.D. in Computer Science from the University of São Paulo (USP) in 2018. He received his M.Sc. from the USP in 2014. He was also a post-doctoral research fellow at the Institute of Computing at UNICAMP before joining the UnB. His research interests are wireless sensor networks, vehicular networks, smart grids, smart home and machine

References (25)

  • LeeJ.H. et al.

    Context-aware inference in ubiquitous residential environments

    Comput. Ind.

    (2014)
  • MahmoodA. et al.

    Home appliances coordination scheme for energy management (hacs4em) using wireless sensor networks in smart grids

  • B.P. Santos, L. Silva, C. Celes, J.B. Borges, B.S.P. Neto, M.A.M. Vieira, L.F.M. Vieira, O.N. Goussevskaia, A....
  • MaM. et al.

    Cityguard: A watchdog for safety-aware conflict detection in smart cities

  • ZanellaA. et al.

    Internet of things for smart cities

    IEEE Internet Things J.

    (2014)
  • DameriR.P.

    Using ICT in smart city

  • ChengJ. et al.

    A Survey on Smart Home Networking

    (2009)
  • StojkoskaB.L.R. et al.

    A review of internet of things for smart home: Challenges and solutions

    J. Cleaner Prod.

    (2017)
  • P. Maia, A. Baffa, E. Cavalcante, F.C. Delicato, T. Batista, P.F. Pires, Uma Plataforma de Middleware para Integração...
  • BonomiF. et al.

    Fog computing: A platform for internet of things and analytics

  • MahmudR. et al.

    Fog computing: A taxonomy, survey and future directions

  • A.A.T.R. Coutinho, E.O. Carneiro, F.G.P. Greve, Computação em Névoa: Conceitos, Aplicações e Desafios, 2016, pp....
  • Cited by (23)

    • Where do all my smart home data go? Context-aware data generation and forwarding for edge-based microservices over shared IoT infrastructure

      2022, Future Generation Computer Systems
      Citation Excerpt :

      A large variety of applications are now able to address various socio-economical issues by processing the sensory data at the very edge of the network. In literature, such applications covering various domains such as, industry 4.0 [18,19], building automation [20], smart cities [17], and healthcare [21] greatly benefit from edge-based computation by adapting to various dynamic events [22,23] or gaining operational and performance enhancements. To meet the challenges associated with data-driven application-specific requirements, the existing research has focused on different aspects of edge-based IoT data management.

    • A hardware and environment-agnostic smart home architecture with containerized on-the-fly service offloading

      2021, Computers and Electrical Engineering
      Citation Excerpt :

      Although Cloud computing provides an effective way of storing, processing, and analyzing large amounts of data, it cannot be a standalone solution for delay-sensitive and data streaming services. As a result, Fog computing enabled IoT platforms are currently gaining more popularity [9]. In [10], authors implemented a smart home energy management system using IoT and Big Data concepts by enabling the remote control of home appliances via a single gateway device.

    • A Methodological Review on EEG Data Reduction in Edge/Fog computing-based IoMT networks

      2023, 2023 14th International Conference on Computing Communication and Networking Technologies, ICCCNT 2023
    View all citing articles on Scopus

    Geraldo P. Rocha Filho ( [email protected]) is an Assistant Professor at the Department of Computer Science (CIC) at University of Brasília (UnB). He received his Ph.D. in Computer Science from the University of São Paulo (USP) in 2018. He received his M.Sc. from the USP in 2014. He was also a post-doctoral research fellow at the Institute of Computing at UNICAMP before joining the UnB. His research interests are wireless sensor networks, vehicular networks, smart grids, smart home and machine learning

    Rodolfo I. Meneguette is an Assistant Professor at Federal Technology Institute. He received his Bachelor’s degree in Computer Science from the University of São Paulo, Brazil, in 2006. He received his master’s degree in 2009. He received his doctorate from the University of Campinas (Unicamp), Brazil, in 2013. He did his post-doctorate in the PARADISE Research Laboratory, University of Ottawa, Canada, in 2017. His research interest are in the areas of vehicular networks, resources management, flow of mobility and vehicular clouds.

    Guilherme Maia is an Assistant Professor of Computer Science at UFMG, Brazil. He got his Ph.D. in Computer Science at this same university in 2013. His research interests include distributed algorithms, mobile computing, wireless sensor networks and vehicular ad hoc networks.

    Gustavo Pessin received his D.Sc. degree (2013) in Computer Science from the University of São Paulo. In 2015, he had a position as a Visiting Scholar at the Massachusetts Institute of Technology. He is currently a Research Associate at the Vale Institute of Technology. He has interest in the area of machine learning, data science, and wireless sensor networks.

    Vinícius P. Gonçalves is PhD in Computer Science and Computational Mathematics (2016) from the University of São Paulo (USP). He was also a research fellow at the University of Arizona (USA) before joining the University of Brasília (UnB). He was a Postdoctorate Researcher Fellow at School of Medicine of USP, as a CAPES Fellowship. Currently, Vinícius P. Gonçalves is an Adjunct Professor at the Institute of Technology at the UnB, Brasília, Brazil. His main research interests are: Human–Computer Interaction, Internet of Things, Cyber–Physical Systems and Mobile Health.

    Li Weigang is a professor and chair of the Department of Computer Science at the University of Brasília (UnB), Brazil. He received his Ph.D. from the Aeronautics Institute of Technology (ITA), Brazil, in 1994. He is a researcher with grant from Brazilian National Council for Scientific and Technological Development (CNPq). He coordinated various research projects from CAPES, CNPq, FINEP, FAPESP and FAPDF and the industry projects with Atech and Boeing Company/Brazil. His research interests include artificial intelligence with emphasis on computation model in air traffic management and data analytics.

    Jó Ueyama is a Full Professor of the Institute of Mathematics and Computer Science (ICMC) at the University of São Paulo (USP). Prof. Ueyama is also a Brazilian Research Council (CNPq) fellow. He completed his Ph.D. in computer science at the University of Lancaster (England) in 2006. Before joining USP, he was a research fellow at the University of Kent at Canterbury (England). Jó has published 50 journal articles and 100 conference papers. His main research interest includes Computer Networks, Distributed Systems and Blockchain. Finally, Prof. Ueyama is the Head of the International Office Affairs at ICMC/USP since August 2016.

    Leandro A. Villas is an Associate Professor and Chair of the Institute of Computing at the University of Campinas (Unicamp), Brazil. Leandro has published 50+ papers in international journals and 120+ in conferences. Six of those papers received the best paper award. Moreover, he received the Latin America Young Professional Award, IEEE Communications Society, and the Excellence Award from the Institute of Computing at Unicamp.

    View full text