bioMCS 2.0: A distributed, energy-aware fog-based framework for data forwarding in mobile crowdsensing

https://doi.org/10.1016/j.pmcj.2021.101381Get rights and content

Abstract

Mobile crowdsensing (MCS) paradigm enables users equipped with energy-constrained smart devices to participate in sensing and reporting of assigned tasks. To achieve seamless communication as well as effective energy and resource management, we leveraged the fog computing platform to propose a centralized, energy-efficient and robust data collection framework, called bioMCS, based on the topological properties of a biological network called transcriptional regulatory network. However, since MCS platforms may potentially entail a high number of mobile users and massive volumes of data traffic, we extend the current work under the name bioMCS 2.0 to conceive a distributed energy-aware data forwarding mechanism where the fog devices function as task data relay nodes. bioMCS 2.0 combines energy-awareness, abundance of subgraphs (called motifs) in the fog network and proximity to the base station to perform efficient task sensing and forwarding in a dynamic scenario where fog devices are both energy constrained and mobile. It also ensures quality of information by accepting task data from reliable smart devices. Extensive simulation on the map of New York City and realistic mobility models suggests that bioMCS 2.0 exhibits comparable performance in terms of data delivery, latency and energy efficiency in comparison with both random next hop (fog node) selection as well as centralized forwarding technique that rely on global network knowledge.

Introduction

In recent years, the unprecedented growth of urban population and unplanned land usage has led to a lack of sustainability in the urban environment. To address this, smart cities are using Information and Communication Technology (ICT) to develop energy-efficient applications augmented with automated decision making, to support various public services in urban spaces. Smart applications need to sense the physical environment in real-time and process the generated data to inform decision making. In recent times, mobile crowdsensing (MCS) [1] has emerged as an enabler of cost-effective sensing infrastructure. It allows citizens possessing smart devices like smartphones, tablets, wearables, etc. to collect rich sensory information about the surroundings. Such sensing can either be opportunistic (sensing through built-in smartphone sensors or IoT devices) or participatory (sensing by humans) [2], and our work will be based upon the latter.

In participatory sensing, the human users are actively engaged in sensing tasks, generated by MCS platforms. Such tasks are based on health care, environmental and traffic monitoring and management, seeking information about the points of interest in the vicinity, rating a particular place, and so on. The users can decide which task to accept and perform sensing and data transfer actions using smartphone-based applications. The data from multiple users are aggregated by the MCS platforms to build a body of knowledge that supports decision making [3]. However, executing MCS tasks incur cost in terms of energy spent from device batteries and/or data subscription plan (if cellular connectivity is used for data transfer), thereby often discouraging participants over time. Hence, there is a need to propose energy-efficient data transfer mechanism so that MCS-based data acquisition remains sustainable for smart city applications.

In our preliminary work [4], we proposed the use of collaborative sensing among users in close proximity by leveraging energy-efficient device-to-device communication, called Wi-Fi direct [5]. In particular, we used the Autonomous mode of Wi-Fi direct, where one of the mobile devices will be centrally selected as a group owner (GO) and other devices (called peer nodes) will get attached to a group. However, such collaborative sensing entails an inherent centrality and is heavily dependent on the functioning of the group owner. If a GO loses its network connectivity or is compromised by an adversary, the information collected by the corresponding peers are either lost or misappropriated. Thus, to prevent such single-point of failure and cope with large user base and data traffic volume, this work proposes a distributed data collection mechanism which enables peer devices to directly communicate with the MCS platform.

In typical MCS systems, the allocation of sensing and reporting tasks is performed on an individual-basis [6]. Urban areas are densely populated with a few thousand smart device users spread across different regions. Hence, transferring and managing data from users is a bottleneck for both the underlying network (consisting of state-of-the-art Wi-Fi access points, gateways, routers, etc.) as well as MCS platform. Thus, we envision a MCS system that uses fog devices which are mobile across any urban space to facilitate energy-efficient and delay-sensitive data transfer. Typically, microservers like notebooks, laptops, etc. which are both energy-limited and mobile, have been considered as fog devices.

In traditional fog computing architectures, the task data generated by smart devices are delivered to a base station (hosting the MCS platform) in a multi-hop fashion, where the fog devices constitute the hops. In each hop, the nearest fog device forwards the data to another fog, until it is transferred via the gateway fog. This results in greater message delay and keeps multiple arbitrary fog nodes engaged majority of the time causing higher energy dissipation. Additionally, it is imperative to ensure connectivity (later defined as network robustness) of the MCS network in the event when a few of the gateway devices are dead. On the contrary, smart city applications have to be sustainable (in terms of energy efficiency and robustness) and demand quicker turnaround time. Often the fog devices may be located at places remote to the MCS platform and their limited energy is dissipated while communicating the sensed data via wireless communication technologies, such as 3G/4G/LTE, Wi-Fi, etc. Also, it is not always be feasible to replenish their batteries nor replace them with fully-charged devices on-the-fly, making energy efficiency a critical precondition for an efficient data transfer framework for fog-based smart city applications.

In the preliminary version of this paper we proposed bioMCS, a bio-inspired collaborative data transfer framework through mobile crowdsensing over fog computing platforms [4]. This solution was based on the network attributes of a biological network called transcriptional regulatory network (TRN) that represents the interaction between proteins, called transcription factors (TFs), and genes in living organisms [7]. In the past, we harnessed the graph properties of TRN to design wireless networking solutions (see Section 2.2 for details on the graph properties and networking applications of TRN). One such graph property of TRN that has been (and can continue to be) exploited in the design of smart networking solutions is the abundance of subgraphs called network motifs. We demonstrated that TRN nodes participating in a high number of a 3-node motif, called Feed Forward Loop, form robust pathways for information flow. We define robustness as the ability of a network to carry out information flow under node and link failures. While we experimentally demonstrated bioMCS to be a highly energy-efficient and robust framework that leverages collaborative sensing to achieve high data delivery and load balancing, it is based on a static and centralized clustering-based mapping strategy and is unlikely to scale well. The following are the significant contributions of this extended work, termed bioMCS 2.0, that set it apart from the conference version.

  • We consider a realistic scenario where the fog devices are mobile as well as energy-constrained. This makes the connectivity of the fog network intermittent rendering the centralized scenario ineffective.

  • We propose a message passing strategy to calculate the motif participation (termed motif centrality) of a fog node in a distributed manner. We include a visualization tool in our customized simulation environment to capture real-time calculation of motif centrality of fogs over time in a mobile setting.

  • We present a data forwarding strategy for fog nodes that combines energy-awareness, motif centrality and route selection based on proximity to the base station. We study how each of these factors contribute toward the next hop selection for data forwarding.

  • We perform extensive simulation experiments on the real map of New York City consisting of 5 boroughs and 59 districts. Furthermore, in addition to two random walk based human mobility models considered in the conference version, we incorporate three realistic human mobility models, namely, Least Action Trip Planning (LATP), Social Network Theoretic (SNT) and ORBIT.

  • We compare the performance of bioMCS2.0, measured in terms of data delivery ratio, communication latency and energy efficiency, to that of a standard centralized data forwarding technique.

The rest of the paper is organized as follows. Section 2 is dedicated to the related works on data acquisition in MCS and TRN, while Section 3 describes the key components in the system. Section 4 covers the distributed motif centrality and data forwarding scheme based on message passing. Section 5 presents the simulation results and analyzes the performance of the proposed data collection framework. Finally, Section 6 draws the conclusions and identifies several future research directions in this line of work.

Section snippets

Related works

In this section, we review the related works in the two relevant areas: (1) frameworks for data acquisition in MCS and (2) application of transcriptional regulatory network (TRN) in wireless communication.

System model

Consider an urban space with thousands of users with smart handheld devices (e.g., smartphones, tablets, wearables, etc.). MCS platforms are characterized by their ability to leverage the sensing capabilities of these devices to collect rich environmental information in order to provide services in various domains, such as environmental monitoring, social networking, healthcare, transportation and safety. We build on top of our prior work to envision a data collection architecture consisting of

BioMCS 2.0 data forwarding scheme

Let us now introduce the proposed data forwarding approach, consisting of three components: (a) distributed FFL motif centrality, (b) energy-awareness, and (c) proximity based route selection.

Results

We create a customized simulation environment enabled with a real-time visualization tool (shared on https://github.com/satunr/bioMCS2.0) based on the Python Simpy library [45] to capture the disparate interactions between BS fog, fog fog, fog smart device. We first present the procedure followed to generate the New York City (NYC) map as well as the properties of the human mobility models used during the analysis. Next, we analyze the efficacy of the three aspects (namely,

Conclusions and future directions

Mobile crowdsensing is a growing platform for task sensing and reporting that may involve a massive user base as well as traffic volume. We propose bioMCS 2.0 — a distributed, energy-aware data forwarding approach for mobile crowdsensing that leverages fog computing. bioMCS 2.0 utilizes a weighted score of factors, such as energy-awareness, FFL motif centrality and proximity to base station (BS) for data forwarding, while ensuring quality of information by accepting task data from reliable

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.

Acknowledgments

This work is partially supported by National Science Foundation (NSF), USA grants under award numbers CNS-1818942, CCF-1725755, CBET-1609642, CBET-1802588, CNS-1545037 and CNS-1545050.

References (53)

  • HasenfratzD.

    Participatory air pollution monitoring using smartphones

    Mob. Sens.

    (2012)
  • SchweizerI et al.

    Noisemap-real-time participatory noise maps

  • HanK et al.

    Taming the uncertainty: Budget limited robust crowdsensing through online learning

    IEEE/ACM Trans. Netw.

    (2016)
  • WangL et al.

    Effsense: A novel mobile crowd-sensing framework for energy-efficient and cost-effective data uploading

    IEEE Trans. Syst. Man Cybern.: Syst.

    (2015)
  • LaneND et al.

    Piggyback crowdsensing (PCS): Energy efficient crowdsourcing of mobile sensor data by exploiting smartphone app opportunities

  • LiuJ et al.

    Data collection for mobile crowdsensing in the presence of selfishness

    EURASIP J. Wireless Commun. Networking

    (2016)
  • BellavistaP et al.

    Scalable and cost-effective assignment of mobile crowdsensing tasks based on profiling trends and prediction: The participact living lab experience

    Sensors

    (2015)
  • JayaramanPP et al.

    Scalable energy-efficient distributed data analytics for crowdsensing applications in mobile environments

    IEEE Trans. Comput. Soc. Syst.

    (2015)
  • SherchanW et al.

    Using on-the-move mining for mobile crowdsensing

  • FiandrinoC et al.

    Sociability-driven framework for data acquisition in mobile crowdsensing over fog computing platforms for smart cities

    IEEE Trans. Sustain. Comput.

    (2017)
  • BarabásiBy Albert-László et al.

    Scale-free

    Sci. Am.

    (2003)
  • RoyS et al.

    Characterization of e. Coli gene regulatory network and its topological enhancement by edge rewiring

  • RoyS et al.

    Design of robust and efficient topology using enhanced gene regulatory networks

    IEEE Trans. Mol. Biol. Multi-Scale Commun.

    (2018)
  • RoyS. et al.

    A bio-inspired approach to design robust and energy-efficient communication network topologies

  • NaziA. et al.

    Robust deployment of wireless sensor networks using gene regulatory networks

  • NaziA et al.

    Efficient communications in wireless sensor networks based on biological robustness

  • Cited by (5)

    • Smart-3DM: Data-driven decision making using smart edge computing in hetero-crowdsensing environment

      2022, Future Generation Computer Systems
      Citation Excerpt :

      As an enabling technology for the Internet of Things (IoT), Mobile Crowd-Sensing (MCS) provides an optimal solution for data collection from a large number of heterogeneous devices [1]. It relies on human–device relationship, where an individual can use his devices, such as smartphone or wearable, to collect data about the physical environment and share it with centralized units [2]. Typically, centralized crowdsensing activities include task allocation, worker recruitment, payment, and data processing.

    • Bio-Inspired Design of Biosensor Networks

      2022, Encyclopedia of Sensors and Biosensors: Volume 1-4, First Edition
    • Influence Spread Control in Complex Networks via Removal of Feed Forward Loops

      2021, Proceedings - International Conference on Computer Communications and Networks, ICCCN
    View full text