Jump to content

WikiJournal Preprints/Mobility-aware Scheduling in Fog Computing: Analysis and Challenges

Add links
From Wikiversity

WikiJournal Preprints
Open access • Publication charge free • Public peer review

WikiJournal User Group is a publishing group of open-access, free-to-publish, Wikipedia-integrated academic journals. <seo title=" Wikiversity Journal User Group, WikiJournal Free to publish, Open access, Open-access, Non-profit, online journal, Public peer review "/>

<meta name='citation_doi' value=>

Article information

Authors: Mir Salim Ul Islam[a][i]  , Ashok Kumar[b]  

See author information ▼
  1. ↑ Department of Computer Science and Engineering, National Institute of Technology Srinagar, Jammu and Kashmir, India
  2. ↑ Department of Computer Application, Chandigarh School of Business, Chandigarh Group of Colleges, Jhanjeri, Punjab, India
  1. ↑ mir.salim27@gmail.com

Abstract

The increasing focus on Fog-IoT results in billions of Internet-connected devices that demand substantial computational power and network bandwidth. These devices are geographically distributed, heterogeneous, computational capacity constrained, inconsistent in behaviour, and generally mobile. Therefore, providing seamless service, irrespective of location and movement of the devices as well as the end-users, makes resource scheduling a significant challenge in the Fog computing paradigm. Several mobility-aware scheduling strategies have been proposed in the literature to efficiently manage the resources for mobile users and devices in the Fog environment. This paper gives a survey of mobility-aware scheduling in the Fog computing environment. It describes the many strategies presented and their benefits and drawbacks. It also includes a complete study and taxonomy of the mobility-aware scheduling field. Further, it delineates open issues and challenges. This work will provide researchers with future research directions and aid them in recognizing the gaps before planning for further research in mobility-aware scheduling.


Introduction

[edit | edit source]

With the advancement in technology and the exponential growth of mobile devices, network traffic has increased manifold in cloud computing. Due to this reason, Latency reduction and faster processing of data for mobile users have become critical challenges in providing seamless connectivity and minimal disruption while the user is moving.[1] The data movement brings the additional issue of integrity and confidentiality because data is moving via a wireless connection to a far distant cloud. Additionally, due to the cloud's location is far from mobile users, so data movement is also affected by variable network strength and phone bandwidth. The solution proposed by Bonomi et al.[2] is to extend cloud capabilities through fog computing architecture. The Fog architecture allows substantial computation, storage, and processing using the Fog devices installed close to the user’s access point. Fog computing, therefore, reduces Latency and bandwidth consumption, improves security, provides context awareness, and renders more efficient services to mobile users.[3][4][5]

Mobility Scenario in Fog Computing

However, mobility also imposes severe challenges for Fog computing due to its distributed and diverse environment. Mobility is recognized by either user-level or device-level contextual information.[6] As the user moves from one location to another, the geographical location of the smart devices also changes. The change in location of the devices raises the issue of searching and rescheduling with mobility management. Efficient re-scheduling requires a well-planned handoff mechanism that is accountable for smoothly de-registering a sensor node from a source access point where the application was initially hosted and registering it to a new access point. Figure 1 depicts these mentioned problems of change in access points while the user moves from one area to another. The services may also get interrupted when there is more distance between the Fog nodes and users.[7] Further, any disruption in communication may lead to an increase in Latency for mobile users.[8]

The significant challenges in mobility management of Fog computing are: (i) Fog node discovery- due to the heterogeneous and mobile nature of the smart mobile devices, the Fog task scheduler faces the issue of finding an optimal Fog node for scheduling the task, (ii) Handover mechanism between users and Fog nodes- imagine a Fog user is moving from location to location and accessing information about his surroundings with the help of a smart device. Due to the frequent change in the user's location, the Fog scheduler may need to repeatedly migrate the user’s task to a different Fog node available in his vicinity. The frequent migration of tasks increases the overhead of scheduling and, further, the restricted signal strength in certain places may lead to a breakdown of task or result delivery, and (iii) the Handover mechanism between Cloud and Fog- Fog nodes have limited capacities and need to continuously communicate with Cloud computing for passing information about the tasks.[9] Due to strict requirements for security, latency, network coverage, and reliability, it becomes challenging to implement an efficient handover mechanism for full mobility support in critical domains such as healthcare[10][11][12] and vehicular systems.[13][14] Thus, mobility significantly impacts the overhead of scheduling policies and the applications' performance, eventually affecting the Quality of Experience (QoE). Therefore, mobility-aware scheduling in Fog computing has observed strong attention from researchers. Our main goal is to provide a detailed review of mobility-aware scheduling in fog computing. The review provides a detailed analysis of existing scheduling strategies that concentrate specifically on mobility awareness in the Fog environment. The following are the main contributions of this paper:

  • This paper presents a detailed survey of mobility-aware scheduling in the Fog computing environment.
  • It provides the details of the different techniques proposed, as well as their advantages and limitations.
  • It provides a detailed analysis and taxonomy of the mobility-aware scheduling field.
  • It identifies several open challenges for future research directions.
[edit | edit source]

Numerous survey studies on Fog computing focus on resource management, job scheduling, and context-aware scheduling. For example, Ghobaei-Arani et al.[15] presented a survey on resource management techniques in fog computing in the form of taxonomy to highlight cutting-edge methods while also addressing unresolved challenges. The various authors in the cloud-fog area provided the task scheduling review. Their benefits and drawbacks, as well as numerous tools and challenges concerning the scheduling techniques and their limitations, were analyzed by Alizadeh et al.[16] Islam et al.[6] thoroughly analyze relevant literature on context-aware scheduling in fog computing. Further, Mouradian et al.[17] review fog computing's significant issues and challenges. But, the critical area of mobility research is still in its initial stage, and most of the review papers contain very few documents on mobility-aware task scheduling in the area of fog computing. Therefore, an extensive and comparative study is required in mobility-constrained fog computing. A deep insight into various techniques which can impact the user QoS is necessary to understand mobility-aware fog computing. This motivates us to carry out a comprehensive survey; to the best of our knowledge, this is the first detailed survey. This survey paper thoroughly examines existing scheduling solutions that focus on user mobility. Moreover, various mobility-aware scheduling techniques are discussed, along with their pros and cons. Further, the impact of mobility parameters on various QoS parameters and context-awareness is also analyzed thoroughly.

Paper Organization

[edit | edit source]

The remainder of this review article is divided into 5 sections: Section 2 discusses mobility-aware scheduling in Fog computing. Section 3 presents the review methodology. Section 4 analyzes and summarizes the considered research papers and compares existing mobility-aware scheduling strategies. Section 5 provides the results drawn after critically examining the existing literature on mobility-aware scheduling policies. Finally, Section 6 presents the conclusion.

Mobility-aware Scheduling in Fog Computing

[edit | edit source]

Mobile device management compromises the fundamental features of Fog computing because whenever a user moves, the distance between them increases, impacting the QoS. Therefore, to keep the computing fog node close to the associated mobile device, the services or tasks need to be migrated from one fog node to another appropriate fog device. The selection of such appropriate fog nodes in a mobile environment deals with two main processes: Estimation of user mobility patterns: User mobility estimation techniques can be probabilistic and deterministic.[18] In a deterministic method, the source and destination are known beforehand, whereas in a non-deterministic technique, periodic estimation has to be done regarding the user's route. Many authors estimate the route of mobile users by leveraging external services like Open Street Maps (http://www.openstreetmap.org), Google Maps APIs (https://cloud.google.com/ maps-platform/)[18], logistics maps,[19] GPRS Here APIs (https://developer. here.com/),[20][21] Lyapunov estimation technique.[22] Specific QoS requirements: The selection of a fog node also depends upon specific quality requirements such as Latency, which many authors are using to select an efficient fog node,[18][10][9] workload, [23][24] cost,[16] etc.

Review Methodology

[edit | edit source]

The review process was conducted considering a review methodology consisting of four phases. The first phase was searching using traditional online database sources based on outlined search keywords. Table 1 lists the keywords used to find the relevant research articles to conduct the review in the Fog computing mobility-aware scheduling area. Second Phase: Limit the search of research articles beginning in 2015, and inclusion and exclusion principles are also used to refine research articles that specifically deal with mobility issues in task scheduling. Finally, in the Third Phase, A total of 20 papers are shortlisted for the review process. Further, Table 2 presents the research questions drafted for this study in mobility-aware scheduling in Fog computing.


Table 1: List of keywords used in the review process
Sno Keyword Description Years
1 Mobility Mobility-aware Fog task scheduling 2015- 2021
2 Mobility environment Mobility environment in Fog task scheduling
3 Mobility factors Mobility factors in Fog task scheduling
4 Mobility Awareness Mobility awareness in Fog task scheduling
5 Mobility management Mobility management in Fog task scheduling
Table 2: List of research questions used to complete the review process
Q.No Research questions
1 What scheduling approaches are used in the Fog computing environment to manage mobility awareness?
2 What are the main limitations considered for mobility-aware scheduling techniques?
3 Which case studies are applied to mobility-aware scheduling techniques?
4 What evaluation tools are used to assess mobility-aware scheduling techniques?
5 What performance indicators are utilized to evaluate mobility-aware scheduling techniques?
6 What are the major open issues concerns in the field of mobility-aware scheduling for future research directions?

Source of Information

[edit | edit source]

To conduct this review, various online sources are listed below, and the research articles were searched using the different keywords mentioned in Table 1.

  • Google Scholar (http://www.scholar.google.co.in)
  • John Wiley & Sons Inc. (https://onlinelibrary.wiley.com/)
  • Elsevier (https://www.elsevier.com/en-in)
  • ACM Digital Library (https://www.acm.org/)
  • Springer (https://www.springer.com/in)
  • IEEE Xplore Digital Library (https://www.ieee.org/)

Quality Assessment

[edit | edit source]

Research papers used quality assessment to filter out the most suitable mobility-based scheduling research articles in fog computing utilising the principle of inclusion and exclusion. Furthermore, in order to obtain high-quality research publications, the Center for Reviews and Dissemination (CRD) recommendations were followed, and each study item was examined for internal and external validation of results.

Literature Analysis

[edit | edit source]

Verma et al.[25] proposed a Server Cloudlet(SC) migration technique to handle users' mobility. The strategy is to select the target SC to migrate the services based on its highest rank. The rank of the SC further depends upon its available RAM, MIPS, and bandwidth.


Maleki et al.[26] designed two mobility-aware computation offloading approaches, sampling-based Online mobile applications to cloudlets(S-OAMC) and Greedy online mobile applications to cloudlets(G-OAMC). The developed system works in three major steps: (i) the future specifications of mobile applications are predicted using the Machine Learning (ML) method, named matrix completion. (ii) The predicted specifications of mobile applications, including future location prediction, also help estimate the offloading cost. (iii) Apply S-OAMC or G-OAMC offloading technique based on minimum cost value.


Abdelmoneem et al.[10] work on a fog-based healthcare architecture system that supports patient movement without altering their Quality of Service (QoS). The QoS in a mobile environment is maintained with the help of their proposed handoff mechanism. The authors modified the traditional Horizontal Handoff (HHO) mechanism[27] into two different phases: (i) handoff decision policy: Received signal Strength (RSS) method is used to detect the network change information of the mobile patients and (ii) handoff strategy: a handoff decision is made based on a comparison between RSS received from the old fog gateways, current threshold value and new fog gateways that are in close proximity to the mobile patient.


Javanmardi et al.[28] consider an imaginative city-based mobility scenario where the user is placing delay-sensitive service while moving. The authors proposed a task scheduling technique that jointly employs fuzzy logic and particle swarm optimization (PSO) algorithm to improve the QoS for mobile users within the city. The proposed algorithm is deployed at the Fog gateway, which further distributes the tasks to Fog devices available in an entire region in order to provide seamless service to mobile users. The main motive of the work is to improve resource utilization in a mobility-aware environment.


Puliafito et al.[29] develop an extension of ifogsim,[30] which supports mobile environment. The authors implement different migration techniques inspired by Bittencourt et al.[31] Different phases are devised to perform migration in a fog mobile environment, which: (i) Before migration phase: migration decision is taken in this phase, based on specific parameters, like user location, speed, the direction of movement, zone, and migration point in order to select the appropriate cloudlet for offloading the services, (ii) During the migration phase: this phase manages, monitors and synchronizes the whole selected migration process and (iii) After migration phase: this phase involves closing the older cloudlet connections with the user and using the new cloudlet for services.


A Blockchain-based Mobility-aware Offloading (BMO) mechanism is designed by Dou et al.,[32] where user mobility prediction is implemented using the Individual-Mobility (IM) model.[33] The idea behind the offloading mechanism is to shift the computational workload to different available Fog Servers (FSs) in the geo-location predicted by the IM model. Further, blockchain technology is deployed to check the authenticity of the forthcoming Fog servers. Finally, accounting is being managed by Fogcoin, similar to Bitcoin, which stores the entire transaction history between the online Fog server and mobile users.


Martin et al.[7] proposed a framework that supports the migration of containers while satisfying the QoS requirements of mobile users. The migration of containers is done in an autonomic manner, by adopting the Monitor-Analyze-Plan-Execute (MAPE) autonomic control loop. The MAPE control loops discuss various steps of migration, like (i) Monitor: used to constantly monitor the environment context, such as the mobility of users that is subsequently used to determine the need to migrate an application module to some other Fog node called target node; (ii) Analyze: applies forecasting techniques to predict the user possible location in the next time step. If the distance between the user and the device is not under certain acceptable limits, a migration decision is made. (iii) Plan: a Genetic Algorithm (GA) is used to identify a suitable Fog node closest to the forecasted location, where migration of the container running user application can be done. (iv) Execute: this step ensures the whole migration process should take place smoothly.


Mass et al. propose a mobility and delay-aware fog server selection scheme[34] called Edge-Process management (EPM) system. The EPM system depends upon the trajectory of a user’s movement, Fog server workload, and user location to select the appropriate Fog server for executing user applications. The system selects or re-selects a Fog Server (FS) based on a score value calculated through available bandwidth, power, distance from the user, and finally, duration of availability in a region.


Mobi-IoST (Mobility-aware Internet of Spatial Things), a real-time mobility-aware framework is presented by Ghosh et al.[35] The authors considered the mobile nature of both IoT devices and Fog nodes, collaboratively called mobile agents. The proposed mobility-aware framework collects a vast amount of Global Positioning System (GPS) data of these mobile agents to predict their movement patterns using various machine learning algorithms. The major components of the framework are, (i) Movement pattern modelling, collecting and modelling GPS log, stay-point, and other contextual location information; (ii) Predicting the following location: human movement semantics is analyzed using all modelled information; (iii) Delivery of result: after the user movement prediction in the previous phase, the system intelligently discovers a capable fog device for data processing in a timely manner.


A middleware solution, URMILA, for managing resources and scheduling tasks in the Fog environment is presented by Shekhar et al. [18] Ubiquitous Resource Management for Interference and Latency-Aware services (URMILA), ensures minimum Service Level Objectives (SLO) violation for latency-sensitive mobile applications across the cloud-Fog environment. The major modules of the proposed system are (i) Route calculation, which calculates the user's possible routes using Google Maps or GPS data; (ii) Latency calculation, the system deploys a data-driven model to estimate the Latency on predicted user routes; (iii) Fog node selection: the system selects a fog server for execution of task on the basis of its instantaneous utilization of the available resources. Further, it selects the Fog server for the entire period of execution, during which mobile users can still access their application through various Wireless Access Points (WAP).


Gia et al.[8] proposed a Handover mechanism for mobility management between fog nodes with the overall objective of consuming minimum energy and delay during handovers. Handover methods frequently rely on one or more measures, such as the Received Signal Strength Indicator (RSSI), the Link Quality Indicator (LQI), and the velocity of objects, to make handover decisions. This proposed system provides emergency services to health monitoring systems and basically works in two different mobility scenarios: (i) Node mobility between indoor or outdoor locations: nodes belonging to indoor location or outdoor location only are considered to be similar, and they're calculated metrics value like RSSI, LQI, velocity, etc.; can be directly used for the handover of services to appropriate gateway, (ii) Node mobility between indoor and outdoor locations: nodes are considered dissimilar, if they belong to indoor and outdoor location both, So, the metrics are re-calculated which introduce some additional parameters like temperature and interference signals in order to make a decision over handover gateway.


Babu and Biswash[36] proposed a mobility management technique that supports node-to-node communication and Fog computing-based architecture for 5G networks. It addresses the technical problems between 5G networks and Fog servers. The mobility-based approach assists mobile nodes in establishing communication while they are in motion. The mobility management technique may also be used to begin N2N communication in dynamic environments. N2N communication schemes for fog networks, on the other hand, provide an effective communication environment for mobile users with highly minimal network usage.


Wang et al.[37] proposed a mobility-aware offloading scheme, that considers an adequate quality and a computation allocation system that deals with the user equipment affairs to maximize the total revenue. The quality of user equipment is delineated by the sojourn time that follows the exponential distribution to reduce the chance of migration and maximize the entire income of user equipment. MILP (mixed-integer non-linear programming) NP-hard problem is modelled and consists of resource allocation and task offloading schemes. So, to solve this problem, a Gini coefficient and genetic algorithm are used to estimate the allocation of resources. The proposed approach can easily handle the mobility of users by minimizing the chances of migration.


Waqas et al.[38] provided a forward-thinking analysis of quality about-mobility in Fog computing by identifying quality challenges, requirements, and options for numerous ideas. The authors also identified outstanding concerns from previous research and summarized the advantages of quality for readers. It allows researchers and developers to avoid common misunderstandings and capture real-world scenarios such as businesses, governments, and educational institutions. Furthermore, it revolutionizes follow-up analysis and differentiates and foregrounds futurity orientations in real-life events involving humans and vehicles in a highly dynamic Fog setting.


Bi et al.[9] introduced software-defined networking-based fog computing architecture by decoupling mobility control and data forwarding. When mobile consumers travel between several access networks, the authors suggested an Optimal Path Selection (OPS) algorithm to preserve service continuity. Mobile customers received seamless and transparent mobility support thanks to efficient signalling operations. In mobile fog computing, the suggested algorithm ensured service continuity, increased handover performance, and achieved high data transfer efficiency.


Niu et al.[21] established a system called mobility-aware and multihop-D2D relaying-based scheduling scheme (MHRC) at Edge nodes near hotspots. The authors exploited concurrent transmissions to improve the performance of the system. The mmWave (millimetre-wave) band of Fog computing was cached, and extensive performance evaluation confirms that MHRC delivers more than the higher expected cached data amount. Name et al.[19] proposed an efficient algorithm to address the problem of resource allocation and user mobility from the Edge of the network to cloud data centres. This algorithm operates on a seamless handover scheme for mobile IPV6 to ease the user mobility challenge and reduce the application response time. The study showed that the task of service delay and packet loss was decreased due to the effect of change in the mobile node position.


Bittencourt et al.[23] examined the subject of resource allocation in the Fog/Cloud environment, taking into account the hierarchical structure. In the context of the Fog paradigm, the authors developed three scheduling algorithms (First come, First serve, delay-priority, and concurrent) that address user mobility and edge computing capabilities. The authors demonstrated that scheduling techniques may be designed to cope with different application classes based on demand from mobile users by leveraging both Fog to the end-user and cloud characteristics in this study.


Velasquez et al.[39] proposed a hybrid strategy for the Fog environment to manage resources for mobility scenarios. The authors applied the orchestrator technique to offer mobility support in a Smart City situation. In this technique, three components, the status monitor, the Planner, and the VM/Container, are employed to monitor, plan and execute the applications. The main aim of this study was to guarantee the QoS and QoE requirements of mobility-based applications and services.


Bittencourt et al.[31] presented a Fog computing architecture focusing on Virtual Machine (VM) migration where each user has a VM running in a cloudlet. In this architecture, the user's location is identified by using GPS, and then the VM is moved to a nearby Fog Cloud. The main aim of this study was to migrate users' data according to their mobility in order to maintain QoS for applications demanding lower Latency and allow smooth handoff mechanisms for mobile users.


From the extensive analysis of the literature, the various mobility-aware scheduling techniques have been classified as shown in Table 3. Further, it presents the advantages and limitations of each technique.

Table 3: Classification of Mobility-aware scheduling techniques
Ref. Technique Advantage(s) Limitation(s)
[25] Ranking of VM
  • Decrease in delay time, migration time, tuple lost value and downtime
  • Case study not discussed
[26] S-OAMC, G-OAMC, Machine learning matrix completion
  • Migration rate decreased
  • Better Scalability
  • Energy utilization of devices not investigated
[32] IM model
  • Provides better mobility support and security
  • Did not investigate synchronization overhead
[10] RSS
  • Reliable and Heterogeneous execution
  • Low scalability
  • No distributed scheduling to minimize response time
[7] MAPE control loop
  • Improved QoS
  • Reduced service downtime
  • No real-time evaluation
  • High energy consumption
  • Low robustness and security
[28] Copy of task to over a region
  • Network Utilization developed
  • Low-Loop delay
  • Fault tolerance reliability is based on Fog gateways only
[18] URMILA
  • Service availability is maintained by delivering the desired QoS
  • Deployment cost minimized
  • Battery longevity ensured
  • No empirical validations
  • No user probabilistic routes
  • Low scalability in terms of distance and speed
[8] RSSI, LQI
  • Promises to keep the connection active with a low latency rate between the system and sensor nodes
  • Consumes more energy
  • Overhead is large for network transmission
  • Coverage and overhead area are undefined between gateways
[29] User trajectories pre- diction using GPS log
  • Provides better mobility support
  • Reduces migration time
  • Low scalability
[36] N2N communication, Data Analytics
  • Fast data access
  • High reliability and scalability- city
  • Minimum overhead and cost
  • High throughput and less delay
  • No real-time cellular network evaluation
  • Low network efficiency
[38] Mobility facets analysis
  • Improved QoS and QoE
  • Latency rate reduced
  • No real-life implementation
  • No reliability and low Latency between dynamic users and fog servers
[37] M-ILP, Sojourn time
  • Cost-effective
  • Migration time reduced
  • Migration cost not considered
  • No real-time implementation
[34] User trajectories prediction using GPS log
  • Conventional delay tolerance
  • High QoS
  • Avoided local task processing cost
  • Efficient in saving battery
  • Handles subtle scenarios with high Latency
  • Smart city not directed through the use of accurate city maps with aid from stimulation setting
[35] Prediction of user location
  • Power consumption and de- lay handled proficiently
  • Power and delay are reduced
  • No acquiring of mobile data usage where location sense and time-series data can be projected to achieve the bandwidth
[21] Relay path planning algorithm
  • Power efficient
  • High spectral efficiency
  • Data is relayed on cached edge nodes and relay nodes
  • Blockage problem due to weak diffraction
[9] OPS
  • Handover performance improved
  • Efficiency of high data communication achieved
  • Guarantees continuity of services
  • It does not guarantee privacy and security
  • Virtual Machine migration not determined
  • The handover process during the optimal path for more logical routing could have been more efficient
[23] Assignment of FS
  • Low Latency
  • Supports dynamic computing and user behaviour
  • There is no prediction of mobility failure
  • Bandwidth and processing not considered in scheduling
[39] Orchestrator
  • Maintains trustworthiness, resilience, and low Latency in a dynamic environment
  • No real implementation has been carried out
[19] Pattern modelling, dictating the following location
  • Application Response time reduced
  • Latency time reduced
  • Services become temporarily inaccessible for some mobile nodes
[31] Forecasting technique
  • Computing capacity provided for storage and processing of data
  • Security concerns associated with both user data and applications not considered

Analytical Discussion

[edit | edit source]

The existing research on Fog computing Mobility-aware scheduling has been analyzed thoroughly. The analysis was performed using the answers given in Table 2. The results drawn through the thorough analysis of the literature are presented in various figures as follows:

Year-wise count of research articles

Figure 2 lists the year-wise count of research papers that are considered for this survey. The bar graph represents the total number of research papers from journals and Conferences from the year 2015 - 2021. The research articles from the journal are 16, and the conference papers are 4. It is observed that more research needs to be conducted on mobility-aware scheduling in Fog computing.

Figure 3 displays an analytical comparison of mobility-aware scheduling approaches in Fog computing based on the content of the represented taxonomy in Figure 7. From the thorough analysis of the literature, four methods have been considered: migration, task offloading, handoff/handover mechanism, and task scheduling. The handoff/handover mechanism has the highest percentage of usage in mobility-aware scheduling, at 30%. The task scheduling and offloading have 25% of us- age in mobility-aware scheduling each. Finally, migration is only 20% of the usage in mobility-aware scheduling. Therefore, these approaches, specifically migration, are still open challenges to address for further research.

Figure 4 depicts various tools that were used for evaluating the mobility-aware scheduling approaches. 18% and 9% of the research articles used iFogSim and Mob-FogSim simulation tools for implementation, respectively. Besides, other simulation tools such as ONE (9%), NS2 (5%), MATLAB (4%), Mininet (5%), and Docker (9%) have been utilized for implementing the proposed techniques in the research articles. Further, pro-Programming languages such as C++ (9%) and Python (9%) and hardware deployments such as Raspberry Pi (5%) and Ardunio (4%) were used for implementing existing case studies based on mobility-aware scheduling.

Percentage of the presented classified approaches in mobility-aware scheduling


The applied case studies are shown in Figure 5, which shows a maximum of 20% of research articles have implemented IoT-based applications. After that, 15% of each research article used Health care and Mobile-based applications. Besides, Smart City and 5G-based applications have been applied in 10% of research articles. Moreover, Surveillance and gaming, Mobile IPV6, and Wireless computing applications are the case studies on which only 5% of research articles exist.

After reviewing numerous research articles based on mobility-aware scheduling, it has been observed that researchers employed various parameters for evaluating the performance of the Mobility-scheduling approaches, as represented in Figure 6. It shows that Time completion (18%) followed by Delay (12%), Network usage (12%), Latency (12%), Energy consumption (10%), and cost (10%) are generally utilized. However, Downtime (4%), Migration time (4%), Makespan (2%), Success ratio (2%), Signal level (2%), Deadline (2%), Makespan (2%), Migration rate (2%), Mobility patterns (2%), Tuple lost (2%), and power consumption (2%) are less exploited parameters.

A taxonomy was compiled after going through the detailed review process, and various techniques have been categorized in Fog computing-based mobility-aware scheduling. Figure 7 presents these categories broadly in Migration, Offloading, Handoff/Handover mechanism, and Scheduling.

Percentage of tools utilized in the literature
Percentage of case studies employed in the literature
Percentage of parameters for evaluating Mobility-aware scheduling in the literature

Open Issues and Challenges

[edit | edit source]

From the thorough analysis of the literature, several open issues and challenges pertaining to the area of mobility-aware scheduling in Fog computing have been identified in order to provide directions for future research exploration. The identified open problems and challenges, depicted in Figure 8, are discussed below.

Task Scheduling

[edit | edit source]

Fog computing consists of several Fog nodes, each of which is a mini Cloud in the vicinity of mobile devices near the Edge of the network. When a mobile device submits a task, the Fog scheduler assigns it to a nearby Fog node(s) for execution. However, as the device moves from one network to another, the task needs to be rescheduled when the device enters a different network. Additionally, Fog nodes have limited capacity and availability; if the mobile user enters into a network where there is no nearby Fog service available, then this leads to a significant delay in service and raises a significant issue of task scheduling for mobile users.[10][40]

Resource Provisioning

[edit | edit source]

Fog computing reduces the workload of Cloud computing by processing the tasks locally near the Edge of the network. However, due to the mobility of the user, the Fog node primarily assigned to a task might not be optimal over time. Therefore, the migration of the task to another Fog node near the user's mobile device is perceived as a necessary solution to resolve this concern.[41] However, such frequent migration over a short time poses the challenge of providing an efficient resource for the task that is capable of performing computation on time and delivering results to users while adhering to QoE.

Mobility-aware Fog Scheduling Taxonomy

Energy Consumption

[edit | edit source]

The placement of fog services at the Edge of the network can provide better QoS to mobile users, resulting in a shorter response time. However, it is practically impossible due to the high deployment cost of new Fog devices, which further raises the significant issue of energy consumption. If too many deployments are done, there will be lots of communication traffic from the Cloud to Fog nodes and servers in order to create copies of the task from one network to another in case of mobility.[42] This results in considerable energy wastage in the form of high bandwidth consumption. This means that where and when to reschedule the task to an efficient Fog node must be carefully determined to minimize energy, response time, and deployment cost.

Quality of Experience (QoE)

[edit | edit source]
Mobility-aware scheduling open issues and challenges

Several mobility-based scheduling algorithms exist, but they need to focus on maximizing the user QoE.[29][8][10][18] Further, they do not analyze the user performance; hence, the QoE of using a service or product is not determined. Therefore, to understand the user gain and loss, the scheduling algorithm needs to focus on enhancing the user QoE.

Resource Management

[edit | edit source]

The mobility of Fog nodes/users demands efficient resource discovery and sharing, resource availability, and task offloading.[43] Few techniques that were proposed to manage the resources effectively did not consider more constraints such as density, latency sensitivity, and mobility of Edge and Fog nodes, and as the number of nodes increases, issues such as scalability and distributing the algorithms arise.[44][45][46] Therefore, more attention needs to be paid towards the mobile Fog computing environment to manage the resources effectively.

Privacy and Security

[edit | edit source]

In [47], a scheduling policy is proposed for the mobile device system to minimize the cost. However, the privacy issues of location and usage patterns were ignored. Additionally, data privacy, access control, and intrusion detection in scheduling policies have been overlooked.[7][48][28] Besides, Fog node devices are normally deployed near the end-user; hence, protection and surveillance are comparatively weak, which can result in a malicious attack.[49][50]

Data availability statement

[edit | edit source]

Not applicable.

Conclusions

[edit | edit source]

Fog computing infrastructure provides services at the Edge of the network. So, to provide support for scheduling and management of mobility awareness, efficient techniques and mechanisms have been proposed. In this survey, research articles on the mobility-aware-scheduling strategies in Fog computing have been thoroughly analyzed. It provides a comparative study among existing mobility-aware scheduling strategies based on vital factors such as techniques proposed, parameters considered, tools utilized for implementation, and case studies considered, along with the advantages and limitations. Further, several open issues and challenges have been identified for future research direction.

Additional information

[edit | edit source]

Data availability statement

[edit | edit source]

Not applicable.

References

[edit | edit source]
  1. ↑ Abdullah, Fatima; Kimovski, Dragi; Prodan, Radu; Munir, Kashif (2021-06-11). "Handover authentication latency reduction using mobile edge computing and mobility patterns". Computing 103 (11): 2667–2686. doi:10.1007/s00607-021-00969-z. ISSN 0010-485X. https://doi.org/10.1007/s00607-021-00969-z. 
  2. ↑ Bonomi, Flavio; Milito, Rodolfo; Zhu, Jiang; Addepalli, Sateesh (2012-08-17). "Fog computing and its role in the internet of things". Proceedings of the first edition of the MCC workshop on Mobile cloud computing. MCC '12 (New York, NY, USA: Association for Computing Machinery): 13–16. doi:10.1145/2342509.2342513. ISBN 978-1-4503-1519-7. https://dl.acm.org/doi/10.1145/2342509.2342513. 
  3. ↑ Puliafito, Carlo; Mingozzi, Enzo; Longo, Francesco; Puliafito, Antonio; Rana, Omer (2019-04-02). "Fog Computing for the Internet of Things: A Survey". ACM Trans. Internet Technol. 19 (2): 18:1–18:41. doi:10.1145/3301443. ISSN 1533-5399. https://dl.acm.org/doi/10.1145/3301443. 
  4. ↑ Dastjerdi, Amir Vahid; Buyya, Rajkumar (2016-08). "Fog Computing: Helping the Internet of Things Realize Its Potential". Computer 49 (8): 112–116. doi:10.1109/mc.2016.245. ISSN 0018-9162. https://doi.org/10.1109/mc.2016.245. 
  5. ↑ Atlam, Hany F.; Walters, Robert J.; Wills, Gary B. (2018-06). "Fog Computing and the Internet of Things: A Review". Big Data and Cognitive Computing 2 (2): 10. doi:10.3390/bdcc2020010. ISSN 2504-2289. https://www.mdpi.com/2504-2289/2/2/10. 
  6. ↑ 6.0 6.1 Islam, Mir Salim Ul; Kumar, Ashok; Hu, Yu-Chen (2021-04). "Context-aware scheduling in Fog computing: A survey, taxonomy, challenges and future directions". Journal of Network and Computer Applications 180: 103008. doi:10.1016/j.jnca.2021.103008. ISSN 1084-8045. https://doi.org/10.1016/j.jnca.2021.103008. 
  7. ↑ 7.0 7.1 7.2 7.3 Martin, John Paul; Kandasamy, A; Chandrasekaran, K (2020-03-09). "Mobility aware autonomic approach for the migration of application modules in fog computing environment". Journal of Ambient Intelligence and Humanized Computing 11 (11): 5259–5278. doi:10.1007/s12652-020-01854-x. ISSN 1868-5137. https://doi.org/10.1007/s12652-020-01854-x. 
  8. ↑ 8.0 8.1 8.2 8.3 Nguyen Gia, Tuan; Rahmani, Amir M.; Westerlund, Tomi; Liljeberg, Pasi; Tenhunen, Hannu (2018). "Fog Computing Approach for Mobility Support in Internet-of-Things Systems". IEEE Access 6: 36064–36082. doi:10.1109/ACCESS.2018.2848119. ISSN 2169-3536. https://ieeexplore.ieee.org/document/8386744/. 
  9. ↑ 9.0 9.1 9.2 9.3 Bi, Yuanguo; Han, Guangjie; Lin, Chuan; Deng, Qingxu; Guo, Lei; Li, Fuliang (2018-05). "Mobility Support for Fog Computing: An SDN Approach". IEEE Communications Magazine 56 (5): 53–59. doi:10.1109/mcom.2018.1700908. ISSN 0163-6804. https://doi.org/10.1109/mcom.2018.1700908. 
  10. ↑ 10.0 10.1 10.2 10.3 10.4 10.5 Abdelmoneem, Randa M.; Benslimane, Abderrahim; Shaaban, Eman (2020-10). "Mobility-aware task scheduling in cloud-Fog IoT-based healthcare architectures". Computer Networks 179: 107348. doi:10.1016/j.comnet.2020.107348. ISSN 1389-1286. https://doi.org/10.1016/j.comnet.2020.107348. 
  11. ↑ Rahmani, Amir M.; Gia, Tuan Nguyen; Negash, Behailu; Anzanpour, Arman; Azimi, Iman; Jiang, Mingzhe; Liljeberg, Pasi (2018-01). "Exploiting smart e-Health gateways at the edge of healthcare Internet-of-Things: A fog computing approach". Future Generation Computer Systems 78: 641–658. doi:10.1016/j.future.2017.02.014. ISSN 0167-739X. https://doi.org/10.1016/j.future.2017.02.014. 
  12. ↑ Kraemer, Frank Alexander; Braten, Anders Eivind; Tamkittikhun, Nattachart; Palma, David (2017). "Fog Computing in Healthcare–A Review and Discussion". IEEE Access 5: 9206–9222. doi:10.1109/ACCESS.2017.2704100. ISSN 2169-3536. http://ieeexplore.ieee.org/document/7927714/. 
  13. ↑ Zhu, Chao; Pastor, Giancarlo; Xiao, Yu; Li, Yong; Ylae-Jaeaeski, Antti (2018-06). "Fog Following Me: Latency and Quality Balanced Task Allocation in Vehicular Fog Computing". 2018 15th Annual IEEE International Conference on Sensing, Communication, and Networking (SECON) (IEEE): 1–9. doi:10.1109/sahcn.2018.8397129. https://doi.org/10.1109/sahcn.2018.8397129. 
  14. ↑ Aljeri, Noura; Boukerche, Azzedine (2020-01-17). "Fog‐enabled vehicular networks: A new challenge for mobility management". Internet Technology Letters 3 (6). doi:10.1002/itl2.141. ISSN 2476-1508. https://doi.org/10.1002/itl2.141. 
  15. ↑ Ghobaei-Arani, Mostafa; Souri, Alireza; Rahmanian, Ali A. (2019-09-06). "Resource Management Approaches in Fog Computing: a Comprehensive Review". Journal of Grid Computing 18 (1): 1–42. doi:10.1007/s10723-019-09491-1. ISSN 1570-7873. https://doi.org/10.1007/s10723-019-09491-1. 
  16. ↑ 16.0 16.1 Alizadeh, Mohammad Reza; Khajehvand, Vahid; Rahmani, Amir Masoud; Akbari, Ebrahim (2020-11-10). "Task scheduling approaches in fog computing: A systematic review". International Journal of Communication Systems 33 (16). doi:10.1002/dac.4583. ISSN 1074-5351. https://onlinelibrary.wiley.com/doi/10.1002/dac.4583. 
  17. ↑ Mouradian, Carla; Naboulsi, Diala; Yangui, Sami; Glitho, Roch H.; Morrow, Monique J.; Polakos, Paul A. (21/2018). "A Comprehensive Survey on Fog Computing: State-of-the-Art and Research Challenges". IEEE Communications Surveys & Tutorials 20 (1): 416–464. doi:10.1109/COMST.2017.2771153. ISSN 1553-877X. http://ieeexplore.ieee.org/document/8100873/. 
  18. ↑ 18.0 18.1 18.2 18.3 18.4 18.5 Shekhar, Shashank; Chhokra, Ajay; Sun, Hongyang; Gokhale, Aniruddha; Dubey, Abhishek; Koutsoukos, Xenofon (2019-05). "URMILA: A Performance and Mobility-Aware Fog/Edge Resource Management Middleware". 2019 IEEE 22nd International Symposium on Real-Time Distributed Computing (ISORC) (IEEE): 118–125. doi:10.1109/isorc.2019.00033. https://doi.org/10.1109/isorc.2019.00033. 
  19. ↑ 19.0 19.1 19.2 Name, Haruna Abdu Manis; Oladipo, Francisca O.; Ariwa, Ezendu (2017-08). "User mobility and resource scheduling and management in fog computing to support IoT devices". 2017 Seventh International Conference on Innovative Computing Technology (INTECH) (IEEE): 191–196. doi:10.1109/intech.2017.8102447. https://doi.org/10.1109/intech.2017.8102447. 
  20. ↑ Turgut, D.; Das, S.K.; Chatterjee, M. (2001). Longevity of routes in mobile ad hoc networks. 4. IEEE. pp. 2833–2837. doi:10.1109/VETECS.2001.944118. ISBN 978-0-7803-6728-9. http://ieeexplore.ieee.org/document/944118/. 
  21. ↑ 21.0 21.1 21.2 Niu, Yong; Liu, Yu; Li, Yong; Zhong, Zhangdui; Ai, Bo; Hui, Pan (2018). "Mobility-Aware Caching Scheduling for Fog Computing in mmWave Band". IEEE Access 6: 69358–69370. doi:10.1109/access.2018.2880031. ISSN 2169-3536. https://doi.org/10.1109/access.2018.2880031. 
  22. ↑ Li, Yun; Xia, Shichao; Zheng, Mengyan; Cao, Bin; Liu, Qilie (2022-01-01). "Lyapunov Optimization-Based Trade-Off Policy for Mobile Cloud Offloading in Heterogeneous Wireless Networks". IEEE Transactions on Cloud Computing 10 (1): 491–505. doi:10.1109/tcc.2019.2938504. ISSN 2168-7161. https://doi.org/10.1109/tcc.2019.2938504. 
  23. ↑ 23.0 23.1 23.2 Bittencourt, Luiz F.; Diaz-Montes, Javier; Buyya, Rajkumar; Rana, Omer F.; Parashar, Manish (2017-03). "Mobility-Aware Application Scheduling in Fog Computing". IEEE Cloud Computing 4 (2): 26–35. doi:10.1109/mcc.2017.27. ISSN 2325-6095. https://doi.org/10.1109/mcc.2017.27. 
  24. ↑ Ahanger, Tariq Ahamed; Tariq, Usman; Nusir​, Muneer (2022-04-17). "Mobility of Internet of Things and Fog Computing: Serious Concerns and Future Directions". International Journal of Communication Networks and Information Security (IJCNIS) 10 (3). doi:10.17762/ijcnis.v10i3.3706. ISSN 2073-607X. https://doi.org/10.17762/ijcnis.v10i3.3706. 
  25. ↑ 25.0 25.1 Verma, Kanupriya; Kumar, Ashok; Ul Islam, Mir Salim; Kanwar, Tulika; Bhushan, Megha (2021). "Rank based mobility-aware scheduling in Fog computing". Informatics in Medicine Unlocked 24: 100619. doi:10.1016/j.imu.2021.100619. ISSN 2352-9148. https://doi.org/10.1016/j.imu.2021.100619. 
  26. ↑ 26.0 26.1 Maleki, Erfan Farhangi; Mashayekhy, Lena; Nabavinejad, Seyed Morteza (2023-01-01). "Mobility-Aware Computation Offloading in Edge Computing Using Machine Learning". IEEE Transactions on Mobile Computing 22 (1): 328–340. doi:10.1109/TMC.2021.3085527. ISSN 1536-1233. https://ieeexplore.ieee.org/document/9444789/. 
  27. ↑ Kassar, Meriem; Kervella, Brigitte; Pujolle, Guy (2008-06). "An overview of vertical handover decision strategies in heterogeneous wireless networks". Computer Communications 31 (10): 2607–2620. doi:10.1016/j.comcom.2008.01.044. ISSN 0140-3664. https://doi.org/10.1016/j.comcom.2008.01.044. 
  28. ↑ 28.0 28.1 28.2 Javanmardi, Saeed; Shojafar, Mohammad; Persico, Valerio; Pescapè, Antonio (2020-08-18). "FPFTS: A joint fuzzy particle swarm optimization mobility‐aware approach to fog task scheduling algorithm for Internet of Things devices". Software: Practice and Experience 51 (12): 2519–2539. doi:10.1002/spe.2867. ISSN 0038-0644. https://doi.org/10.1002/spe.2867. 
  29. ↑ 29.0 29.1 29.2 Puliafito, Carlo; Gonçalves, Diogo M.; Lopes, Márcio M.; Martins, Leonardo L.; Madeira, Edmundo; Mingozzi, Enzo; Rana, Omer; Bittencourt, Luiz F. (2020-05-01). "MobFogSim: Simulation of mobility and migration for fog computing". Simulation Modelling Practice and Theory. Modeling and Simulation of Fog Computing 101: 102062. doi:10.1016/j.simpat.2019.102062. ISSN 1569-190X. https://www.sciencedirect.com/science/article/abs/pii/S1569190X19301935. 
  30. ↑ Gupta, Harshit; Vahid Dastjerdi, Amir; Ghosh, Soumya K.; Buyya, Rajkumar (2017-06-21). "iFogSim: A toolkit for modeling and simulation of resource management techniques in the Internet of Things, Edge and Fog computing environments". Software: Practice and Experience 47 (9): 1275–1296. doi:10.1002/spe.2509. ISSN 0038-0644. https://doi.org/10.1002/spe.2509. 
  31. ↑ 31.0 31.1 31.2 Bittencourt, Luiz Fernando; Lopes, Marcio Moraes; Petri, Ioan; Rana, Omer F. (2015-11). "Towards Virtual Machine Migration in Fog Computing". 2015 10th International Conference on P2P, Parallel, Grid, Cloud and Internet Computing (3PGCIC) (IEEE). doi:10.1109/3pgcic.2015.85. https://doi.org/10.1109/3pgcic.2015.85. 
  32. ↑ 32.0 32.1 Dou, Wanchun; Tang, Wenda; Liu, Bowen; Xu, Xiaolong; Ni, Qiang (2020-12-01). "Blockchain-based Mobility-aware Offloading mechanism for Fog computing services". Computer Communications 164: 261–273. doi:10.1016/j.comcom.2020.10.007. ISSN 0140-3664. https://www.sciencedirect.com/science/article/abs/pii/S0140366420319460. 
  33. ↑ Song, Chaoming; Koren, Tal; Wang, Pu; Barabási, Albert-László (2010-09-12). "Modelling the scaling properties of human mobility". Nature Physics 6 (10): 818–823. doi:10.1038/nphys1760. ISSN 1745-2473. https://doi.org/10.1038/nphys1760. 
  34. ↑ 34.0 34.1 Mass, Jakob; Chang, Chii; Srirama, Satish Narayana (2019-06). "Edge Process Management: A case study on adaptive task scheduling in mobile IoT". Internet of Things 6: 100051. doi:10.1016/j.iot.2019.100051. ISSN 2542-6605. https://doi.org/10.1016/j.iot.2019.100051. 
  35. ↑ 35.0 35.1 Ghosh, Shreya; Mukherjee, Anwesha; Ghosh, Soumya K.; Buyya, Rajkumar (2020-10-01). "Mobi-IoST: Mobility-Aware Cloud-Fog-Edge-IoT Collaborative Framework for Time-Critical Applications". IEEE Transactions on Network Science and Engineering 7 (4): 2271–2285. doi:10.1109/tnse.2019.2941754. ISSN 2327-4697. https://doi.org/10.1109/tnse.2019.2941754. 
  36. ↑ 36.0 36.1 Babu, S.; Biswash, Sanjay Kumar (2019-09-03). "Fog computing–based node‐to‐node communication and mobility management technique for 5G networks". Transactions on Emerging Telecommunications Technologies 30 (10). doi:10.1002/ett.3738. ISSN 2161-3915. https://doi.org/10.1002/ett.3738. 
  37. ↑ 37.0 37.1 Wang, Dongyu; Liu, Zhaolin; Wang, Xiaoxiang; Lan, Yanwen (2019). "Mobility-Aware Task Offloading and Migration Schemes in Fog Computing Networks". IEEE Access 7: 43356–43368. doi:10.1109/access.2019.2908263. ISSN 2169-3536. https://doi.org/10.1109/access.2019.2908263. 
  38. ↑ 38.0 38.1 Waqas, Muhammad; Niu, Yong; Ahmed, Manzoor; Li, Yong; Jin, Depeng; Han, Zhu (2019). "Mobility-Aware Fog Computing in Dynamic Environments: Understandings and Implementation". IEEE Access 7: 38867–38879. doi:10.1109/access.2018.2883662. ISSN 2169-3536. https://doi.org/10.1109/access.2018.2883662. 
  39. ↑ 39.0 39.1 Velasquez, Karima; Abreu, David Perez; Goncalves, Diogo; Bittencourt, Luiz; Curado, Marilia; Monteiro, Edmundo; Madeira, Edmundo (2017-08). Service Orchestration in Fog Environments. IEEE. pp. 329–336. doi:10.1109/FiCloud.2017.49. ISBN 978-1-5386-2074-8. http://ieeexplore.ieee.org/document/8114500/. 
  40. ↑ Kaur, Navjeet; Kumar, Ashok; Kumar, Rajesh (2021-06-04). "A systematic review on task scheduling in Fog computing: Taxonomy, tools, challenges, and future directions". Concurrency and Computation: Practice and Experience 33 (21). doi:10.1002/cpe.6432. ISSN 1532-0626. https://doi.org/10.1002/cpe.6432. 
  41. ↑ Maleki, Erfan Farhangi; Mashayekhy, Lena (2020-05). "Mobility-aware computation offloading in edge computing using prediction". 2020 IEEE 4th International Conference on Fog and Edge Computing (ICFEC) (IEEE): 69–74. doi:10.1109/icfec50348.2020.00015. https://doi.org/10.1109/icfec50348.2020.00015. 
  42. ↑ Zhao, Xuhui; Shi, Yan; Chen, Shanzhi (2020-12). "MAESP: Mobility aware edge service placement in mobile edge networks". Computer Networks 182: 107435. doi:10.1016/j.comnet.2020.107435. ISSN 1389-1286. https://doi.org/10.1016/j.comnet.2020.107435. 
  43. ↑ Yousefpour, Ashkan; Fung, Caleb; Nguyen, Tam; Kadiyala, Krishna; Jalali, Fatemeh; Niakanlahiji, Amirreza; Kong, Jian; Jue, Jason P. (2019-09). "All one needs to know about fog computing and related edge computing paradigms: A complete survey". Journal of Systems Architecture 98: 289–330. doi:10.1016/j.sysarc.2019.02.009. ISSN 1383-7621. https://doi.org/10.1016/j.sysarc.2019.02.009. 
  44. ↑ Liu, Wei; Nishio, Takayuki; Shinkuma, Ryoichi; Takahashi, Tatsuro (2014-09). "Adaptive resource discovery in mobile cloud computing". Computer Communications 50: 119–129. doi:10.1016/j.comcom.2014.02.006. ISSN 0140-3664. https://doi.org/10.1016/j.comcom.2014.02.006. 
  45. ↑ He, Jianhua; Wei, Jian; Chen, Kai; Tang, Zuoyin; Zhou, Yi; Zhang, Yan (2018-04). "Multitier Fog Computing With Large-Scale IoT Data Analytics for Smart Cities". IEEE Internet of Things Journal 5 (2): 677–686. doi:10.1109/jiot.2017.2724845. ISSN 2327-4662. https://doi.org/10.1109/jiot.2017.2724845. 
  46. ↑ Nishio, Takayuki; Shinkuma, Ryoichi; Takahashi, Tatsuro; Mandayam, Narayan B. (2013). "Service-oriented heterogeneous resource sharing for optimizing service latency in mobile cloud". Proceedings of the first international workshop on Mobile cloud computing & networking - MobileCloud '13 (New York, New York, USA: ACM Press): 19. doi:10.1145/2492348.2492354. https://doi.org/10.1145/2492348.2492354. 
  47. ↑ He, Xiaofan; Liu, Juan; Jin, Richeng; Dai, Huaiyu (2017-12). Privacy-Aware Offloading in Mobile-Edge Computing. IEEE. pp. 1–6. doi:10.1109/GLOCOM.2017.8253985. ISBN 978-1-5090-5019-2. http://ieeexplore.ieee.org/document/8253985/. 
  48. ↑ Yang, Ming; Ma, Hao; Wei, Shuang; Zeng, You; Chen, Yefeng; Hu, Yuemei (2020). "A Multi-Objective Task Scheduling Method for Fog Computing in Cyber-Physical-Social Services". IEEE Access 8: 65085–65095. doi:10.1109/ACCESS.2020.2983742. ISSN 2169-3536. https://ieeexplore.ieee.org/document/9049337/. 
  49. ↑ Hu, Pengfei; Ning, Huansheng; Qiu, Tie; Song, Houbing; Wang, Yanna; Yao, Xuanxia (2017-10). "Security and Privacy Preservation Scheme of Face Identification and Resolution Framework Using Fog Computing in Internet of Things". IEEE Internet of Things Journal 4 (5): 1143–1155. doi:10.1109/jiot.2017.2659783. ISSN 2327-4662. https://doi.org/10.1109/jiot.2017.2659783. 
  50. ↑ Han, Guangjie; Liu, Li; Chan, Sammy; Yu, Ruiyun; Yang, Yu (2017-03). "HySense: A Hybrid Mobile CrowdSensing Framework for Sensing Opportunities Compensation under Dynamic Coverage Constraint". IEEE Communications Magazine 55 (3): 93–99. doi:10.1109/mcom.2017.1600658cm. ISSN 0163-6804. https://doi.org/10.1109/mcom.2017.1600658cm. 
Retrieved from "https://en.wikiversity.org/w/index.php?title=WikiJournal_Preprints/Mobility-aware_Scheduling_in_Fog_Computing:_Analysis_and_Challenges&oldid=2694159"