Offloading mobile data traffic for QoS-aware service provision in vehicular cyber-physical systems

https://doi.org/10.1016/j.future.2015.10.004Get rights and content

Highlights

  • A Wi-Fi and VANET offloading model is established and the offloading capacity is quantified.

  • A multi-objective combinatorial problem for mobile data traffic offloading is formulated and optimized.

  • The proposed offloading approach of Vehicular Cyber-Physical Systems (VCPSs) supports global QoS guarantee and service provisioning.

Abstract

Owing to the increasing number of vehicles in vehicular cyber-physical systems (VCPSs) and the growing popularity of various services or applications for vehicles, cellular networks are being severely overloaded. Offloading mobile data traffic through Wi-Fi or a vehicular ad hoc network (VANET) is a promising solution for partially solving this problem because it involves almost no monetary cost. We propose combination optimization to facilitate mobile data traffic offloading in emerging VCPSs to reduce the amount of mobile data traffic for the QoS-aware service provision. We investigate mobile data traffic offloading models for Wi-Fi and VANET. In particular, we model mobile data traffic offloading as a multi-objective optimization problem for the simultaneous minimization of mobile data traffic and QoS-aware service provision; we use mixed-integer programming to obtain the optimal solutions with the global QoS guarantee. Our simulation results confirm that our scheme can offload mobile data traffic by up to 84.3% while satisfying the global QoS guarantee by more than 70% for cellular networks in VCPSs.

Introduction

In recent years, vehicular cyber-physical systems (VCPSs) have been proposed to exploit the latest advances in sensing, computing, communications, and networking technologies to improve the intelligence, safety, efficiency, resiliency, and environmental compatibility of transportation systems  [1], [2].

With the rapid development of VCPSs, an increasing number of vehicles with wireless communication capabilities can connect to Wi-Fi or cellular networks, thereby gaining access to various mobile services or applications such as safety applications, driving assistance, and multimedia content sharing. Owing to the widespread use of vehicles and growing service demands, mobile data traffic is increasing at an unexpected rate in VCPSs. According to the latest Cisco forecasts  [3], global mobile traffic will increase tenfold between 2014 and 2019, and monthly global mobile data traffic will surpass 24.3 exabytes by 2019. Because of this exponential growth in mobile data traffic, cellular networks will be overloaded. In particular, during peak-hour traffic in urban areas, service provision will suffer from extreme performance hits in terms of limited network bandwidth, dropped calls, and unreliable coverage.

Rapid growth in mobile data traffic leads to a sharp drop in the quality of service provision  [4]. To provide effective quality of service (QoS), the most straightforward solution is to increase the cellular network capacity by adding more base stations with a smaller cell size, such as picocells and femtocells, or by deploying 4G networks  [5]. Hence, service providers around the world are busy rolling out 4G networks to meet the growing end-user demand for better QoS such as higher bandwidth and faster connectivity on the go. However, although 4G or LTE networks provide better QoS, when vehicular capabilities are combined with these networks, the widespread adoption of advanced vehicular applications, such as video sharing, cloud applications, and big data services, will further increase mobile data traffic. For example, in 2013, a 4G connection generated 14.5 times more traffic on average than a non-4G connection  [3]. Therefore, even if the capacity of existing networks is enhanced, the future demands of users and applications will quickly exceed the network capacity. Simply using cellular networks for vehicular Internet access may aggravate the overload problem and degrade the service performance of both non-vehicular and vehicular users  [6]. Thus, rapid growth in mobile data traffic leads to a growing need for offloading solutions, and it becomes essential for operators to provide various offloading schemes in VCPSs.

There are two types of schemes for achieving mobile data traffic offloading: Wi-Fi offloading  [5], [6], [7] and vehicular ad hoc network (VANET) offloading  [8], [9], [10]. Cellular network congestion can be alleviated by delivering data originally targeted for cellular networks by Wi-Fi (i.e., Wi-Fi offloading)  [6]. It has been shown that approximately 65% of cellular traffic can be offloaded by merely switching from cellular networks to Wi-Fi when Wi-Fi connectivity is available  [5], [6], [11]. In urban environments, vehicles with low mobility signal nearby Wi-Fi access points (APs) when traveling along a road so that cellular traffic can be delivered to the vehicles through the drive-thru Internet  [12] in an opportunistic manner. This opportunistic Wi-Fi offloading offers unique features in VCPSs. At present, vehicles are ranked third, behind homes and offices, in terms of places where people spend the most time daily  [13]. Moreover, vehicles are becoming increasingly intelligent and connected, as they are equipped with on-board units (OBUs). According to forecasts of the European Telecommunications Standards Institute  [14], by 2027, nearly all vehicles will be equipped with OBUs. OBUs are devices that provide vehicle-to-vehicle (V2V) links and vehicle-to-infrastructure (V2I) communications, where Wi-Fi APs and base stations of cellular networks belong to the infrastructure. Moreover, a dedicated frequency band, 5.86 to 5.92 GHz, has been allocated for IEEE 802.11p vehicular communications. Therefore, considering VANET for offloading cellular traffic (i.e., VANET offloading) represents an attractive solution.

Using the above-mentioned technology as a foundation, some notable studies  [5], [6], [7], [8], [9], [10] have focused on offloading mobile data traffic. Although these offloading schemes can significantly reduce the mobile data traffic in cellular networks, they are not sufficiently efficient for VCPSs owing to two factors. First, most current studies focus on Wi-Fi offloading or VANET offloading and assume homogeneous offloading capacity for each Wi-Fi AP and vehicle in VCPSs. At first glance, this assumption appears to be reasonable, but it is unreasonable in a VCPS environment because different vehicles or newly deployed Wi-Fi APs are usually added to VCPSs to provide services; thus, different vehicular and Wi-Fi configurations form a heterogeneous mobile data traffic offloading environment. Second, an excellent mobile data traffic-offloading scheme satisfies the maximum traffic offloading requirements for service provision and also guarantees QoS. Therefore, QoS also plays an important role in determining the success or failure of offloading schemes. However, traditional offloading schemes mainly focus on the local QoS guarantee of mobile data traffic offloading and rarely consider the global QoS guarantee of service provision. The local QoS guarantee of mobile data traffic offloading is to offload traffic from a node with its QoS requirements independently on the other nodes. Even if the local QoS guarantee approach is useful in Wi-Fi environments, it is not suitable for mobile data traffic offloading with end-to-end QoS constraints (e.g., maximum total response time) because such global constraints cannot be verified locally. The global QoS guarantee approach aims at solving the problem on the composite mobile data traffic offloading level. The aggregated QoS values of all possible nodes from Wi-Fi or VANET are computed and the node combination that maximizes mobile data traffic offloading while satisfying global constraints is selected. Hence, finding a tradeoff between mobile data traffic offloading and the global QoS guarantee is very important for VCPSs.

In this paper, we propose an alternative mobile data traffic offloading approach for VCPSs. In contrast to the previous studies described above, our study removes the assumption of offloading capacity homogeneity and considers the global QoS guarantee of QoS-aware service provision. First, we establish the Wi-Fi and VANET offloading models and quantify the offloading capacity of each Wi-Fi AP or vehicle. Then, mobile data traffic offloading is formulated as a multi-objective combinatorial optimization problem that aims to simultaneously optimize possibly conflicting objectives. The objectives include making efficient use of multidimensional resources, satisfying the global QoS guarantee, and reducing mobile data traffic. A mixed-integer programming search method is used to efficiently search for global optimal solutions within a reasonable runtime. We implement our approach and compare it with the other approaches using the QualNet simulator.1 The simulation results indicate that our approach significantly reduces mobile data traffic while satisfying the global QoS guarantee in VCPSs.

The remainder of this paper is organized as follows. Section  2 reviews related studies. Section  3 establishes the Wi-Fi and VANET offloading models for VCPSs, quantifies their offloading capacity, and presents the problem statement. Section  4 describes the proposed approach to mobile data traffic offloading for QoS-aware service provision in VCPSs. Section  5 presents performance evaluations for comparing our solution against existing solutions. Finally, Section  6 summarizes our findings and concludes the paper.

Section snippets

Related work

Some notable schemes have been proposed to provide efficient mobile data traffic offloading for VCPSs.

From the Wi-Fi offloading perspective, Wi-Fi is recognized as one of the primary offloading technologies  [11]. P. Deshpande et al.  [15] proposed a prefetching mechanism motivated by the effective prediction of the mobility and connectivity of vehicles; the data can be redundantly prefetched by subsequent Wi-Fi APs. Subsequently, L. Kyunghan et al.  [16] presented a delayed offloading

Wi-Fi and VANET offloading models

Unlike traditional approaches, our approach takes QoS into consideration, i.e., QoS as a constraint should be guaranteed when we use Wi-Fi and VANET to offload mobile data traffic.

Recently, as an increasing number of Wi-Fi APs are deployed over roads in VCPSs, vehicles have greater opportunities to access mobile services via Wi-Fi on the go. The communication region of a Wi-Fi AP is the geographical region where vehicles can send and receive service from the Wi-Fi. In the communication region,

System model

As shown in Fig. 1, the system model is based on a VCPS architecture that consists of three systems: Wi-Fi APs, an LTE Advanced infrastructure, and a VANET network. The architecture includes a VANET network that consists of vehicles moving on a road where Wi-Fi APs are deployed. The geographical region of a Wi-Fi AP in which vehicles are moving and exchanging information with this Wi-Fi AP is referred to as ROW. The geographical region of a base station in which vehicles are moving and

Performance evaluation

We evaluated our Wi-Fi and VANET offloading models via numerical analysis and verified their effectiveness. We also implemented our offloading approach and compared it with other approaches based on the QualNet simulator.

Conclusions

Although existing offloading schemes can significantly reduce mobile data traffic in cellular networks, they are not sufficiently efficient for homogeneous vehicles and global QoS guarantee in VCPSs. In this study, first, we established the Wi-Fi and VANET offloading models and quantified the offloading capacity of each Wi-Fi AP or vehicle. Next, we derived a formulation to evaluate the potential of Wi-Fi and VANET to offload mobile data traffic. Then, mobile data traffic offloading was

Acknowledgments

This work is supported by the National Natural Science Foundation of China under Grant Nos. 61202435 and 61272521 and the Natural Science Foundation of Beijing under Grant No. 4132048.

Shangguang Wang is an associate professor at the State Key Laboratory of Networking and Switching Technology, Beijing University of Posts and Telecommunications. He received his Ph.D. degree in computer science at Beijing University of Posts and Telecommunications of China in 2011. His Ph.D. thesis was awarded as an outstanding doctoral dissertation by BUPT in 2012. His research interests include Service Computing, Mobile Services, and QoS Management.

References (22)

  • L. Xu et al.

    Toward effective service scheduling for human drivers in vehicular cyber-physical systems

    IEEE Trans. Parallel Distrib. Syst.

    (2012)
  • L. Uichin et al.

    FleaNet: A virtual market place on vehicular networks

    IEEE Trans. Veh. Technol.

    (2010)
  • Cisco Visual Networking Index: Global Mobile Data Traffic Forecast Update 2014–2019 White Paper, in, Cisco Systems,...
  • S. Wang et al.

    Efficient service selection in mobile information systems

    Mobile Inf. Syst.

    (2014)
  • F. Rebecchi et al.

    Data offloading techniques in cellular networks: A survey

    IEEE Commun. Surv. Tutor.

    (2015)
  • C. Nan, L. Ning, Z. Ning, X.S. Shen, J.W. Mark, Opportunistic WiFi offloading in vehicular environment: a queueing...
  • K. Xin, C. Yeow-Khiang, S. Sumei, Mobile data offloading through a third-party WiFi access point: an operator’s...
  • L. Yong et al.

    Coding or not: Optimal mobile data offloading in opportunistic vehicular networks

    IEEE Trans. Intell. Transp. Syst.

    (2014)
  • G. El Mouna Zhioua, Z. Jun, H. Labiod, N. Tabbane, S. Tabbane, VOPP: a VANET offloading potential prediction model, in:...
  • G. El Mouna Zhioua, H. Labiod, N. Tabbane, S. Tabbane, A traffic QoS aware approach for cellular infrastructure...
  • A. Aijaz et al.

    A survey on mobile data offloading: Technical and business perspectives

    IEEE Wirel. Commun.

    (2013)
  • Cited by (77)

    • ANDROIDOFF:Offloading android application based on cost estimation

      2019, Journal of Systems and Software
      Citation Excerpt :

      However, this paper mainly focuses on determining which parts shall be offloaded in mobile edge computing, and the two issues above are orthogonal to the problem in this paper. Some related approaches can be introduced to reduce this threat, such as supporting multi-user cases via game-theoretic model (Chen, 2014; Chen et al., 2016) and supporting complex mobility models via other offloading decision algorithms (Wang et al., 2016b; Lei et al., 2016). The threat to external validity also includes the selected calibration dataset in RQ3.

    • Vehicular Communication Network Enabled CAV Data Offloading: A Review

      2023, IEEE Transactions on Intelligent Transportation Systems
    • IoT Service Runtime Fault Tolerance Mechanism Based on Flink Dynamic Checkpoint

      2023, Communications in Computer and Information Science
    • Target node selection for data offloading in partially connected vehicular ad hoc networks

      2023, International Journal of Ad Hoc and Ubiquitous Computing
    View all citing articles on Scopus

    Shangguang Wang is an associate professor at the State Key Laboratory of Networking and Switching Technology, Beijing University of Posts and Telecommunications. He received his Ph.D. degree in computer science at Beijing University of Posts and Telecommunications of China in 2011. His Ph.D. thesis was awarded as an outstanding doctoral dissertation by BUPT in 2012. His research interests include Service Computing, Mobile Services, and QoS Management.

    Tao Lei received the B.Eng. and M.Eng. degrees from Hubei University for Nationalities and North China University of Water Resource and Electric Power, in 2009 and 2013, respectively. He is currently working toward the Ph.D. degree at the Beijing University of Posts and Telecommunications. His research interests include VANET and Multi-agent Systems.

    Lingyan Zhang received an M.E. degree in computer science and technology from the Institute of Network Technology, Beijing University of Posts and Telecommunications, in 2012. Currently, she is a Ph.D. candidate at the State Key Laboratory of Networking and Switching Technology, Beijing University of Posts and Telecommunications. Her research interests include VANET and Multi-agent Systems.

    Ching-Hsien Hsu is a professor in the department of computer science and information engineering at Chung Hua University, Taiwan. His research includes high performance computing, cloud computing, parallel and distributed systems, and ubiquitous/pervasive computing and intelligence. He has been involved in more than 100 conferences and workshops as various chairs and more than 200 conferences/workshops as a program committee member. He is the editor-in-chief of an international journal on Grid and High Performance Computing and has served on the editorial board for approximately 20 international journals.

    Fangchun Yang received his Ph.D. degree in communication and electronic systems from Beijing University of Posts and Telecommunication in 1990. He is currently a professor at the Beijing University of Posts and Telecommunication, China. He has published 6 books and more than 80 papers. His current research interests include network intelligence, services computing, communications software, soft switching technology, and network security. He is a fellow of the IET.

    View full text