Elsevier

Computer Networks

Volume 174, 19 June 2020, 107226
Computer Networks

Incentive mechanisms for mobile data offloading through operator-owned WiFi access points

https://doi.org/10.1016/j.comnet.2020.107226Get rights and content

Abstract

Due to the explosive growth of mobile data traffic, it has become a common practice for Mobile Network Operators (MNOs, also known as operators or carriers) to utilize cellular and WiFi resources simultaneously through mobile data offloading. However, existing offloading technologies are mainly established between operators and third-party WiFi resources, which cannot reflect users dynamic traffic demands. Therefore, MNOs have to design an effective incentive framework, encouraging users to reveal their valuations on resources. In this paper, we propose a novel bid-based Heterogeneous Resources Allocation (HRA) framework. It can enable operators to efficiently utilize both cellular and operator-own WiFi resources simultaneously, where the decision cost of user is strictly controlled. Through auction-based mechanisms, it can achieve dynamic offloading with awareness of users valuations. And the operator-domain offloading effectively avoids anarchy brought by users selfishness and lack of information. More specifically, HRA-Profit and HRA-Utility, are proposed to achieve the maximal profit and social utility, respectively. addition, based on Stochastic Multi-Armed Bandit model, the newly proposed HRA-UCB-Profit and HRA-UCB-Utility are able to gain near-optimal profit and social utility under incomplete user context information. All mechanisms have been proven to be truthful and satisfy individual rationality, while the achieved profit of our mechanism is within a bounded difference from the optimal profit. In addition, the trace-based simulations and evaluations have demonstrated that HRA-Profit and HRA-Utility increase the profit and social utility by up to 40% and 47%, respectively, compared with benchmarks. And the cellular utilization rate is kept at a favorable level under the proposed mechanisms. HRA-UCB-Profit and HRA-UCB-Utility restrict pseudo-regret ratios under 20%.

Introduction

With the explosive growth of intelligent mobile devices and bandwidth-consuming mobile applications, mobile data traffic has experienced dramatic growth in the past decade [1]. There exists a shortage of cellular capacity, especially in busy regions during peak periods. It is often too expensive or sometimes even impossible for Mobile Network Operators (MNOs) (we use the terms MNO, operator and carrier interchangeably in this paper) to deploy enough cellular resources that can meet the peak traffic demand [2], [3], [4]. Once excessive traffic exceeds the cellular capacity, it will introduce high congestion cost to MNOs [5], [6], [7]. Therefore, MNOs have to utilize other complementary technologies to enhance transmission capability.

To reduce the pressure on the cellular networks, WiFi has been widely used by mobile users to offload mobile data from cellular networks at specific scenarios. It has been envisioned that one of the future trends for MNOs is to utilize both licensed (e.g., cellular) and unlicensed (e.g., WiFi) spectrums in Heterogeneous Networks (HetNet) [8], [9], [10], [11], [12], [13]. For example, Bennis et al.  [8] constitute a cost-effective integration of multiple infrastructures, efficiently coping with peak traffic and heterogeneous Quality of Service (QoS) requirements. Unlicensed Long-Term Evolution (LTE-U) is also proposed by industries to introduce Long-Term Evolution (LTE) in unlicensed spectrums [14], which will improve the throughput of radio access networks.

WiFi can be divided into third-party WiFi and carrier WiFi (i.e., operator-owned WiFi). Third-party WiFi has provided widely deployed infrastructures for mobile data offloading. In contrast, infrastructures of carrier WiFi are deployed and operated by carriers. Existing offloading technologies [11], [12], [15], [16] mainly focus on how to utilize third-party WiFi, rather than carrier WiFi. In fact, carrier WiFi should also be explored and exploited considering the following advantages. First, many MNOs (e.g., AT&T, Verizon, China Mobile, and Vodafone) have widely deployed WiFi Access Points (APs) [17]. Most of these APs are deployed in busy regions, so they can offload the peak traffic from cellular networks. Second, operator-owned WiFi APs can help improve the QoS. During peak hours, third-party WiFi APs may also serve heavy traffic. In comparison, operators can reserve bandwidth on self-owned APs for peak-hour cellular offloading. In addition, the security and privacy issues [18] brought by third-party WiFi can also be solved by controlling the access to operator-owned WiFi APs. Moreover, third-party WiFi APs can also be utilized by MNOs to provide offloading services [11], [19], which can be regarded as the resources of MNOs. Therefore, we focus on how to utilize carrier WiFi resources from the perspective of economics, to alleviate the pressure on cellular networks.

To achieve the integrated utilization of cellular and carrier WiFi resources, user context information about their valuations on resources cannot be neglected. It is the foundation for achieving economic optimization targets [20], [21], like the maximal profit of the operator and social utility (i.e., the sum of utilities of all users). On the other hand, operator-dominant offloading (i.e., dynamic resource allocation is dominated by the operator), which has the ability to take into account user context information, can achieve better resource allocation performance, because the operators are better aware of network status. In addition, small-scale central control with users’ valuations cannot deal with the anarchy brought by users’ local views on the global network condition as well as their selfish behaviours, but also satisfy delay requirements. These relevant factors have been considered in our work, which are further verified through the trace-based evaluations.

However, several challenges are brought forth considering users’ valuations in heterogeneous resource utilization. First, incentive mechanisms are required to motivate users to reveal their true valuations on resources. Second, the operator profit and social utility must be considered under the premise of true users’ valuations. Third, the interactions among users and MNOs should minimize the costs of users.

To solve the above issues under such circumstances, we propose a novel bid-based operator-dominant mobile data offloading framework between cellular and carrier WiFi networks, to achieve the Heterogeneous Resource Allocation (HRA). In summary, the proposed HRA framework in this paper is mainly due to the following motivations.

  • Different from a simpler resource scheduler operated by the MNO alone, the newly proposed offloading mechanisms should enable users to express their willingness to use WiFi resources according to different scenarios.

  • In addition to alleviating the pressure of cellular networks, the newly proposed framework should also encourage users to reveal their true valuation on WiFi resources, so that operators can achieve economic optimization targets.

  • Different from allowing users to freely decide whether to use WiFi resources, the operator should achieve offloading on the basis of considering the global network condition, avoiding the anarchy brought by users local views.

Thus, MNOs can realize dynamic WiFi pricing and resource allocation, while mobile users can achieve dynamic data offloading according to their bids. More specifically, all auction mechanisms proposed in this paper are established between MNOs and mobile users, instead of MNOs and third-party resource owners. And it has been demonstrated through theoretical analysis that, they can solve the above challenges to encourage users to claim their true valuations and help MNOs make better use of heterogeneous resources.

The implementation of HRA framework needs the support of MNOs or regulators. For operators, the goal is to maximize their profits with some constraints such as traffic performance and user experience assurance. For regulators, the policy is formulated in order to maximize social utility. Thus, from the perspectives of both operators and regulators, two auction mechanisms, HRA-Profit and HRA-Utility, are designed to achieve the maximal profit and social utility, respectively. In addition, HRA-UCB-Profit and HRA-UCB-Utility are proposed to gain near-optimal profit and social utility under incomplete user information. The following summarizes the contributions of this paper.

  • Through offloading between cellular and carrier WiFi networks, the bid-based operator-dominant offloading framework for the first time takes users’ valuations on resources into consideration, and effectively avoids anarchy brought by users selfishness and lack of information.

  • To enable the newly proposed framework to be applied to more complex scenarios with incomplete information, HRA-UCB-Profit and HRA-UCB-Utility are proposed to optimize operator profit and social utility.

  • Compared with classical auction-based mechanisms, we cope with the challenges brought by HRA’s distinct features. And the proposed mechanisms have been proven to be truthful and satisfy individual rationality.

  • Through theoretical analysis, the profit of HRA-Profit has been proven to be within the bound of the optimal profit. And extensive evaluations demonstrate the efficiency of the proposed mechanisms, compared with benchmarks.

The rest of this paper is organized as follows. Section 2 builds the system model to formulate the problem. Section 3 proposes and theoretically analyzes the bid-based dynamic resource allocation framework. Furthermore, Section 4 proposes a model to analyze the resource allocation with incomplete information. And Section 5 sets up trace-based simulations and analyzes the performance of the proposed framework. Section 6 analyzes the implementation issues of the framework. Section 7 introduces the related work. Finally, Section 8 concludes the paper.

Section snippets

System model and problem formulation

The key problem is to determine the allocation of cellular and operator-owned WiFi resources, especially properly pricing WiFi resources, so as to achieve dynamic resource allocation with considering MNO’s profit, social utility, resource utilization efficiency and QoS. In this section, we first introduce the system model through an example, illustrated in Fig. 1. After explaining the relationship between the system participants and the main parameters through Fig. 1, we further provide a high

Bid-based resource allocation framework

We propose a bid-based resource allocation framework, as shown in Fig. 3. The allocation of cellular and WiFi resources is achieved through the operator-dominant mobile data offloading, i.e., the cellular data of some users is offloaded to WiFi networks through a decision-making auction. More specifically, we first summary the heterogeneous resource allocation, focusing on the distinctive features from classical auction-based allocation. We then discuss the trigger policy of HRA and how users

Resource allocation with incomplete information

As discussed in Section 3.3, the actual data-rates of users are unknown to MNOs when they determine the winners of the resource allocation. Instead, we use estimated data-rates (i.e., Eq. (12)), which will affect the precision of the optimization result. And the estimation factor β in Eq. (12) depends on the historical data. However, in the real-world wireless data market, there are inevitably some new users entering the market, and these users do not have sufficient historical data to support

Simulation and evaluation

In this section, we simulate the proposed mechanisms based on real trace data and evaluate their performances compared with that of benchmark resource allocation methods.

Implementation issues

In this section, we discuss the implementation issues of the proposed framework from multiple perspectives, which may provide some useful suggestions for framework deployment.

1) Controller: The controller involves three function modules: i) Information module obtains cellular load information from the Base Station Controller (BSC), user’s available networks from BSs and APs, user’s bids from APs, and etc.; ii) Computation module implements the auction mechanism and the WiFi Pricing algorithm;

Related work

To solve the conflicts between the shortage of cellular resources and the increasing mobile data demands, both industry and academia have proposed various solutions.

Temporal Dynamics: From operator’s perspective, cellular traffic obeys obvious diurnal and weekly patterns [24], [25], allowing researchers to propose dynamic time-dependent pricing strategies [4], [5], [25], [52], which encourage users to shift their delay-tolerant demands during peak hours to off-peak periods. With time-dependent

Conclusion

In this paper, we propose a novel bid-based operator-dominant offloading framework for the dynamic allocation of MNO’s heterogeneous wireless resources. The HRA framework not only enables operators to efficiently utilize both cellular and carrier WiFi simultaneously, but also encourages users to reveal their valuations on resources through auctions, without increasing users’ decision cost. Thus, the operator-dominant offloading can avoid anarchy brought by users selfishness and lack of

CRediT authorship contribution statement

Yi Zhao: Writing - original draft, Data curation, Software, Methodology, Conceptualization. Ke Xu: Conceptualization, Methodology, Writing - review & editing, Funding acquisition. Yifeng Zhong: Conceptualization, Methodology, Software. Xiang-Yang Li: Conceptualization, Writing - review & editing. Ning Wang: Conceptualization, Writing - review & editing. Hui Su: Writing - review & editing. Meng Shen: Writing - review & editing. Ziwei Li: Writing - review & editing.

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.

Acknowledgment

This work was in part supported by National Science Foundation for Distinguished Young Scholars of China with No. 61825204, No. 61625205, National Natural Science Foundation of China with No. 61932016, No. 61751211, No. 61520106007, No. 61972039, Beijing Outstanding Young Scientist Program with No. BJJWZYJH01201910003011, Beijing National Research Center for Information Science and Technology (BNRist) with No. BNR2019RC01011, Key Research Program of Frontier Sciences, CAS. No. QYZDY-SSW-JSC002,

Yi Zhao received the B. Eng. degree from the School of Software and Microelectronics, Northwestern Polytechnical University, Xi’an, China, in 2016. Currently, he is pursuing the Ph.D. degree in the Department of Computer Science and Technology of Tsinghua University, Beijing, China. His research interests include network economics, game theory, machine learning, social network, and recommended system. He is a student member of IEEE, and a student member of ACM.

References (66)

  • X. Gong et al.

    When social network effect meets congestion effect in wireless networks: data usage equilibrium and optimal pricing

    IEEE J. Select. Area Commun. (JSAC)

    (2017)
  • Cisco, Cisco visual networking index: forecast and trends, 2017-2022 white paper, 2019, (([Online], Available:...
  • A. Apostolaras et al.

    A mechanism for mobile data offloading to wireless mesh networks

    IEEE Trans. Wireless Commun. (TWC)

    (2016)
  • Y. Zhao et al.

    Variety Matters: A New Model for the Wireless Data Market under Sponsored Data Plans

    Proceedings of IEEE/ACM IWQoS

    (2019)
  • S. Sen et al.

    Time-Dependent pricing for multimedia data traffic: analysis, systems, and trials

    IEEE J. Sel. Area. Commun. (JSAC)

    (2019)
  • S. Ha et al.

    TUBE: time-dependent pricing for mobile data

    Proceedings of ACM SIGCOMM

    (2012)
  • R.T. Ma

    Usage-Based pricing and competition in congestible network service markets

    IEEE/ACM Trans. Netw. (ToN)

    (2016)
  • M. Zou et al.

    On optimal service differentiation in congested network markets

    IEEE/ACM Trans. Netw. (ToN)

    (2018)
  • Z. Lu et al.

    Easybid: enabling cellular offloading via small players

    Proceedings of IEEE INFOCOM

    (2014)
  • Y. He et al.

    On wifi offloading in heterogeneous networks: various incentives and trade-Off strategies

    IEEE Commun. Surv. Tutor.

    (2016)
  • Z. Zhou et al.

    Licensed and unlicensed spectrum allocation in heterogeneous networks

    IEEE Trans. Commun. (TC)

    (2017)
  • H. Yu et al.

    Mobile data offloading for green wireless networks

    IEEE Wireless Commun.

    (2017)
  • M. Li et al.

    Mobile data offloading with uniform pricing and overlaps

    IEEE Trans. Mobile Comput. (TMC)

    (2019)
  • Z. Zhou et al.

    Multi-Scale dynamic allocation of licensed and unlicensed spectrum in software-Defined hetnets

    IEEE Netw.

    (2019)
  • Qualcomm, LTE advanced in unlicensed spectrum, (([Online], Available:...
  • H. Shah-Mansouri et al.

    An incentive framework for mobile data offloading market under price competition

    IEEE Trans. Mobile Comput. (TMC)

    (2017)
  • J. Liu et al.

    An incentive mechanism combined with anchoring effect and loss aversion to stimulate data offloading in IoT

    IEEE Internet Thing J. (JIoT)

    (2019)
  • AT&T

    AT&T expands wi-Fi hotzone pilot project to additional cities

    AT&T Press Rel.

    (2010)
  • E.H.H. Kure et al.

    Distributed uplink offloading for IoT in 5G heterogeneous networks under private information constraints

    IEEE Internet Things J. (JIoT)

    (2018)
  • G. Iosifidis et al.

    A double-auction mechanism for mobile data-Offloading markets

    IEEE/ACM Trans. Netw. (ToN)

    (2015)
  • K. Xu et al.

    Can P2P technology benefit eyeball ISPs? A cooperative profit distribution answer

    IEEE Trans. Parallel Distrib. Syst. (TPDS)

    (2014)
  • K. Poularakis et al.

    Mobile data offloading through caching in residential 802.11 wireless networks

    IEEE Trans. Netw. Serv. Manag. (TNSM)

    (2016)
  • J. Lee et al.

    Economics of wifi offloading: trading delay for cellular capacity

    IEEE Trans. Wireless Commun. (TWC)

    (2014)
  • Y. Zhao et al.

    Understand love of variety in wireless data market under sponsored data plans

    IEEE J. Sel. Area Commun. (JSAC)

    (2020)
  • M.Z. Shafiq et al.

    Characterizing and modeling internet traffic dynamics of cellular devices

    Proceedings of ACM SIGMETRICS

    (2011)
  • L. Zhang et al.

    Time dependent pricing in wireless data networks: flat-rate vs. usage-based schemes

    Proceedings of IEEE INFOCOM

    (2014)
  • 3GPP, Access Network Discovery and Selection Function (ANDSF), 2015, (([Online], Available:...
  • V. Ganesan, Mobile device with automatic switching between cellular and WiFi networks, 2017. US Patent...
  • Y. Yang, X. Zhou, H. Jeong, J.S. Abuan, G. Johar, Y. Shen, Seamless video pipeline transition between WiFi and cellular...
  • P. Taylor

    AT&T Imposes Usage Caps on Fixed-Line Broadband

    Financ. Times,

    (2011)
  • L. Zheng et al.

    Customized data plans for mobile users: feasibility and benefits of data trading

    IEEE J. Select. Area. Commun.(JSAC)

    (2017)
  • L. Zheng et al.

    Optimizing data plans: usage dynamics in mobile data networks

    Proceedings of IEEE INFOCOM

    (2018)
  • V. Valancius et al.

    How many tiers? Pricing in the internet transit market

    Proceedings of ACM SIGCOMM

    (2011)
  • Cited by (6)

    Yi Zhao received the B. Eng. degree from the School of Software and Microelectronics, Northwestern Polytechnical University, Xi’an, China, in 2016. Currently, he is pursuing the Ph.D. degree in the Department of Computer Science and Technology of Tsinghua University, Beijing, China. His research interests include network economics, game theory, machine learning, social network, and recommended system. He is a student member of IEEE, and a student member of ACM.

    Ke Xu received his Ph.D. from the Department of Computer Science & Technology of Tsinghua University, Beijing, China, where he serves as a full professor. He has published more than 200 technical papers and holds 11 US patents in the research areas of next-generation Internet, Blockchain systems, Internet of Things, and network security. He is a member of ACM and senior member of IEEE. He has guest-edited several special issues in IEEE and Springer Journals. He is an editor of IEEE IoT Journal. He is Steering Committee Chair of IEEE/ACM IWQoS.

    Yifeng Zhong received his PhD degree in computer science and engineering from Tsinghua University, in 2016. He is now an expert in Migu Culture Technology Co. Ltd, Beijing, China.

    Xiang-Yang Li is a professor and executive dean at School of Computer Science and Technology at University of Science and Technology of China. He is an ACM Fellow, an IEEE Fellow and an ACM Distinguished Scientist. He was a professor at the Illinois Institute of Technology. Dr. Li received MS (2000) and PhD (2001) degree at Department of Computer Science from University of Illinois at Urbana-Champaign, a Bachelor degree at Department of Computer Science and a Bachelor degree at Department of Business Management from Tsinghua University, P.R. China, both in 1995. His research interests include mobile computing, cyber physical systems, security and privacy, and data sharing and trading. He and his students won eight best paper awards and one best demo/poster award. He has served as an editor of several international journals such as ACM/IEEE Transactions on Networking, IEEE TPDS, IEEE TMC and so on.

    Ning Wang is a Reader at the Institute for Communication Systems (5G Innovation Centre), University of Surrey, UK. He received his B.Eng (Honours) degree from the Changchun University of Science and Technology, P.R. China in 1996, his M.Eng degree from Nanyang University, Singapore in 2000, and his Ph.D. degree from the University of Surrey in 2004 respectively. His research interests mainly include mobile content delivery, context-aware networking, network resource management.

    Hui Su received the Ph.D. from the Department of Computer Science and Technology of Tsinghua University, Beijing, China, in 2017. His research interests include network economics and wireless network.

    Meng Shen received the B.Eng degree from Shandong University, Jinan, China in 2009, and the Ph.D degree from Tsinghua University, Beijing, China in 2014, both in computer science. Currently he serves in Beijing Institute of Technology, Beijing, China, as an associate professor. His research interests include privacy protection for cloud and IoT, blockchain applications, and encrypted traffic classification. He received the Best Paper Runner-Up Award at IEEE IPCCC 2014. He is a member of the IEEE.

    Ziwei Li received his B. Eng. degree and M. Eng. degree from the Department of Information Science and Technology at Beijing Forestry University, China, in 2011 and 2013, respectively. Currently, he is pursuing a Ph.D. degree in the Department of Computer Science and Technology at Tsinghua University, Beijing, China. His research interests include machine learning, wireless networks, and mobile computing.

    View full text