LI

NS

research
group

Learning, Incentives, and Optimization
for Networked Systems

Our project ``Towards a Marketplace of Networked Services for the Next Billion Devices'' aims to enable cross-device applications by developing mechanisms that incentivize devices to cooperate with each other. For instance, external sensors may use cellular and Wi-Fi networks to send measurements to a group of smartphones running augmented reality applications; these smartphones can then work together to analyze the sensor data and display the right information to their users. This page gives an overview of the research and education initiatives that are a part of this project, specifically summarizing their goals, activities, and outputs. This work is funded by Carlee's NSF CAREER award, which has supported Carlee, Xiaoxi, Madhu, Ethan, and Taejin.

Intellectual Merit: Research Initiatives

Goals of this work: The overall intellectual goal of this project is to develop a theory of marketplaces for networked services, focusing in particular on markets for network connectivity and computing resources. Our current activities are focused on three of the main challenges in enabling cross-device applications: building infrastructure for users to respond to such needs, the dynamic nature of device needs and the mobile nature of today's computing devices. When devices like smartphones move, it is difficult to coordinate them as their value to an application may depend on their current location. Some of our most recent publications consider how users should pay for dynamic amounts of resources that match their demands at different times (INFOCOM 2019, JSAC 2019), the mechanics of how users could easily implement those decisions (ICC Workshop 2021, ICDCS demo 2020), how applications with relatively static demands can take advantage of dynamically priced resources (AISTATS 2021, ToN 2020 and 2021, JSAC 2019), incentivizing vehicles or other mobile devices to move to locations where they will be most useful (HotMobile 2022, SECON 2021, IoTJ 2020, TMC 2019), and the utility of mobile devices for collecting or caching data (INFOCOM 2020, INFOCOM 2019).

Network and computing infrastructure
    A. F. Baarzi, G. Kesidis, C. Joe-Wong and M. Shahrad, On Merits and Viability of Multi-Cloud Serverless, ACM Symposium on Cloud Computing 2021 (vision paper).
    M. Harishankar, J. Zuo, S. V. Iyer, P. Tague and C. Joe-Wong, Datanet: Enabling Seamless, Metered and Trusted Internet Connectivity without Subscriptions, IEEE ICC 2021 Workshop Towards Standardized Secured IoT B5G Networking - Artificial Intelligence and Blockchain (invited paper).
    X. Li, S. Gomena, L. Ballard, J. Li, E. Aryafar and C. Joe-Wong, A Community Platform for Research on Pricing and Distributed Machine Learning, demo at IEEE ICDCS 2020.
Dynamic application demands
    H. Jiang, X. Zhang and C. Joe-Wong, DOLL: Distributed OnLine Learning using Preemptible Cloud Instances, MAMA 2022, held in conjunction with ACM SIGMETRICS/IFIP Performance 2022.
    X. Zhang, J. Wang, L.-F. Lee, T. Yang, A. Kalra, G. Joshi and C. Joe-Wong, Machine Learning on Volatile Instances, IEEE/ACM Transactions on Networking, 2021. Preliminary version appeared in IEEE INFOCOM 2020.
    Y. Ruan, X. Zhang, S.-C. Liang and C. Joe-Wong, Towards Flexible Device Participation in Federated Learning, AISTATS 2021.
    S. Liu, C. Joe-Wong, J. Chen, C. G. Brinton, C. W. Tan and L. Zheng, Economic Viability of a Virtual ISP, IEEE/ACM Transactions on Networking, 2020. Preliminary version appeared in IEEE INFOCOM 2017.
    Y. Jiang, M. Shahrad, D. Wentzlaff, D. H. K. Tsang and C. Joe-Wong, Burstable Instances for Clouds: Performance Modeling, Equilibrium Analysis, and Revenue Maximization, IEEE/ACM Transactions on Networking, 2020. Preliminary version appeared in IEEE INFOCOM 2019. This work was featured in a news story on CMU Silicon Valley's website.
    M. Harishankar, S. Pilaka, P. Sharma, N. Srinivasan, C. Joe-Wong, and P. Tague, Procuring Spontaneous Session-level Resource Guarantees for Real-time Applications: An Auction Approach, IEEE Journal on Selected Areas in Communications, 2019.
    S. Sen, C. Joe-Wong, S. Ha and M. Chiang, Time-Dependent Pricing for Multimedia Data Traffic: Analysis, Systems, & Trials, IEEE Journal on Selected Areas in Communications, 2019.
Incentivizing and evaluating device mobility
    S. Lin, Y. Yao, H. Y. Noh, P. Zhang and C. Joe-Wong, A Neural-Based Bandit Approach to Mobile Crowdsourcing, ACM HotMobile 2022.
    P. Kortoci, A. Mehrabi, C. Joe-Wong and M. Di Francesco, Incentivizing Opportunistic Data Collection for Time-Sensitive IoT Applications, IEEE SECON 2021.
    X. Chen, S. Xu, J. Han, H. Fu, X. Pi, C. Joe-Wong, L. Zhang, H. Noh and P. Zhang, PAS: Prediction Based Actuation System for City-scale Ride Sharing Vehicular Mobile Crowdsensing, IEEE Internet of Things Journal, 2020. Preliminary version appeared in the 4th International Science of Smart City Operations and Platforms Engineering (SCOPE) Workshop, co-located with CPS-IoT Week 2019.
    S. Xu, X. Chen, C. Joe-Wong, P. Zhang and H. Noh, iLoCuS: Incentivizing Vehicle Mobility to Optimize Sensing Distribution in Crowd Sensing, IEEE Transactions on Mobile Computing, 2019.
    Y. Ruan and C. Joe-Wong, On the Economic Value of Mobile Caching, IEEE INFOCOM 2020.
    P. Kortoci, L. Zheng, C. Joe-Wong, M. Di Francesco and M. Chiang, Fog-based Data Offloading in Urban IoT Scenarios, IEEE INFOCOM 2019.

Broader Impacts: Education and Outreach

Goals and Activities: The findings from this project will be integrated into an education plan that emphasizes the role of user and application needs in shaping the evolution of Internet-based technologies. So far, we have developed a new undergraduate/graduate course on machine learning that takes cross-device applications as a motivating example, and a workshop course for middle school girls that looks at how incentives can be used to impact energy usage in homes. The project has also supported ECOFEC (ECOnomics of Fog, Edge and Cloud) 2019, a workshop co-located with IEEE INFOCOM 2019 that Carlee co-chaired, BlockNet (ACM MobiHoc Workshop on Blockchain for Network Resource Sharing), a workshop co-located with ACM MobiHoc 2020 that Carlee and Madhu co-chaired, and a special issue in the IEEE Journal on Selected Areas of Communications (Smart Data Pricing for Next Generation Networks) for which Carlee served as an associate editor.

    IEEE Open Journal of the Communications Society, Special Issue on the Optimization and Economics of Fog/Edge Networks. Guest Editors: Carlee Joe-Wong, Eitan Altman, Luoyi Fu, Sangtae Ha and David Starobinski.
    ACM MobiHoc Workshop on Blockchain for Network Resource Sharing (BlockNet 2020).
    Introduction to Machine Learning for Engineers, CMU 18-461/661 (Spring 2022). First offered in fall 2018 by Carlee Joe-Wong and Virginia Smith.
    IEEE Journal on Selected Areas in Communications, Special Issue on Smart Data Pricing for Next Generation Networks. Guest Editors: Mung Chiang, Rachid El-Azouzi, Lin Gao, Jianwei Huang, Carlee Joe-Wong and Soumya Sen.
    Economics of Energy Choices, Summer Engineering Experience for Girls (SEE) workshop, July 2018.
    First International IEEE Workshop on the ECOnomics of Fog, Edge, and Cloud Computing (ECOFEC 2019), co-located with IEEE INFOCOM 2019.