CAREER: Taming Wireless Devices Cross-Layer Errors with Assistive Networked Edges

Project Synopsis

      Wireless devices are inherently faulty due to imperfect electronics and noisy environments, which can result in multifaceted data errors in computing, caching, and communications (C3). These errors have been widely deemed harmful, so a great number of error-control mechanisms are built into wireless devices’ hardware and software layers to avoid errors. However, data errors can be benign or even beneficial to design objectives (e.g., perturbing gradients when training deep neural networks may help the model escape local optima) while existing error-control studies that either consider the “worst case” for absolute error removal or reactively adapt to the dynamics of errors are passive, non-functional, and often redundant.

      Figure 1 shows a demonstrative example of different wireless applications, for which existing designs adjust error-correction margins according to the naturally evolved error modalities (X axis). Clearly, existing designs either waste error-correction margins (yellow-shaded area for error-free applications) or leave random and non-usable errors (dashed area for error-tolerant applications). Motivated by this research gap, this project’s ideology of “turning waste into wealth” by proactively harvesting, rendering, and controlling errors offers an innovation that will push the frontier of the state of the art. This project proposes three research thrusts to achieve this vision.

      (1): Harvest and render errors on hierarchical layers of C3 (as rendering techniques);

      (2): Design an optimization and control framework to integrate errors for system gains (as control algorithms);

      (3): Design an error-embracing assistive networked edge to support the error-rendering technique (as networking protocols).

      In our prior studies [TMC'19, ICDCS'17], we adopted this idea in an OFDM system as shown in Figure 2. Specifically, IEEE 802.11a adopts a stair-case link adaptation technique in accordance to the measured SNR. To cope with the worst-case channel conditions, there always exists unused channel coding capability. In light of it, we intentionally render some symbol errors by nullifying the transmit power of a selective set of OFDM sub-carriers - the one that inherently suffers from strong frequency selective fading. These "silence symbols'' are viewed by the receiver as erasure errors, which can be corrected by bit interleaving and residual coding capability. Moreover, the temporal-spectral distribution of “silence symbols” can be encoded to carry extra information for implicit communications, as shown in Figure 2(b). Based on this design, we showed significant performance gains in throughput and power saving simultaneously.

Personnel and Collaborators

  • PhD students
  • Steven Puckett (Fall 2021 - )
    MS., Electrical Engineering, University of Alabama at Birmingham
    Email: scp0019 at uah dot edu
    Role: protocol design in edge networks for low-power IoT
    * co-advise with Dr. Thomas Morris at the University of Alabama in Huntsville

    Punya Satish Gouda (Fall 2021 - )
    MS., Electrical Engineering, University of Alabama in Huntsville
    Email: pgouda at ncsu dot edu
    Role: error taming techniques in the wireless physical layer

  • Collaborators
  • Dr. Na Gong
    W. Nicholson Associate Professor
    Department of Electrical and Computer Engineering
    University of South Alabama
    Email: nagong at southalabama dot edu

    Research Progress

  • Publications (underscored are my students)
      [1] Privacy By Memory Design: Visions and Open Problems
              J. Liu, N. Gong, to appear at IEEE MICRO (IEEE MICRO).
      [2] A Secure and Efficient Protocol for LoRa Using Cryptographic Hardware Accelerators
              S. Puckett, J. Liu, Y. Seong-Moo, and T. Morris, to appear at IEEE Internet of Things Journal (IEEE IoT-J).
      [3] Towards Anonymous yet Accountable Authentication for Public Wi-Fi Hotspot Access with Permissionless Blockchains
              Y. Niu, L. Wei, C. Zhang, J. Liu, and Y. Fang, IEEE Transactions on Vehicular Technology (IEEE TVT), Vol. 72, no. 3, pp. 3904-3913, 2023.
      [4] Privacy Preservation in Multi-Cloud Secure Data Fusion for Infectious-Disease Analysis
              J. Liu, C. Zhang, K. Xue, and Y. Fang, IEEE Transactions on Mobile Computing (IEEE TMC), Vol. 22, no. 7, pp. 4212-4222, 2023.
  • Education and Outreach Activities

    under development


    Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.