Browsing by Author "Yin, Zhimeng"

Now showing 1 - 1 of 1
  • Thumbnail Image
    Item
    IoT Networking: From Isolation to Collaboration
    (2020-09) Yin, Zhimeng
    Emerging Internet of Things (IoT) dramatically enriches every aspect of our daily life through pervasive IoT objectives. To satisfy the specific requirements of disparate IoT applications, researchers introduce various wireless protocols, such as WiFi, ZigBee, Bluetooth, and LTE. Designed independently, heterogeneous IoT protocols have incompatible Physical/MAC layers and lack direct communication. As a result, their spectrum competition leads to wireless interference that downgrades many important aspects of IoT performance such as delay, throughput, and energy consumption. This dissertation aims at enabling secure collaboration among the isolated IoT protocols for achieving synergic performance. First, it introduces cross-technology communication (CTC) for connecting heterogeneous IoT devices through C-Morse, a technique that slightly perturbs the transmission timing of existing WiFi packets for constructing dedicated and unique energy patterns. The energy patterns are detected through energy sensing, which is generally available across different types of IoT. To further boost the CTC throughput, it leverages all WiFi traffic (such as data packets and control frames) while being transparent for upper-layer applications without causing a significant delay. We evaluate C-Morse on testbeds such as WARP and MicaZ. Experiments demonstrate that C-Morse achieves a free side channel, with a throughput of 936bps and negligible delay for applications and end-users. Based on the direct connectivity offered by CTC, this dissertation presents ECC - a collaborative coexistence technique that uniquely enables explicit channel coordination among heterogeneities for improving spectrum efficiency. Specifically, ECC generates the guaranteed white space using WiFi CTS, which is then explicitly notified to ZigBee through CTC for immediate use. ECC's technical highlight lies in protecting low-power ZigBee communication from wireless interference while maintaining transparency to WiFi applications. This is achieved through dynamic adjustment of CTS duration with respect to traffic amount and spectrum availability. Our evaluation on commercial platforms shows that ECC achieves a 1.8x ZigBee packet reception ratio through collaborative coexistence. Despite the success of CTC, it also leads to potential security issues across heterogeneous systems. To ensure secure IoT applications, this dissertation further proposes CTCMon, a general detection framework for identifying CTC attacks based on the received payload at commodity devices. Since CTC techniques need to approximate the target waveform with heterogeneous devices, the transmitted waveform inevitably contains distinctive signal patterns. By examining these unique signal patterns like fingerprints, CTCMon reliably distinguishes CTC from regular wireless transmissions. Extensive experiments on multiple testbeds demonstrate the general applicability and reliability of CTCMon.