Kun Woo Cho

Tomorrow’s massive-scale Internet-of-Things (IoT) sensor networks are poised to drive uplink traffic demand, especially in areas of dense deployment. To meet this demand, however, network designers leverage tools that often require accurate estimates of Channel State Information (CSI), which incurs a high overhead and thus reduces network throughput. Furthermore, the overhead generally scales with the number of clients, and so is of special concern in such massive IoT sensor networks. While prior work has used transmissions over one frequency band to predict the channel of another frequency band on the same link, this paper takes the next step in the effort to reduce CSI overhead: predict the CSI of a nearby but distinct link. We propose Cross-Link Channel Prediction (CLCP), a technique that leverages multi-view representation learning to predict the channel response of a large number of users, thereby reducing channel estimation overhead further than previously possible. CLCP’s design is highly practical, exploiting existing transmissions rather than dedicated channel sounding or extra pilot signals. We have implemented CLCP for two different Wi-Fi versions, namely 802.11n and 802.11ax, the latter being the leading candidate for future IoT networks. We evaluate CLCP in two large-scale indoor scenarios involving both line-of-sight and non-line-of-sight transmissions with up to 144 different 802.11ax users. Moreover, we measure its performance with four different channel bandwidths, from 20 MHz up to 160 MHz. Our results show that CLCP provides a 2x throughput gain over baseline and a 30% throughput gain over existing prediction algorithms.

We present Wall-Street, a smart surface designed for vehicles to boost 5G mmWave connectivity for passengers. It improves mmWave connections in three ways: (1) it steers outdoor mmWave signals into the vehicle, ensuring all users have coverage; (2) it enables the vehicle to receive data from the current cell while measuring signals from potential handover cells, allowing smooth transitions without interrupting service; (3) during handovers, it combines/splits signals from/to both the current and new cells, creating a make-before-break connection. Our demonstration shows Wall-Street's versatile signal manipulation abilities. These include steering single beams, simultaneously reflecting and transmitting beams for neighboring cell measurements with concurrent communication, and combining or splitting beams for seamless cell transitions.

We present the design and implementation of WaveFlex, the first smart surface that enhances Private 5G networks operating under the shared-license framework in the Citizens Broadband Radio Service frequency band. WaveFlex works in the presence of frequency diversity: multiple nearby base stations operating on different frequencies, as dictated by a Spectrum Access System coordinator. It also handles time dynamism: due to the dynamic sharing rules of the CBRS band, base stations occasionally switch channels, especially when priority users enter the network. Finally, WaveFlex operates independently of the network itself, not requiring access to nor modification of the gNB or UEs, yet it remains compliant with and effective on prevailing cellular protocols. We have designed and fabricated WaveFlex on a custom multi-layer PCB, software defined radio based network monitor, and supporting control software and hardware. Our experimental evaluation benchmarks operational Private 5G and LTE networks running at full line rate. In a realistic indoor office scenario, 5G experimental results demonstrate an 8.58~dB average SNR gain, and an average throughput gain of 10.77 Mbps under a single gNB, and 12.84 Mbps under three gNBs, corresponding to throughput improvements of 18.4% and 19.5%, respectively.

Tomorrow's massive-scale IoT sensor networks are poised to drive uplink traffic demand, especially in areas of dense deployment. To meet this demand, however, network designers leverage tools that often require accurate estimates of Channel State Information (CSI), which incurs a high overhead and thus reduces network throughput. Furthermore, the overhead generally scales with the number of clients, and so is of special concern in such massive IoT sensor networks. While prior work has used transmissions over one frequency band to predict the channel of another frequency band on the same link, this paper takes the next step in the effort to reduce CSI overhead: predict the CSI of a nearby but distinct link. We propose Cross-Link Channel Prediction (CLCP), a technique that leverages multi-view representation learning to predict the channel response of a large number of users, thereby reducing channel estimation overhead further than previously possible. CLCP's design is highly practical, exploiting existing transmissions rather than dedicated channel sounding or extra pilot signals. We have implemented CLCP for two different Wi-Fi versions, namely 802.11n and 802.11ax, the latter being the leading candidate for future IoT networks. We evaluate CLCP in two large-scale indoor scenarios involving both line-of-sight and non-line-of-sight transmissions with up to 144 different 802.11ax users and four different channel bandwidths, from 20 MHz up to 160 MHz. Our results show that CLCP provides a 2× throughput gain over baseline and a 30% throughput gain over existing prediction algorithms.

Electroencephalogram (EEG) activity from the brain is a promising biological marker that can serve as personal identification. Despite substantial efforts, it still remains unsolved problems to quantify EEG feature distribution for brain biometrics due to the complexity of the brain. In this study, we attempt to tackle EEG-based identification challenges by exploiting a novel distribution model. The distribution dissimilarity is measured by Mallows distance, a cluster similarity sensitive distance that is robust to signal noises. Specifically, EEG signals are decomposed through several statistical feature extraction methods, autoregressive (AR) model, discrete wavelet transform (DWT), and fast Fourier transform (FFT). With the dataset obtained from the real-world application, our proposed system achieves the f-score accuracy of 96.18% and half total error rate of 2.223%, which demonstrates the feasibility and effectiveness of utilizing EEG biometrics for personal identification applications.