LightLLM: A Versatile Large Language Model for Predictive Light Sensing

We propose LightLLM, a model that fine tunes pre-trained large language models (LLMs) for light-based sensing tasks. It integrates a sensor data encoder to extract key features, a contextual prompt to provide environmental information, and a fusion layer to combine these inputs into a unified representation. This combined input is then processed by the pre-trained LLM, which remains frozen while being fine-tuned through the addition of lightweight, trainable components, allowing the model to adapt to new tasks without altering its original parameters. This approach enables flexible adaptation of LLM to specialized light sensing tasks with minimal computational overhead and retraining effort. We have implemented LightLLM for three light sensing tasks: light-based localization, outdoor solar forecasting, and indoor solar estimation. Using real-world experimental datasets, we demonstrate that LightLLM significantly outperforms state-of-the-art methods, achieving 4.4x improvement in localization accuracy and 3.4x improvement in indoor solar estimation when tested in previously unseen environments. We further demonstrate that LightLLM outperforms ChatGPT-4 with direct prompting, highlighting the advantages of LightLLM's specialized architecture for sensor data fusion with textual prompts.

Mitigating Automotive Radar Interference using Onboard Intelligent Reflective Surface

The use of automotive radars is gaining popularity as a means to enhance a vehicle's sensing capabilities. However, these radars can suffer from interference caused by transmissions from other radars mounted on nearby vehicles. To address this issue, we investigate the use of an onboard intelligent reflective surface (IRS) to artificially increase a vehicle's effective radar cross section (RCS), or its "electromagnetic visibility." Our proposed method utilizes the IRS's ability to form a coherent reflection of the incident radar waveform back towards the source radar, thereby improving radar performance under interference. We evaluated both passive and active IRS options. Passive IRS, which does not support reflection amplification, was found to be counter-productive and actually decreased the vehicle's effective RCS instead of enhancing it. In contrast, active IRS, which can amplify the reflection power of individual elements, effectively combats all types of automotive radar interference when the reflective elements are configured with a 15-35 dB reflection gain.

A New Architecture for Energy Efficient Fault Detection Using Energy Harvesters

The current battery-powered fault detection system for vibration monitoring has a rather limited lifetime. This is because the high-frequency sampling (typically tens of kilo-Hertz) required for vibration monitoring results in high energy consumption in both the analog-to-digital (ADC) converter and wireless transmissions. This paper proposes a new fault detection architecture that can significantly reduce the energy consumption of the ADC and wireless transmission. Our inspiration for the new architecture is based on the observation that the many tens of thousand of data samples collected for fault detection are ultimately transformed into a small number of features. If we can generate these features directly without high frequency sampling, then we can avoid the the energy cost for ADC and wireless transmissions. We propose to use piezoelectric energy harvesters (which can be designed to have different frequency responses) and integrators to obtain these features in an energy-efficient manner. By using a publicly available data set for ball bearing fault detection (which was originally sampled at 51.2kHz) and piezoelectric energy harvester models, we can produce features, which when sampled at 0.33Hz, give a fault detection accuracy of 89% while reducing the sampling requirement by 4 orders-of-magnitude.

Path Planning for the Dynamic UAV-Aided Wireless Systems using Monte Carlo Tree Search

For UAV-aided wireless systems, online path planning attracts much attention recently. To better adapt to the real-time dynamic environment, we, for the first time, propose a Monte Carlo Tree Search (MCTS)-based path planning scheme. In details, we consider a single UAV acts as a mobile server to provide computation tasks offloading services for a set of mobile users on the ground, where the movement of ground users follows a Random Way Point model. Our model aims at maximizing the average throughput under energy consumption and user fairness constraints, and the proposed timesaving MCTS algorithm can further improve the performance. Simulation results show that the proposed algorithm achieves a larger average throughput and a faster convergence performance compared with the baseline algorithms of Q-learning and Deep Q-Network.

Deep Learning for Radio-based Human Sensing: Recent Advances and Future Directions

While decade-long research has clearly demonstrated the vast potential of radio frequency (RF) for many human sensing tasks, scaling this technology to large scenarios remained problematic with conventional approaches. Recently, researchers have successfully applied deep learning to take radio-based sensing to a new level. Many different types of deep learning models have been proposed to achieve high sensing accuracy over a large population and activity set, as well as in unseen environments. Deep learning has also enabled detection of novel human sensing phenomena that were previously not possible. In this survey, we provide a comprehensive review and taxonomy of recent research efforts on deep learning based RF sensing. We also identify and compare several publicly released labeled RF sensing datasets that can facilitate such deep learning research. Finally, we summarize the lessons learned and discuss the current limitations and future directions of deep learning based RF sensing.

Simultaneous Energy Harvesting and Gait Recognition using Piezoelectric Energy Harvester

Piezoelectric energy harvester, which generates electricity from stress or vibrations, is gaining increasing attention as a viable solution to extend battery life in wearables. Recent research further reveals that, besides generating energy, PEH can also serve as a passive sensor to detect human gait power-efficiently because its stress or vibration patterns are significantly influenced by the gait. However, as PEHs are not designed for precise measurement of motion, achievable gait recognition accuracy remains low with conventional classification algorithms. The accuracy deteriorates further when the generated electricity is stored simultaneously. To classify gait reliably while simultaneously storing generated energy, we make two distinct contributions. First, we propose a preprocessing algorithm to filter out the effect of energy storage on PEH electricity signal. Second, we propose a long short-term memory (LSTM) network-based classifier to accurately capture temporal information in gait-induced electricity generation. We prototype the proposed gait recognition architecture in the form factor of an insole and evaluate its gait recognition as well as energy harvesting performance with 20 subjects. Our results show that the proposed architecture detects human gait with 12% higher recall and harvests up to 127% more energy while consuming 38% less power compared to the state-of-the-art.