DSO National Laboratories

Accurate real-time Radio Frequency (RF) signal strength prediction is essential for optimizing wireless communication systems, enabling efficient network deployment, and supporting emerging technologies. In this paper, we propose a Generative Adversarial Network (GAN)-based approach for real-time prediction of real (or measured) RF signal strengths. In particular, we train a GAN model using both simulation and a limited number of real-world measurement data. Then, we use the GAN to predict the measured signal strength anywhere on the provided urban map. The proposed GAN model is applied to real-world urban environments and its performance is evaluated against actual measurement data. The results validate the effectiveness of the GAN-based approach in accurately predicting RF signal strength, demonstrating its potential for improving wireless network planning and optimization.

This paper presents the electrical design methods and results of a high-performance through silicon interposer. The interposer integrates an FPGA, several RF chips, and multiple components, leading to a multi-functional RF transceiver. The design details, including integration schematic, wafer electrical characterization, interconnection design, RF design of transmitter and receiver, signal/power integrity analysis, and measurement, are systematically elaborated.

Colorectal cancer is the most common cancer in Singapore and the third most common cancer worldwide. Blood in a person's stool is a symptom of this disease, and it is usually detected by the faecal occult blood test (FOBT). However, the FOBT presents several limitations - the collection process for the stool samples is tedious and unpleasant, the waiting period for results is about 2 weeks and costs are involved. In this research, we propose a simple-to-use, fast and cost-free alternative - a stool recognition neural network that determines if there is blood in one's stool (which indicates a possible risk of colorectal cancer) from an image of it. As this is a new classification task, there was limited data available, hindering classifier performance. Hence, various Generative Adversarial Networks (GANs) (DiffAugment StyleGAN2, DCGAN, Conditional GAN) were trained to generate images of high fidelity to supplement the dataset. Subsequently, images generated by the GAN with the most realistic images (DiffAugment StyleGAN2) were concatenated to the classifier's training batch on-the-fly, improving accuracy to 94%. This model was then deployed to a mobile app - Poolice, where users can take a photo of their stool and obtain instantaneous results if there is blood in their stool, prompting those who do to seek medical advice. As "early detection saves lives", we hope our app built on our stool recognition neural network can help people detect colorectal cancer earlier, so they can seek treatment and have higher chances of survival.

Wireless communication using radio frequency (RF) signal is developing rapidly in recent years. Many transmission systems are workingin RF signal range, so RF signal prediction is becoming an emerging topic. Many deep learning methods have been applied to do the prediction exactly. Generative Adversarial Network (GAN) has been the major breakthroughs in deep learning over the past few years, it is a neural network-based generative model which aims to mimic some underlying distribution given a dataset of samples. In this paper, we propose to use online GAN for RF signal prediction. In the future, we will apply online GAN in real scenarios.

Ensuring data privacy and protection has become paramount in the era of deep learning. Unlearnable examples are proposed to mislead the deep learning models and prevent data from unauthorized exploration by adding small perturbations to data. However, such perturbations (e.g., noise, texture, color change) predominantly impact low-level features, making them vulnerable to common countermeasures. In contrast, semantic images with intricate shapes have a wealth of high-level features, making them more resilient to countermeasures and potential for producing robust unlearnable examples. In this paper, we propose a Deep Hiding (DH) scheme that adaptively hides semantic images enriched with high-level features. We employ an Invertible Neural Network (INN) to invisibly integrate predefined images, inherently hiding them with deceptive perturbations. To enhance data unlearnability, we introduce a Latent Feature Concentration module, designed to work with the INN, regularizing the intra-class variance of these perturbations. To further boost the robustness of unlearnable examples, we design a Semantic Images Generation module that produces hidden semantic images. By utilizing similar semantic information, this module generates similar semantic images for samples within the same classes, thereby enlarging the inter-class distance and narrowing the intra-class distance. Extensive experiments on CIFAR-10, CIFAR-100, and an ImageNet subset, against 18 countermeasures, reveal that our proposed method exhibits outstanding robustness for unlearnable examples, demonstrating its efficacy in preventing unauthorized data exploitation.

The data that support the findings of this study are available from the corresponding author upon reasonable request.

The RF design for the wire bond and flip-chip Power Amplifier (PA) MMIC on the High Resistivity (HiR) Si-interposer are presented in this paper. The HiR Si-interposer is to support multiple off-the-shelf RF and Baseband ICs for complex advanced RF frontend modules using heterogeneous integration. A two-sided thermal cooling solution is adopted for this high-performance integration architecture. The high-power dissipation MMIC is integrated using flip-chip assembly so that the heat can be dissipated through the top heat spreader. For the other ICs, they are integrated using wire bond assembly so that their heat will be dissipated through the bottom HiR Si-interposer.

The ongoing COVID-19 pandemic has popularised the usage of Bluetooth signals from devices to assist the process of contact tracing. The purpose of our project was to utilise Bluetooth signals for an alternate purpose: to read devices’ Received Signal Strength Indicator (RSSI) values in order to create a robot that would assist in cluster detection and dispersion, hence easing the enforcement of Safe Management Measures (SMM).