Artificial Intelligence Generated Content (AIGC) services can efficiently satisfy user-specified content creation demands, but the high computational requirements pose various challenges to supporting mobile users at scale. In this paper, we present our design of an edge-enabled AIGC service provisioning system to properly assign computing tasks of generative models to edge servers, thereby improving overall user experience and reducing content generation latency. Specifically, once the edge server receives user requested task prompts, it dynamically assigns appropriate models and allocates computing resources based on features of each category of prompts. The generated contents are then delivered to users. The key to this system is a proposed probabilistic model assignment approach, which estimates the quality score of generated contents for each prompt based on category labels. Next, we introduce a heuristic algorithm that enables adaptive configuration of both generation steps and resource allocation, according to the various task requests received by each generative model on the edge.Simulation results demonstrate that the designed system can effectively enhance the quality of generated content by up to 4.7% while reducing response delay by up to 39.1% compared to benchmarks.

The impressive performance of ChatGPT and other foundation-model-based products in human language understanding has prompted both academia and industry to explore how these models can be tailored for specific industries and application scenarios. This process, known as the customization of domain-specific foundation models, addresses the limitations of general-purpose models, which may not fully capture the unique patterns and requirements of domain-specific data. Despite its importance, there is a notable lack of comprehensive overview papers on building domain-specific foundation models, while numerous resources exist for general-purpose models. To bridge this gap, this article provides a timely and thorough overview of the methodology for customizing domain-specific foundation models. It introduces basic concepts, outlines the general architecture, and surveys key methods for constructing domain-specific models. Furthermore, the article discusses various domains that can benefit from these specialized models and highlights the challenges ahead. Through this overview, we aim to offer valuable guidance and reference for researchers and practitioners from diverse fields to develop their own customized foundation models.

Website Fingerprinting (WF) attacks can effectively identify the websites visited by Tor clients via analyzing encrypted traffic patterns. Existing attacks focus on identifying different websites, but their accuracy dramatically decreases when applied to identify fine-grained webpages, especially when distinguishing among different subpages of the same website. WebPage Fingerprinting (WPF) attacks face the challenges of highly similar traffic patterns and a much larger scale of webpages. Furthermore, clients often visit multiple webpages concurrently, increasing the difficulty of extracting the traffic patterns of each webpage from the obfuscated traffic. In this paper, we propose Oscar, a WPF attack based on multi-label metric learning that identifies different webpages from obfuscated traffic by transforming the feature space. Oscar can extract the subtle differences among various webpages, even those with similar traffic patterns. In particular, Oscar combines proxy-based and sample-based metric learning losses to extract webpage features from obfuscated traffic and identify multiple webpages. We prototype Oscar and evaluate its performance using traffic collected from 1,000 monitored webpages and over 9,000 unmonitored webpages in the real world. Oscar demonstrates an 88.6% improvement in the multi-label metric Recall@5 compared to the state-of-the-art attacks.

We demonstrate mechanical threats classification including jackhammers and excavators, leveraging wavelet transform of MIMO-DFS output data across a 57-km operational network link. Our machine learning framework incorporates transfer learning and shows 93% classification accuracy from field data, with benefits for optical network supervision.

Dell revealed details for new models in its Inspiron and Latitude laptop lines at IFA 2024. The company announced in May that it would be powering several of its new devices with Qualcomm's Snapdragon X Plus, and now we have more information about how those processors will work in Dell's collection of Copilot PCs. The Inspiron 14 and the Latitude 5455 can have either the 8-core or 10-core Snapdragon X Plus processors. The 10-core option has clock speeds up to 3.4GHz while the newly announced 8-core goes up to 3.24 GHz. Both versions have the same NPU for AI tasks, which offers up to 45 TOPS (trillions of operations per second) in machine learning performance in support of Microsoft's Copilot AI platform.

As MTC networks continue to grow rapidly, managing and optimizing resources has become a crucial challenge for ensuring scalability. Additionally, the low-power/complexity of the MTC devices (MTD) present another challenge in terms of data management and network availability. These factors depend on the limited battery lifetime of the devices and their ability to implement algorithms, making energy efficiency and optimization critical enablers for future MTC networks (Shafiq et al., 2020). Characterizing and modeling MTC traffic is crucial for optimizing wireless IoT networks by tailoring management strategies to specific application needs (Sharma & Wang, 2019, 2018). With this, significant energy savings may be achieved, which is crucial due to the limited battery lifespan inherent in IoT networks, thereby improving network efficiency and scalability Shafiq et al. (2020). This may be enabled by the exploitation of accurate machine learning (ML)-based traffic predictors (Kato et al., 2016). For example, idle channel monitoring is responsible for wasting over half of the energy consumed in these networks (Mughees et al., 2020) while the results in (Ruiz-Guirola et al., 2022) indicated that up to 38% of the consumed energy could be saved by exploiting a prediction method when using sleep modes like wake-up radio or discontinuous reception. Unfortunately, the use of ML requires numerous labeled data obtained from extensive, large-scale dataset (Aldahiri et al., 2021). A significant stage of the ML process is the data analysis, which can be a difficult task.

The next-generation radio access network (RAN), known as Open RAN, is poised to feature an AI-native interface for wireless cellular networks, including emerging satellite-terrestrial systems, making deep learning integral to its operation. In this paper, we address the nonconvex optimization challenge of joint subcarrier and power allocation in Open RAN, with the objective of minimizing the total power consumption while ensuring users meet their transmission data rate requirements. We propose OpenRANet, an optimization-based deep learning model that integrates machine-learning techniques with iterative optimization algorithms. We start by transforming the original nonconvex problem into convex subproblems through decoupling, variable transformation, and relaxation techniques. These subproblems are then efficiently solved using iterative methods within the standard interference function framework, enabling the derivation of primal-dual solutions. These solutions integrate seamlessly as a convex optimization layer within OpenRANet, enhancing constraint adherence, solution accuracy, and computational efficiency by combining machine learning with convex analysis, as shown in numerical experiments. OpenRANet also serves as a foundation for designing resource-constrained AI-native wireless optimization strategies for broader scenarios like multi-cell systems, satellite-terrestrial networks, and future Open RAN deployments with complex power consumption requirements.

The rapid progress in Deep Learning (DL) and Large Language Models (LLMs) has exponentially increased demands of computational power and bandwidth. This, combined with the high costs of faster computing chips and interconnects, has significantly inflated High Performance Computing (HPC) construction costs. To address these challenges, we introduce the Fire-Flyer AI-HPC architecture, a synergistic hardware-software co-design framework and its best practices. For DL training, we deployed the Fire-Flyer 2 with 10,000 PCIe A100 GPUs, achieved performance approximating the DGX-A100 while reducing costs by half and energy consumption by 40%. We specifically engineered HFReduce to accelerate allreduce communication and implemented numerous measures to keep our Computation-Storage Integrated Network congestion-free. Through our software stack, including HaiScale, 3FS, and HAI-Platform, we achieved substantial scalability by overlapping computation and communication. Our system-oriented experience from DL training provides valuable insights to drive future advancements in AI-HPC.

Recent research in communications-aware robotics has been propelled by advancements in 5G and emerging 6G technologies. This field now includes the integration of Multi-Rotor Aerial Vehicles (MRAVs) into cellular networks, with a specific focus on under-actuated MRAVs. These vehicles face challenges in independently controlling position and orientation due to their limited control inputs, which adversely affects communication metrics such as Signal-to-Noise Ratio. In response, a newer class of omnidirectional MRAVs has been developed, which can control both position and orientation simultaneously by tilting their propellers. However, exploiting this capability fully requires sophisticated motion planning techniques. This paper presents a novel application of omnidirectional MRAVs designed to enhance communication security and thwart eavesdropping. It proposes a strategy where one MRAV functions as an aerial Base Station, while another acts as a friendly jammer to secure communications. This study is the first to apply such a strategy to MRAVs in scenarios involving eavesdroppers.

If you have an account with Comcast Xfinity, then you also have a year-long subscription to the Perplexity Pro AI answer engine. Perplexity announced the special deal on Threads. Perplexity Pro differs from the company's free option by allowing unlimited quick answers from a choice of AI models, including GPT-4o, Claude-3 and Sonar Large. Engadget hasn't reviewed the service, but if you're already paying for Xfinity, free seems like a good price for you to make up your own mind on its value. All you have to do to get your free year of Perplexity is to log into your Xfinity Rewards account and obtain a promo code.