This paper presents a novel graph neural network (GNN) based framework for efficient clock tree synthesis (CTS) optimization in complex System-on-Chip designs. As technology nodes advance to 5nm and below, traditional CTS methodologies face significant challenges in optimizing power, performance, and skew metrics while managing exponentially growing design complexity. We propose a specialized GNN architecture incorporating bidirectional message passing mechanisms and attention components to effectively capture critical clock network characteristics. The framework implements a multi-objective optimization approach that simultaneously addresses power consumption, insertion delay, and clock skew constraints through reinforcement learning techniques. Our hybrid methodology integrates GNN-based predictions with conventional CTS algorithms, achieving a synergistic workflow that preserves design rule compliance while enhancing optimization capabilities. Experimental evaluation across multiple benchmark circuits and industrial SoC designs demonstrates average reductions of 8.7% in clock power, 6.3% in maximum skew, and 1.8% in insertion delay compared to state-of-the-art commercial tools, while simultaneously reducing runtime by 56.2%. The performance advantages scale favorably with increasing design complexity, showing sublinear computational growth compared to the superlinear scaling of traditional methods. The framework demonstrates robust performance across diverse application domains including mobile processors, automotive controllers, and AI accelerators, validating its practical applicability in advanced technology nodes.

Show more

The rapid expansion of wireless communication networks, driven by the increasing demand for high-speed connectivity and the exponential growth of IoT devices, presents significant challenges to traditional signal processing methods. As Beyond 5G (B5G) and 6G technologies continue to evolve, wireless networks must address issues related to spectrum congestion, dynamic channel conditions, and interference management while maintaining low latency and high energy efficiency. Traditional signal processing approaches struggle to adapt to these dynamic environments, necessitating AI-driven adaptive signal processing frameworks. This study investigates the integration of artificial intelligence (AI) and machine learning (ML) in adaptive signal processing, focusing on Blind Spot Awareness Sensing (BSS), Edge Learning (EL), and Radio Frequency (RF) signal reflection. By using unsupervised learning for blind spectrum sensing, federated learning for distributed optimization, and AI-driven RF reflection techniques for wireless sensing, it is demonstrated that AI models enhance detection precision, optimize spectrum utilization, and improve anti-interference performance.

Show more

Integrating PLCs with emerging IoT technologies for industrial automation has transformed the manufacturing ecosystems towards a smarter and data-driven one. This systematic review explores the synergistic potential of IoT's connectivity and PLCs' reliability in modern industrial settings. It analyses trends such as edge computing, AI-driven analytics and digital twins for technical challenges and proposes future directions using blockchain integration and 5G-enabled automation to support them. This review synthesizes academic literature, industry case studies and technological frameworks to outline a roadmap for resilient, efficient and adaptable industrial systems.

Show more

The hybrid algorithm model developed in this study successfully broke the "black box effect" of cross-border capital flows by integrating the three engines of random forest, support vector machine, and LSTM neural network. Based on full-cycle data of the S&P 500, FTSE 100, and Shanghai Composite indices from 2019 to 2024, the model achieved an 85% response rate in warning of volatility in Southeast Asian emerging markets, 37 percentage points lower than the forecast error of the traditional ARCH-GARCH model. The specially designed timing analysis module successfully captured the abnormal signal of the daily withdrawal of $12.7 billion of northbound funds during the global market meltdown in March 2020, and issued a liquidity depletion warning 18 hours in advance by analyzing the dynamic correlation between the VIX fear index and the offshore RMB exchange rate.

Show more