Current Retrieval-Augmented Generation systems use uniform processing, causing inefficiency as simple queries consume resources similar to complex multi-hop tasks. We present SymRAG, a framework that introduces adaptive query routing via real-time complexity and load assessment to select symbolic, neural, or hybrid pathways. SymRAG’s neuro-symbolic approach adjusts computational pathways based on both query characteristics and system load, enabling efficient resource allocation across diverse query types. By combining linguistic and structural query properties with system load metrics, SymRAG allocates resources proportional to reasoning requirements. Evaluated on 2,000 queries across HotpotQA (multi-hop reasoning) and DROP (discrete reasoning) using Llama-3.2-3B and Mistral-7B models, SymRAG achieves competitive accuracy (97.6–100.0% exact match) with efficient resource utilization (3.6–6.2% CPU utilization, 0.985–3.165s processing). Disabling adaptive routing increases processing time by 169–1151%, showing its significance for complex models. These results suggest adaptive computation strategies are more sustainable and scalable for hybrid AI systems that use dynamic routing and neuro-symbolic frameworks.

Internet of Vehicles (IoV) is growing rapidly with the potential to revolutionize transportation systems. Considering the promising future and potential contributions of IoV’s technology, it has attracted the attention of researchers, industry stakeholders, and potential intruders. However, the IoV’s network topological infrastructure faces several connectivity and communication challenges, along with security issues that are beyond the scope of current literature. Although each aspect and challenge has its own consequences, this work focuses on Physical Layer Security (PhyLaySec) threats, which are the most devastating because they undermine the trust of all stakeholders associated with this technology. In the literature, this topic is bearly focused, which demonstrates that the existing PhyLaySec countermeasures would not be able to counter future security challenges in IoV in terms of vehicle-to-vehicle (V2V) authentication, vehicle-to-infrastructure (V2I) authentication, vehicles-to-everything (V2X) authentication, etc., due to factors such as high vehicle mobility, dynamic network topologies, limited bandwidth, and ultra-fast communication. Therefore, this paper aims to provide a systematic review of state-of-the-art PhyLaySec techniques from 2017 to 2025, with a focus on their strengths and weaknesses. Through our review, we identify key open research questions that require further investigation to enhance the security of IoV’s technologies. Moreover, we highlight potential future research directions that aim to ensure the foolproof security of IoV technology with respect to underlined challenges. Finally, we acknowledge that this is the first paper to comprehensively address the topic of PhyLaySec of IoV technology, which makes it a valuable resource for researchers and professionals working in this field.

The aim of this study is to investigate into $\mu \mathrm{AFL}$, a non-intrusive, feedback-driven fuzzing framework, evaluated on Cortex M4 embedded systems and Unix platforms, focusing on the STM32F407VE Cortex M4 microcontroller. By leveraging the SEGGER J-Trace Pro for trace collection, it demonstrates $\mu$AFL’s utility beyond its traditional scope, showcasing its efficacy in both embedded and general-purpose computing environments. Our analysis, enriched by juxtaposing $\mu$AFL’s capabilities with traditional AFL, emphasizes the adaptability and effectiveness of fuzzing methodologies in firmware security enhancement. Furthermore, the study provides a deep understanding of fuzzing execution on different hardware, presenting an execution strategy for the STM32F407VE that highlights the framework’s potential in identifying vulnerabilities, evidenced by tests on specific firmware programs such as an LED blinking program integrated with semihosting breakpoints and ETM tracing. The use of uninitialized memory sections and strategically placed break-points offers significant insights into the firmware’s execution flow. The results of our comparative analysis clearly show that $\mu \mathrm{AFL}$ excels at uncovering vulnerabilities, reinforcing the need for evolving fuzzing methodologies to build stronger security systems for embedded devices. This contribution underscores the importance of refining fuzzing techniques to meet the intricate security demands of contemporary computing environments.

In this paper, we focus on addressing the challenges of detecting malicious attacks in networks by designing an advanced Explainable Intrusion Detection System (xIDS). The existing machine learning and deep learning approaches have invisible limitations, such as potential biases in predictions, a lack of interpretability, and the risk of overfitting to training data. These issues can create doubt about their usefulness, and transparency, and decrease the trust of involved stakeholders. To overcome these challenges, we propose an ensemble learning technique called the "EnsembleGuard". This approach uses the predicted outputs of multiple models, including tree-based (LightGBM, GBM, Bagging, XGBoost, CatBoost) and deep learning models such as neural network (LSTM (long short-term memory networks) and GRU (gated recurrent unit), to maintain a balance and achieve trustworthy results. Our work is unique because it combines both tree-based and deep learning models to design an interpretable and explainable meta-model through model distillation. By considering the predictions of all individual models, our neta-model effectively addresses key challenges, and ensures both explainable and reliable results. We evaluate our model using well-known datasets, including UNSW-NB15, NSL-KDD, and CIC-IDS-2017, to assess its reliability against various types of attacks. During analysis, we found that our model outperforms both tree-based models and other comparative approaches when it comes to different kinds of attack scenarios.