Chadi Assi

In Vehicular Fog Computing (VFC), the per-vehicle processing capacity and mobility dynamics exert climacteric impacts on the task offloading performance of in-proximity RoadSide Units (RSUs) in terms of critical Quality-of-Service (QoS) metrics (e.g., task admission and Head-of-Line (HoL) blocking probabilities as well as the task expiry probability and mean buffering delay). Precisely, RSU buffer exhaustion, RSUto- Vehicle (R2V) denial of task release and, hence, resulting increases in the average system response time are undesirable but often inevitable whenever incoming vehicles into the RSU's range suffer from limited task processing resources and/or happen to be navigating at high speeds; hence, limiting their capability to successfully process and complete tasks within their relatively short residency within the RSU's coverage range. Existing work have already highlighted the possible occurrence of the abovementioned R2V Task Offloading (R2VTO) shortcomings. However, very few publications have quite marginally quantified the impact of vehicle processing resource availabilities and mobility dynamics on the R2VTO QoS performance. Also, the vehicles' non-cooperativeness has been largely overlooked. This work aims at closing these gaps through the establishment of a theoretical stochastic framework, throughout which the task release feasibility is derived from first principles (as opposed to being exogenously pre-defined in prior work) and the above-mentioned QoS metrics are expressed as a function of the vehicles's processing capacity, mobility characteristics and cooperativeness probability. Extensive simulations are then conducted to verify the correctness, validity and accuracy of the derived expressions and evaluate the R2VTO's performance as a function of each of the considered metrics as well as existing scheduling disciplines.

The increasing sophistication and complexity of Advanced Persistent Threats (APTs) pose significant challenges to security practitioners. To proactively protect against these threats, security practitioners rely on the generation of testflows, structured sequences of actions designed to verify whether the tactics and behaviors of an APT are present within their organization. However, manually creating such testflows is time-consuming, error-prone, and highly dependent on expert knowledge. Moreover, existing automated approaches suffer from several limitations, including validity, efficiency, and insufficient domain adaptation. To address these challenges, this paper introduces CTI-RAGFlow, to automate the generation of relevant, valid, and effective testflows from unstructured threat reports tailored to specific organizational environments. CTI-RAGFlow introduces three key contributions: (i) a dual-ontology approach, that integrates both a system ontology representing the operational environment and a cybersecurity ontology capturing adversary tactics, techniques, and procedures, improving the precision and accuracy of generated testflows; (ii) a fact-based context retrieval mechanism that combines a hypergraph structured knowledge base with a Retrieval-Augmented Generation pipeline using Large Language Models; and (iii) a fully automated testflow generation process that minimizes manual effort, reduces human error, and facilitates the generation of valid testflow. We evaluate CTI-RAGFlow against three widely used LLM models (e.g., base and fine-tuned models) using publicly available CTI reports for three well-known APTs (e.g., APT41, APT29, APT28). The results show that CTI-RAGFlow outperforms the baselines in terms of semantic relevance, coverage, validity, and effectiveness in verifying multi-stage cyberattack scenarios.

This paper presents an experimental and simulationbased evaluation of two data center network (DCN) topologies: the widely adopted hierarchical Clos architecture and STRAT, a flat, expander-based topology designed around passive optical interconnects. While Clos offers proven scalability and performance, it incurs hardware complexity and suffers from congestion in oversubscribed scenarios. STRAT eliminates aggregation and spine layers entirely—using only Top-of-Rack (ToR) switches interconnected via static optical patch panels—to reduce cost, simplify deployment, and enhance path diversity. Our goal is to assess these topologies based on their inherent architectural properties—namely throughput, congestion resilience, scalability, and cost—without relying on congestion control protocols or centralized traffic engineering. To this end, we adopt simple forwarding schemes based purely on local information: ECMP for Clos, and ECMP with Dynamic Group Multipath (DGM) for STRAT. We evaluate both topologies on a physical testbed built from commercial Ethernet switches and further validate scalability through packet-level simulations of networks with up to 256 switches and 1,024 hosts using OMNeT++. We also introduce DEALER, a lightweight routing algorithm tailored to STRAT’s topology, and evaluate its effectiveness in dynamic conditions. Our results show that STRAT achieves up to 43% higher throughput and requires approximately 40% fewer switches than a comparable Clos topology. These gains are further supported by Load Area Under Curve (LAUC) analysis and congestion hotspot visualizations. Overall, our study highlights STRAT as a compelling and practical alternative to conventional DCN architectures, offering deployable scalability, improved performance under load, and reduced infrastructure cost.

The multi-base-station cooperative (MBC) ISAC mechanism is promising to satisfy the higher-quality sensing requirements of accuracy and coverage in 6G. STAR-RIS can help ISAC BSs provide cooperative sensing and communication (S&C) across full-space radio signal coverage. Short-packet transmission (SPT) is expected to be performed by ISAC BSs in 6G vertical applications to provide enhanced ultra-reliable low-latency S&C services. This paper proposes a STAR-RIS-assisted heterogeneous (SRH) MBC ISAC mechanism within the finite blocklength (FBL) regime. Multiple ISAC BSs and a communication-only (Comm-only) BS collaborate to provide S&C services to users and the target using STAR-RISs and cooperative detection results. A composite SPT (C-SPT) time frame structure is proposed by dynamically designing the S&C block lengths from the perspectives of the sub-frame and the entire transmission block. The systemic sum achievable capacity (SAC) over the entire transmission block is maximized by jointly optimizing the pilot and data block lengths and powers, as well as the active and passive beamforming of ISAC BSs and STAR-RISs under constraints of temporal, spatial, and power resources. The non-convex optimization problem is solved by combining the methods of alternating optimization (AO), quadratic transform (QT), sequential rotation (SR), and particle swarm optimization (PSO). Simulation results demonstrate the superiority of the proposed SRH-MBC ISAC mechanism with the C-SPT frame structure in achieving higher systemic communication performance with the assistance of STAR-RIS.

This study investigates a downlink rate-splitting multiple access (RSMA) scenario in which multiple base stations (BSs), employing a coordinated multi-point (CoMP) transmission scheme, serve users equipped with movable antenna (MA) technology. Unlike traditional fixed-position antennas (FPAs), which are subject to random variations in wireless channels, MAs can be strategically repositioned to locations with more favorable channel conditions, thereby achieving enhanced spatial diversity gains. To leverage these advantages and maximize the achievable sum rate, we formulate an optimization problem that jointly determines the optimal transmit beamforming vectors at the BSs, the common stream allocation for different users, and the optimal positioning of the MAs, all while ensuring compliance with quality of service (QoS) constraints. However, the formulated problem is non-convex and computationally challenging due to the strong interdependence among the optimization variables. Traditional methods for solving large-scale optimization problems typically incur prohibitively high computational complexity. To address the above challenge, we propose a gradient-based meta-learning (GML) algorithm that operates without pre-training and is well-suited for handling large-scale optimization tasks. Numerical results demonstrate the effectiveness and accuracy of the proposed approach, achieving near-optimal performance (exceeding 97% compared to the optimal solution). Moreover, the MA-enabled CoMP-RSMA model significantly outperforms conventional benchmark schemes, yielding performance gains of up to 190% over the spatial division multiple access (SDMA)-based scheme and 80% over the RSMA FPA-based model. Finally, the proposed approach is shown to mitigate the sum-rate upper-bound limitations imposed by interference in SDMA, achieving superior performance with fewer BSs.