Ignacio Sanudo Olmedo

arXiv2026

Measurement-Based Performance Evaluation of SmartRSUs with Heterogeneous Antenna Architectures for V2X Communications

This paper presents a measurement-based performance evaluation of two custom Smart Roadside Units (SmartRSUs) featuring different V2X antenna architectures. The first configuration integrates GNSS and communication antennas into an all-in-one rooftop module, whereas the second uses external dual ITS-G5 (IEEE 802.11p) antennas operating at 5.9~GHz and a dedicated GNSS antenna. Both systems are built upon a proprietary On-Board Unit (OBU) platform adapted for infrastructure deployment. The experimental campaign evaluates key V2X communication metrics, including coverage, received signal strength indicator (RSSI), packet loss, and end-to-end latency in both transmission (OBU-to-infrastructure) and reception (infrastructure-to-OBU) directions. To ensure objective validation, a commercial off-the-shelf V2X Roadside Unit is co-located on the same infrastructure and used as a performance benchmark, providing ground-truth reference measurements under identical environmental conditions through a controlled co-located deployment. Results highlight the impact of antenna design and placement on communication reliability and latency, revealing trade-offs between integrated and external antenna configurations in real-world deployment scenarios. The findings provide practical insights for the design and optimization of next-generation SmartRSUs in cooperative intelligent transportation systems (C-ITS).

cs.NI↗

arXiv2021

Contention-Aware GPU Partitioning and Task-to-Partition Allocation for Real-Time Workloads

In order to satisfy timing constraints, modern real-time applications require massively parallel accelerators such as General Purpose Graphic Processing Units (GPGPUs). Generation after generation, the number of computing clusters made available in novel GPU architectures is steadily increasing, hence, investigating suitable scheduling approaches is now mandatory. Such scheduling approaches are related to mapping different and concurrent compute kernels within the GPU computing clusters, hence grouping GPU computing clusters into schedulable partitions. In this paper we propose novel techniques to define GPU partitions; this allows us to define suitable task-to-partition allocation mechanisms in which tasks are GPU compute kernels featuring different timing requirements. Such mechanisms will take into account the interference that GPU kernels experience when running in overlapping time windows. Hence, an effective and simple way to quantify the magnitude of such interference is also presented. We demonstrate the efficiency of the proposed approaches against the classical techniques that considered the GPU as a single, non-partitionable resource.

cs.DC↗