Date : 12/05/2026
Hour: 13.00
Student Name & ID: Kadir Yılmaz & 255105113
Supervisor: Dr. Öğr. Üyesi Metin Öztürk
Topic: Knowledge Base Update Optimization for Semantic Communication in Vehicular Networks
Online Meeting Link : A-424
Abstract: Semantic communication has emerged as a promising paradigm for enabling efficient and meaningful data exchange in next-generation vehicular communication systems and connected car environments. In such systems, the knowledge base plays a critical role in supporting semantic encoding, transmission, and interpretation. However, maintaining an up-to-date and relevant knowledge base under dynamic vehicular conditions remains a key challenge, as frequent updates may improve semantic accuracy but also introduce additional communication overhead, latency, and resource consumption. In this work, the knowledge base management process in semantic communication is formulated as a multi-objective optimization problem, where semantic accuracy, latency, and bandwidth usage are jointly considered. A control mechanism is introduced to regulate the frequency and extent of knowledge base updates, enabling an adaptive balance between communication efficiency and semantic performance. The proposed framework analyzes different update strategies and evaluates their impact on overall system performance in vehicular communication scenarios. This study provides insights into the design of resource-efficient and scalable semantic communication systems for connected vehicles operating in dynamic environments.