https://scholar.dgist.ac.kr/handle/20.500.11750/13854 2026-04-05T16:42:34Z https://scholar.dgist.ac.kr/handle/20.500.11750/57191 Title: Edge-Server Workload Characterization in Vehicular Computation Offloading: Semantics and Empirical Analysis Author(s): Kim, BaekGyu; Gangadharan, Deepak Abstract: Edge server-assisted computation offloading enables vehicles to leverage server compute resources to deliver connected services, overcoming the limitations of onboard resources. Understanding the compute workloads of edge servers is crucial for effective resource management and scheduling, yet this task is challenging due to the complex interplay of factors such as vehicle mobility and computation offloading patterns. To address this, we propose an empirical analysis framework that systematically characterizes the compute workloads of edge servers. We begin by formalizing the relationships among three key aspects: local load (generated by vehicles), composite load (imposed on edge servers), and traffic flow (vehicle mobility patterns). Our framework then uses models of the local load and traffic flow as inputs to generate the composite loads on edge servers. Experiments were conducted by injecting between 600 and 5,000 vehicles per hour in two distinct geographical areas, New York City and Tampa. We provide a quantitative analysis demonstrating how the composite loads on edge servers vary with changes in traffic flows, geographical areas, and offloading patterns. Authors 2024-05-31T15:00:00Z https://scholar.dgist.ac.kr/handle/20.500.11750/57186 Title: Trustworthy Compiler Design for Generating Concrete Scenarios Toward Certifying Autonomous Driving Safety Author(s): Kim, BaekGyu Abstract: Generating synthetic driving scenarios in virtual environments is being actively studied for autonomous driving systems, whose safety is challenging to verify in physical environments. While several works proposed a way to describe such driving scenarios called Operational Design Domain (ODD), their syntax still lacks formal semantics, which makes it difficult to design trustworthy compilers -a software that converts an abstract scenario expressed in ODDs into a concrete scenario executable in a virtual environment. We propose a systematic way to formalize the lane-level driving scenarios described using a subset of ASAM OpenSCENARIO syntax. In particular, our formalization exploits the concept of separating concerns in which an abstract scenario is described without exposing the low-level behavioral details of autonomous driving; the details can then be auto-filled during the concrete scenario generation via different objectives. We define the decision variables that need to be determined when generating concrete scenarios; then, we propose a set of constraint categories that restrict the valuation space of the variables according to the syntactic structure of an abstract scenario. We implemented two compiler designs as exemplar based on two solvers: one with SMT (Satisfiability Modulo Theories) and another MILP (Mixed-Integer Linear Programming) solvers, and show the qualitative and quantitative analysis of concrete scenarios auto-generated from an abstract one. 2025-01-31T15:00:00Z https://scholar.dgist.ac.kr/handle/20.500.11750/46277 Title: Toward Large-Scale Test for Certifying Autonomous Driving Software in Collaborative Virtual Environment Author(s): Kim, BaekGyu; Kang, Eunsuk Abstract: Virtual simulation environments are widely used to test autonomous driving software by creating highly complex driving scenarios that are non-trivial to set up in a physical environment. However, the current practice of using the virtual test still does not fully utilize its potential to build a much larger scale test. We propose a perspective and research vision to build a large-scale test architecture in which participants collaboratively construct, execute and analyze complex test scenarios at scale in the virtual world. In particular, the architectural concept is built on the existing concept of the Collaborative Virtual Environment (CVE) that has been successfully applied in other domains, such as entertainment or military training applications. The proposed domain-specific architectural requirements extend the CVE to include the following necessary properties - selective sharing and collaboration - to test autonomous driving software. In addition, the test architectural concept is explained as to how a large number of participants interact with each other collaboratively to build and execute diverse test scenarios at scale. Finally, we explain the new research directions to make this test architectural concept realized for testing autonomous driving software. 2023-06-30T15:00:00Z https://scholar.dgist.ac.kr/handle/20.500.11750/17494 Title: Time Efficient Offloading Optimization in Automotive Multi-access Edge Computing Networks Using Mean-Field Games Author(s): Kang, Yuhan; Wang, Haoxin; Kim, BaekGyu; Xie, Jiang; Zhang, Xiao-Ping; Han, Zhu Abstract: Emerging connected vehicular services, such as intelligent driving and high-definition (HD) map, are gaining increasing interest with the fast development of multi-access edge computing (MEC). For most time-sensitive and computation-intensive vehicular services, the data offloading process significantly influences the capacity and performance of MEC, especially when the number of connected vehicles is enormous. In this work, we consider data offloading optimization for a large-scale automotive MEC network. The problem is challenging due to the large number of connected vehicles and the complicated interaction between vehicles and edge servers. To tackle the scalability problem, we reformulate the original offloading optimization problem into a Mean-Field-Game (MFG) problem by abstracting the interaction among the connected vehicles as a distribution over their state spaces of task sizes, known as the mean-field term. To solve the problem efficiently, we propose a G-prox Primal-Dual-Hybrid-Gradient (PDHG) algorithm that transforms the MFG problem into a saddle-point problem. Based on our developed MFG model and G-prox PDHG algorithm, we propose the first data offloading scheme whose computation time is independent of the number of connected vehicles in automotive MEC systems. Extensive evaluation results corroborate the superior performance of our proposed scheme compared with the state-of-the-art methods. © 2023 IEEE 2023-04-30T15:00:00Z