검색 상세

A Study of Mobile Edge Computing System Architecture for Vehicular Cloud Media Services on Highway

  • SangYubLEE (이상엽, 대학원 컴퓨터정보학과 내장형시스템 및 실시간 컴퓨팅전공)

초록/요약

In this thesis, we investigate a method to reduce data traffic load on Data Center Cloud (DCC) for media contents in a congested road or highway environment. We present a new mobile edge network architecture to handle increasing amounts of traffic, quality requirements, advanced driver assistance system for autonomous driving and new cloud computing demands on a highway. We propose a hierarchical cloud computing architecture to enhance reducing data load and download latency performance by using adaptive data load distribution with buses, which play the role of an edge computing server. A vehicular dynamic cloud is based on a wireless architecture including Wireless Local Area Network (WLAN) and infrastructure Long Term Evolution and Long Term Evolution Advance (LTE/LTE-A) communication is used for data transmission between moving buses and cars. The main advantages of the proposed architecture include a reduction in data loading for the top layer cloud server and an effective data distribution on a traffic jam highway where moving vehicles require video on demand (VOD) services from the server. Through a description of a real environment based on an NS-2 network simulation, we conduct experiments to validate the proposed architecture with the flow of motorways. Additionally, we investigate how much cost gain and network traffic reduction is shown by the internally applied in bus according to the network type as the Mobile Edge Computing Server (MECS). We validate the performance of the proposed architecture using an actual in-vehicle network type consisting of a dual network ring. Through the proposed process, we show the feasibility and effectiveness of the connected car media service on a highway from various aspects such as data load density, cost charge, service latency and change in packet quantity.

more

목차

I. INTRODUCTION １７
II. RELATED WORK ２１
II.1. Motivation ２１
II.2. Automotive Network Technologies ２２
II.2.1. Vehicular Networking ２３
II.3. In-Vehicle Network in Bus ２５
II.3.1. Media Oriented System Transport (MOST) ２６
II.3.2. MOST Ethernet Packet (MEP) ２８
II.3.3. Implementation of MOST Network in Bus ２９
III. MECS SYSTEM ARCHITECTURE ３２
III.1. General Vehicular Cloud Service Model ３３
III.1.1. Vehicular Cloud Computing Architecture ３３
III.1.1.1. Network as a Service (NaaS) ３４
III.1.1.2. Storage as a Service (STaaS) ３５
III.1.1.3. Cooperation as a Service (CaaS) ３６
III.1.1.4. Entertainment as a Service (ENaaS) ３７
III.1.2. Application of General Vehicular Cloud ３８
III.1.2.1. Parking Lot as a Static Data Center ３８
III.1.2.2. Traffic Light Management as a Dynamic Data Center ３９
III.1.2.3. Self-organized high occupancy vehicle (HOV) as a Dynamic Data Center ４０
III.2. Connected Car Media Service ４１
III.3. System Mechanism ４３
III.3.1. System Architecture ４４
III.3.1.1. Data scheduling method (DSM) ４５
III.3.1.2. Car selection algorithm (CSA) ４７
III.3.1.3. Bus information management (BIM) ４９
III.4. Extended MECS Model ５１
III.4.1. Dual MOST Network System ５１
III.4.1.1. Implementation of Dual MOST Network System ５４
III.4.2. Extended MECS Service ５７
IV. PERFORMANCE EVALUATION ５８
IV.1. Evaluation of the Proposed Architecture ６０
IV.1.1. Data Load Analysis ６０
IV.1.2. Data Load Optimization for Extended MECS Model ６２
IV.1.3. Download and Service Latency Analysis ６４
IV.2. Performance Evaluation of MECS based on Single MOST Network System ６９
IV.2.1.1. Performance Analysis ７１
IV.3. Performance Evaluation of Extended MECS based on Dual MOST Network System ７５
IV.4. Performance Comparison between MECS and Extended MECS Service Model ７８
V. CONCLUSIONS ８１
REFERENCES ８２

more
반출 Meta View 목록