C-ITS 첨단도로 환경을 위한 엣지 네트워크 기반 교통상황인지 시스템의 설계 및 구현 = A Design and Implementation of Traffic Situation Recognition System Based on Edge Network for C-ITS Advanced Road Environments

국문 초록 (Abstract)

차세대 교통시스템으로 주목받고 있는 C-ITS(Cooperative-Intelligent
Transport Systems)는 교통 분야의 최우선 해결과제인 교통안전과, 교
통편의를 목표로 연구개발 및 현장 적용에 박차를 가하고 있다. 이에 따
라 C-ITS의 요소 기술인 교통 정보 수집 기술, 도로 구성요소 간 연결
기술, 교통정보 서비스 등 다양한 연구개발이 진행되고 있다. 그러나 현
재 서비스되고 있는 교통시스템은 중앙집중식 교통정보 서비스 구조, 부
족하고 미흡한 첨단 도로 인프라 및 도로 인프라 간 연결 기술은 고도화
된 정보처리 서비스가 요구되는 C-ITS를 실현하기에 한계점이 있다.
본 논문에서는 이러한 한계점을 극복하기 위해 도로 인프라 간의 연결
및 정보처리 환경인 엣지 네트워크를 제안하였다. 엣지 네트워크는
RSSI(Received Signal Strength Indicator)와 메시지 수신 횟수를 기반으
로 인접한 노드의 그룹을 생성하는 하는 무선 네트워크 구축 기법으로,
생성된 그룹 내의 정보 교환을 통해 효율적인 엣지 컴퓨팅 환경을 지원
한다. 제안한 엣지 네트워크를 기반으로 교통흐름과 돌발 사고를 인지할
수 있는 교통상황인지 시스템을 구축하였다. 구축한 교통상황인지 시스
템은 교통정보 수집 및 제공 역할의 IoT 디바이스, IoT 디바이스 간 연
결 기법인 엣지 네트워크, 교통흐름 및 교통사고 인지하는 교통상황 인
지 알고리즘, 교통상황 모니터링 소프트웨어로 구성된다. 그리고 도로현
장실험을 통해 교통흐름과 교통사고를 인지하고, 도로 이용자에게 정보
를 전달하는 교통정보 서비스의 예를 제시하였다. 교통흐름 인지 기능은
상시 동작하여 IoT 디바이스를 통해 교통흐름 상태를 표시하고, 사고발
생 시 사고 상황을 인지하여 IoT 디바이스를 통해 현장에서 사고정보를
표시한다. 교통상황 모니터링 소프트웨어에서는 지도 기반 모니터링 기
능과 및 IoT 디바이스 상태관리 기능을 제공한다. 또한, 엣지 네트워크
의 안정성과 효율성을 검증하기 위해 기존의 RSSI 기반 메시 네트워크
와 제안한 엣지 네트워크를 대상으로 실험을 진행하였다. 실험결과 엣지
네트워크에서의 데이터 전송률은 평균 97.04%, 데이터 전송 시간은 평균
252.33ms, 네트워크 장애 복구시간은 평균 50798ms의 결과를 보였다.

다국어 초록 (Multilingual Abstract)

C-ITS(Cooperative-Intelligent Transport Systems), which is attracting
attention as a next-generation transportation system, is spurring R&D
and field application with the goal of traffic safety and transportation
convenience, which are the top priorities in the transportation field.
Various R&D such as information collection technology, road component
connection technology, and traffic information service are underway as
element technologies for C-ITS. However, the transportation system
currently being serviced has limitations in implementing C-ITS, which
requires advanced information processing services, due to the centralized
traffic information service structure, insufficient and insufficient
state-of-the-art road infrastructure and connection technology between
road infrastructures.
In this dissertation, to overcome these limitations, an edge network,
which is an environment for connection and information processing
between road infrastructures, is proposed. Edge networks are wireless
network construction techniques that create groups between adjacent nodes
based on RSSI and the number of message receptions, and support an
efficient edge computing environment by exchanging information between
nodes in the generated group. Based on the proposed edge network, a
traffic situation recognition system was established to recognize traffic
flows and unexpected accidents. The established traffic situation awareness
system consists of IoT devices that collect and provide traffic information,
edge network that is a connection technique between IoT devices, a traffic
situation recognition algorithm that recognizes traffic flows and accidents,
and traffic situation monitoring software. In addition, an example of a
traffic information service that recognizes traffic flow and traffic accidents
through road field experiments and delivers information to road users was
presented. The traffic flow recognition function operates all the time to
display the traffic flow status through the IoT device, recognizes the
accident situation when an accident occurs, and displays accident
information on the spot through the IoT device. The traffic condition
monitoring software provides map-based monitoring functions and IoT
device status management functions. In addition, to verify the stability and
efficiency of the edge network, experiments were conducted on the
existing RSSI-based mesh network and the proposed edge network. As a
result of the experiment, the average data transmission rate in the edge
network was 97.04%, the average data transmission time was 252.33ms,
and the average network failure recovery time was 50798ms.

목차 (Table of Contents)

• Ⅰ. 서론················································································································· 1
• 1. 연구의 배경 ······························································································· 1
• 2. 연구의 목적 ······························································································· 3
• 3. 논문의 구성 ······························································································· 4
• Ⅱ. 관련 연구······································································································ 5
• 1. 국내·외 교통정보시스템 사례 ································································ 5
• 가. 국내 교통정보시스템 ········································································ 5
• 나. 국외 교통정보시스템 ········································································ 8
• 2. 첨단교통정보와 교통 인프라································································· 9
• 가. 첨단교통정보의 정의 ········································································ 9
• 나. 정보수집 및 제공용 교통 인프라················································ 11
• 3. 교통정보 활용 관련연구 ······································································· 11
• 가. 돌발상황 검지 알고리즘의 분류 ·················································· 11
• 나. 교통량 정보 활용 및 상황인지 관련연구 ·································· 15
• 4. 엣지 컴퓨팅과 메시 네트워크 ··························································· 16
• 가. 엣지 컴퓨팅의 개념 ········································································ 16
• 나. 메시 네트워크의 개념 및 라우팅 알고리즘 ······························ 19
• 다. RSSI 기반 라우팅 알고리즘 관련연구 ······································· 27
• 5. 선행연구 고찰························································································· 28
• Ⅲ. 엣지 네트워크 기반 교통상황인지 시스템의 구현······················ 29
• 1. 시스템 설계 ····························································································· 29
• 가. 설계 고려사항 ·················································································· 29
• 나. 전체 시스템 구성 ············································································ 30
• 다. IoT 디바이스··················································································· 31
• 라. 엣지 네트워크 ·················································································· 33
• 마. 교통상황인지 알고리즘 ·································································· 38
• 바. 교통상황 모니터링 소프트웨어···················································· 48
• 2. 시스템 구현 ····························································································· 53
• 가. IoT 디바이스··················································································· 53
• 나. 엣지 네트워크 ·················································································· 57
• 다. 교통상황인지 알고리즘 ·································································· 60
• 라. 교통상황 모니터링 소프트웨어···················································· 61
• Ⅳ. 실험 및 결과 분석··················································································· 63
• 1. 교통상황인지 시스템의 도로 현장실험············································· 63
• 가. 실험환경 정의 ·················································································· 63
• 나. 교통흐름 인지 실험결과 ································································ 65
• 다. 교통사고 인지 실험결과 ································································ 68
• 2. 엣지 네트워크 성능 분석····································································· 72
• 가. 실험환경 정의 ·················································································· 72
• 나. 데이터 전송률 실험결과 ································································ 74
• 다. 데이터 전송시간 실험결과 ···························································· 78
• 라. 네트워크 장애복구 시간 실험결과·············································· 82
• 마. 네트워크 실험결과 종합 ································································ 86
• Ⅴ. 결론··············································································································· 88
• 참고문헌············································································································ 90
• ABSTRACT ···································································································· 96

참고문헌 (Reference)

1. 인공지능(AI) 기술 을 활용한 부산광역시 교통사고 예방 시스템, 조우진, 정석찬, 윤다현, 김진석, 이다희, 한국정보기술학회 2020년도 종합학술대회, pp.514-517, , 2020

2. V2I 통신환경을 활용 한 연동교차로 교통신호 실시간 제어 연구, 한국 ITS학회 논문지, 제17권, 제3호, 한음, 윤일수, 이상수, 장기태, 박병규, 통권77호, , 2018