상세정보

 Vehicle to vehicle (V2V) communication based on Institute of Electrical and Electronic Engineers (IEEE) Wireless Access in Vehicular Environment (WAVE) system paves the path for innovated applications in driving safety, drive-assistance, and future autonomous vehicle technology. Although the standards have been established, there still remain some ill-understood aspects of the technological components of the IEEE WAVE systems. In the IEEE WAVE, the periodic broadcast of the safety beacon enables proximity awareness in the V2V context. To maximize the level of awareness and consequently the driving safety, the beacon transmission at the highest allowed rate is desired in principle. One difficulty in the reliable delivery of the beacon is channel congestion that occurs when the allocated channel capacity is exceeded, and many researchers have proposed various congestion control algorithms to avoid this problem. Most of them, however, have glossed over the impact of unfairness problem that all vehicles in the network do not have similar opportunities to communicate with nearby vehicles. The other source of beacon delivery failure is the hidden terminal problem in V2V communication. Few researchers have addressed this problem, besides there is no consideration for directional communication aspect. In this dissertation, we investigate and solve the harmful influence of these congestion problems in V2V communication through mathematical analysis and simulation experiments. 
 To avoid the unfairness problem, we propose a distributed but coordinated congestion control algorithm that leads to stability and relevance to the given vehicle density pattern. By adding a jitter insertion scheme to information sharing that has been introduced in the newest standard of V2V communication, we solve the global synchronization problem caused by information sharing.
 We also show that proposed Application Layer Timing (ALT) can significantly mitigate hidden terminal problem. In addition, using a geographic scheduling algorithm to resolve the hidden terminal problem in directional communication, we show that it can achieve a significantly higher safety message delivery rate than omnidirectional messaging.
 The proposed solutions in this dissertation are simple to implement – they do not require modification of any related standards and can be immediately deployed by embedding a few lines of code in the safety application. Also, they are orthogonal to other existing congestion control algorithms. Namely, they are not designed to replace existing approaches. To the best of our knowledge, this dissertation is the first to tackle the hidden terminal problem for directionally broadcast safety messages through application layer scheduling.

Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . i
Contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iv
List of Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii
List of Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . viii
1 Introduction 1
2 Unfairness of Distributed Rate Control 7
2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.2 Related work . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
2.3 Evaluation experiments . . . . . . . . . . . . . . . . . . . . . . 12
2.3.1 Threshold-based control . . . . . . . . . . . . . . . . . . 15
2.3.2 Hysteresis-based control . . . . . . . . . . . . . . . . . . 21
2.3.3 Anomaly in real-life rate control schemes . . . . . . . . 31
2.4 Proposed solution: Mean-checked threshold-based control . . . 33
2.4.1 Dynamics . . . . . . . . . . . . . . . . . . . . . . . . . . 35
2.4.2 Proof . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
2.4.3 Performance . . . . . . . . . . . . . . . . . . . . . . . . 42
2.5 Other conditions . . . . . . . . . . . . . . . . . . . . . . . . . . 46
2.5.1 Channel fading . . . . . . . . . . . . . . . . . . . . . . . 47
2.5.2 Random start time . . . . . . . . . . . . . . . . . . . . . 49
2.5.3 Sliding window for channel measurements . . . . . . . . 50
2.5.4 Different adaptation constants . . . . . . . . . . . . . . 51
2.5.5 Uneven vehicle distribution . . . . . . . . . . . . . . . . 53
2.5.6 Random mobility pattern . . . . . . . . . . . . . . . . . 54
2.6 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
3 Global Syncronization in Information Sharing 58
3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
3.2 Related work . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
3.3 Global synchronization . . . . . . . . . . . . . . . . . . . . . . . 60
3.4 Proposed solution: Random jitter insertion . . . . . . . . . . . 63
3.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
4 Hidden Terminal Problem in Omnidirectional V2V Communication 69
4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
4.2 Related work . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
4.3 Finding the hidden terminals from application . . . . . . . . . . 73
4.4 Proposed solution: Application layer timing control . . . . . . . 79
4.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
5 Hidden Terminal Problem in Directional V2V Communication 85
5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
5.2 Related work . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
5.3 Problem definition . . . . . . . . . . . . . . . . . . . . . . . . . 89
5.3.1 Impact of the hidden terminal problem in vehicular communication
. . . . . . . . . . . . . . . . . . . . . . . . . 90
5.3.2 Worst-case analysis . . . . . . . . . . . . . . . . . . . . . 95
5.4 Proposed solution: Geographical scheduling . . . . . . . . . . . 106
5.4.1 Spatial structure for scheduling . . . . . . . . . . . . . . 106
5.4.2 Temporal structure for scheduling . . . . . . . . . . . . 109
5.4.3 Scheduling algorithm . . . . . . . . . . . . . . . . . . . . 111
5.4.4 Performance evaluation . . . . . . . . . . . . . . . . . . 117
5.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
6 Conclusion 145
References 149
Summary (in Korean) 161