On the Design of Synchronous Traffic Protocols for Intelligent Intersections

M.Sc. Daniel Markert

AutorIn

M.Sc. Daniel Markert

Titel

On the Design of Synchronous Traffic Protocols for Intelligent Intersections

Zitierfähige Url:

https://nbn-resolving.org/urn:nbn:de:bsz:ch1-qucosa2-1015238

Erstveröffentlichung

2026

Datum der Einreichung

20.06.2025

Datum der Verteidigung

12.01.2026

DOI

https://doi.org/10.60687/2026-0021

Abstract (EN)

Intelligent intersections aim to replace conventional traffic lights with traffic protocols that dynamically schedule connected and automated vehicles (CAVs) in all directions, as opposed to greatly static traffic lights that are currently in use. Traffic protocols can be synchronous or asynchronous in nature, with synchronous such traffic protocols enforcing efficient crossing patterns to achieve a higher throughput. However, due to the open-ended nature of realistic intersections, current synchronous traffic protocols are generally confined to specific types of traffic or infrastructure, which greatly restricts their applicability in the real world. To address this issue, this thesis proposes several design concepts to increase the real-world applicability of synchronous traffic protocols to a wider range of settings. More specifically, three points are covered. First, deterministic methods fail to provide meaningful estimates of the maximum number of vehicles at the intersection, which is paramount to assess communication reliability and, in the end, guarantee safety. In contrast, probabilistic estimates can greatly reduce pessimism and overdesign compared to deterministic approaches while still retaining safety. These and other benefits or the proposed approach are illustrated by means of a detailed case study and simulations using OMNeT++. Second, to guarantee safety, traffic protocols must enforce sufficiently large gaps between vehicles on different lanes, taking their dimensions into account. In particular, existing such protocols are designed for the longest possible vehicle resulting in space-hungry intersections, that require modifications in the infrastructure (in particular, broader roads/lanes). Moreover, these do not allow for vehicles that are exceptionally longer than the ones considered at design time (e.g., extra long trucks, or buses, etc.). In this thesis, to overcome this limitation, all proposed approaches handle overlength vehicles as exceptions to compensate for their low probability of occurrence, relaxing space requirements on the intersections. This is implemented in a single-crossing traffic protocol called SV-LTR (Single-Vehicle LTR) and further extended by a two-speed scheme to account for realistic driving and turning behavior and augmented with a platooning mode called PB-LTR (Platooning-Based LTR), where vehicles on opposing lanes cross with reduced inter-vehicle distances while considering the delay to perpendicular lanes. Finally, FleXS-TP (Flexible Synchronous Traffic Protocol) dynamically forms crossing patterns without requiring specific vehicle arrival orders. By maintaining a synchronous core while scheduling vehicles individually, FleXS-TP combines the efficiency of synchronous protocols under well-behaved traffic with the flexibility of asynchronous approaches under randomized traffic, achieving substantial throughput in both scenarios, as demonstrated by realistic SUMo simulations. These results illustrate the proposed techniques’ effectiveness in enhancing safety, throughput, and space efficiency, while broadening the applicability of synchronous traffic protocols todiverse and realistic traffic conditions.

Freie Schlagwörter (DE)

Intelligente Verkehrssysteme, Autonome Fahrzeuge

Freie Schlagwörter (EN)

Intelligent Intersections, Connected Autonomous Vehicles (CAVs), Traffic Protocols

Klassifikation (DDC)

004.62

005.1

Normschlagwörter (GND)

Verkehrssystem, Autonomes Fahrzeug

GutachterIn

Prof. Dr. Alejandro Masrur
Jun.-Prof. Dr. Stefan Reitmann

BetreuerIn Hochschule / Universität

Prof. Dr. Alejandro Masrur

Den akademischen Grad verleihende / prüfende Institution

Technische Universität Chemnitz, Chemnitz

Version / Begutachtungsstatus

angenommene Version / Postprint / Autorenversion

URN Qucosa

urn:nbn:de:bsz:ch1-qucosa2-1015238

Veröffentlichungsdatum Qucosa

21.01.2026

Dokumenttyp

Dissertation

Sprache des Dokumentes

Englisch

Lizenz / Rechtehinweis

CC BY 4.0

Inhaltsverzeichnis

1. Introduction
  1.1. Scope and Motivation
  1.2. Contributions
  1.3. Thesis Structure

2. Related Work
  2.1. Asynchronous Arrival Pattern
  2.2. Synchronous Arrival Pattern
  2.3. Adjacent Approaches
    2.3.1. Platooning
    2.3.2. Machine Learning
    2.3.3. Game Theory
  2.4. Summary

3. Fundamentals of Intelligent Intersections and Traffic Protocols
  3.1. Models and Assumptions — Synchronization
    3.1.1. Vehicle Lengths and Sectors
    3.1.2. Overlength Penalty
    3.1.3. Cycles and the Two-Speed Scheme
    3.1.4. Synchronization Strategies

4. Probabilistic Reliability Modeling
  4.1. Vehicle Length Distribution
    4.1.1. Deterministic Approach
    4.1.2. Probabilistic Approach
  4.2. Traffic Density and Direction
    4.2.1. Probabilities
    4.2.2. Deriving Vehicle Count
  4.3. Communication Protocol / Scheme
    4.3.1. Medium Access Control
    4.3.2. Physical Layer
  4.4. Fallback Mechanisms

5. Space Efficiency
  5.1. Ballroom Intersection Protocol (BRIP)
  5.2. Single-Vehicle Left, Through, Right (SV-LTR) Protocol
    5.2.1. Drive Through Only / Drive Through and Right Turns
    5.2.2. Drive Through and Left Turns
    5.2.3. Right and Left Turns
    5.2.4. Considering Overlength Vehicles
    5.2.5. Algorithm Performance

6. Platooning and Fairness
  6.1. Platooning-Based Left, Through, Right (PB-LTR) Protocol
    6.1.1. Drive-Through Platoon Crossing
    6.1.2. Left-Turn Platoon Crossing
    6.1.3. Considering Right Turns
    6.1.4. Maximum Blocking Time

7. Flexible Synchronous Traffic Protocol (FleXS-TP)
  7.1. Blocking Chart and Blocking Patterns
  7.2. Square Sectors
    7.2.1. Drive Through
    7.2.2. Left Turns
    7.2.3. Right Turns
    7.2.4. Overlength Vehicles
  7.3. Variable Sector Length

8. Simulations and Algorithms
  8.1. Simulation of Urban Mobility (SUMo)
    8.1.1. Generation of Well-Behaved and Randomized Traffic
    8.1.2. Simulation of Unsorted Traffic without Contention
    8.1.3. Conventional Traffic Lights
    8.1.4. Simulation of BRIP
    8.1.5. Simulation of SV-LTR
    8.1.6. Simulation of FleXS-TP
  8.2. Performance of Vehicle Count Estimation Algorithm

9. Evaluation
  9.1. Probabilistic Vehicle Count Estimation and Communication Reliability
    9.1.1. Probabilistic Vehicle Count Estimation
    9.1.2. Impact on Communication Reliability
    9.1.3. Fallback Mechanism — Examples
  9.2. Space Efficiency
  9.3. Throughput Part I: LTR vs. BRIP
    9.3.1. SV-LTR vs. BRIP
    9.3.2. PB-LTR vs. BRIP
    9.3.3. Summary: LTR Versions vs. BRIP
  9.4. Throughput Part II: SV-LTR vs. FleXS-TP
    9.4.1. Well-Behaved Traffic
    9.4.2. Randomly Generated Traffic

10. Discussion and Conclusion
  10.1. Summary of Findings
    10.1.1. Chapter 4 — Reducing Deterministic Pessimism
    10.1.2. Chapter 5 — Infrastructure-Agnostic Design
    10.1.3. Chapter 6 — Platooning for Well-Behaved Traffic
    10.1.4. Chapter 7 — Increasing Flexibility towards Traffic Composition
  10.2. Reflecting Design Choices / Limitations
    10.2.1. Vehicle Reordering
    10.2.2. Coping with System Accidents
    10.2.3. System-Level Safety
    10.2.4. Centralized vs. Distributed Protocols
  10.3. Outlook

Bibliography

Appendix A. Code Samples and Extra Material
  A.1. Example Route File
  A.2. Randomized Traffic Route File Generation
  A.3. BRIP Type I Route File Generation
  A.4. Variable Vehicle Length Alone

Nutzungshinweise für die digitalen Objekte in Qucosa