Source-linked AI summary
Maneuver Coordination Using V2I to Improve Traffic Efficiency at Intersections
Lorenzo Farina, Marco Rapelli, Barbara Mavì Masini, Claudio Casetti, Alessandro Bazzi
TL;DR
Urban congestion and air pollution motivate better intersection management, while existing maneuver-coordination standards define messages and protocols but not traffic-management algorithms. The paper proposes a FIFS algorithm and communication protocol for CAVs coordinating with an intersection controller, and reports travel-time and emissions advantages over priority, traffic-light, and FIFO approaches.
Problem
The paper addresses intersection traffic efficiency and emissions, noting that coordination standards define messages and protocols but not traffic-management algorithms.
Source: §I · p. 1Method
The paper proposes a cellular-based FIFS algorithm and a communication protocol for controller-coordinated CAVs to cross an unregulated intersection without stopping unless congestion is heavy.
Source: §I · pp. 1–2Results
The proposed solution improves travel time and emissions compared with priority, traffic-light, and FIFO-managed intersections.
Source: §VI · p. 7Takeaways & Limitations
The paper presents controller-coordinated CAV crossing without routine stops as a practical approach for unregulated intersections, except under heavy congestion.
Source: §I · p. 1, §VI · p. 7
Abstract
from arXiv · showhide
Traffic efficiency and air pollutants are major issues in urban transport systems, as they are critical points for traffic flows and people living in urban areas, respectively. In response to this challenge, connected and autonomous vehicles (CAVs) have garnered increasing attention, with the potential to address the above challenges through cooperative maneuvering and real-time communication. In this paper, we consider an intersection served by a multi-access edge computing (MEC) where CAVs communicate with a controller to pass the intersection without the need of signals and without stopping unless in presence of heavy congestion. For this purpose, we propose a first-in-first-scheduled (FIFS) algorithm, and we specify the communication protocol for its practical implementation. The solution is validated through simulations with varying traffic densities, showing that it provides significant gains in travel time and emissions compared to both uncontrolled intersections, with or without traffic lights, and a benchmark solution derived from the literature.
I. INTRODUCTION
The paper proposes a cellular-based FIFS algorithm and communication protocol for coordinating CAVs at unregulated intersections, targeting shorter travel times and lower emissions. It addresses the gap between coordination standards and traffic-management algorithms, and evaluates the proposal through simulations.
I. INTRODUCTION
- Urban congestion and air pollution motivate intersection management, with drivers estimated to spend 54 extra hours yearly in congestion and WHO reporting 7 million annual air-pollution deaths,.Source: §I · p. 1
- Standards define messages and protocols for maneuver coordination, but do not provide traffic-management algorithms; related proposals often simplify either communication or control.The paper frames integrating both aspects as a gap.Source: §I · p. 1
- The proposed cellular-based FIFS algorithm schedules CAVs through an unregulated intersection without stopping unless congestion is heavy.Source: §I · p. 1
- The proposal considers turning vehicles, uses a single trajectory-definition phase, accommodates vehicle-specific motion choices, and details the vehicle-controller communication protocol.Source: §I · pp. 1–2
- Simulations with variable vehicle density show reduced travel time and lower emissions for the proposal.The supplied introduction passage does not specify comparator values.Source: §I · p. 2
II. RELATED WORKS
Prior intersection-coordination studies explore reservation, trajectory optimization, and resequencing, while this paper pairs a FIFS algorithm with a communication protocol and a single trajectory-definition phase. Its model details a single-lane, four-way intersection with control and conflict zones, V2I communication, and vehicle-specific trajectories.
II. RELATED WORKS
- Prior work explores reservation-based FIFO scheduling, turning trajectories, and resequencing, while repeated trajectory changes are identified as a potential concern for practical communication protocols.This paper presents a FIFS algorithm and related protocol using a single trajectory-definition phase.Source: §II · p. 2
III. MODEL DESIGN
- The paper details its algorithm for a classic single-lane, four-way intersection and leaves extensions to other intersection types for future investigation.Source: §III · p. 2
- The model divides the intersection into a control zone, where the algorithm starts, and conflict subzones, which only one vehicle may occupy at a time.A vehicle occupies two subzones while transitioning between them.Source: §III · p. 2
- The scenario assumes CAVs use V2I communication with an infrastructure-connected MEC server hosting the intersection controller.The controller schedules conflict-subzone reservations using maneuver-coordination messages.Source: §III · p. 2
- Each vehicle independently defines its trajectory according to the controller's schedule, accommodating differences in vehicle characteristics.Source: §III · pp. 2–3
IV. FIRST-IN FIRST-SCHEDULED ALGORITHM
The algorithm starts when an ego-vehicle enters the control zone and is organized into three parts plus a backup mode.
IV. FIRST-IN FIRST-SCHEDULED ALGORITHM
- The ego-vehicle initiates the algorithm upon entering the control zone.Source: §IV · p. 3
- The procedure has three parts and a backup mode discussed at the end of the section.Source: §IV · p. 3
- Algorithm 1 describes ego-vehicle commands, while Algorithm 2 describes controller instructions.Source: §IV · p. 3
A. Part 1: Preliminary evaluation at the EGO
The ego-vehicle estimates its earliest conflict-zone arrival, proposes crossing times to the controller, and uses the response to set its motion profile or enter backup mode.
A. Part 1: Preliminary evaluation at the EGO
- On control-zone entry, the ego-vehicle estimates a no-traffic trajectory and holds speed during ΔTinit to allow the communication exchange to finish.The estimated conflict-zone arrival time is denoted t#; ΔTinit should reflect estimated communication delay.Source: A. Part 1: Preliminary evaluation at… · p. 3
- Arrival-time planning minimizes time to the intersection while constraining acceleration and deceleration, with example position, speed, and acceleration profiles shown.For a lower initial speed, the example accelerates toward the target speed after the initial interval.Source: A. Part 1: Preliminary evaluation at… · p. 3, Figure 2 · p. 4
- The controller postpones proposals until conflicts are cleared, sends a prescription, and waits for ego-vehicle confirmation or a new proposal.The ego-vehicle confirms and sets the trajectory when it can adhere to the prescription.Source: A. Part 1: Preliminary evaluation at… · p. 3
- The ego-vehicle proposes entry and subzone times, retries with a recalculated trajectory after a response timeout, and enters backup mode if it gets too close to the conflict zone.Source: A. Part 1: Preliminary evaluation at… · p. 4
B. Part 2: Evaluation and prescription from the controller
The controller checks proposed crossing profiles for conflicts before, within, and after the intersection, postponing arrival until the crossing path is conflict-free.
B. Part 2: Evaluation and prescription from the controller
- Before the conflict zone, the controller delays the ego-vehicle if its estimated profile comes within the minimum safety distance of the vehicle ahead.The implementation compares positions every 0.5 s and adds 0.2 s when distance falls below 2 m.Source: 1) Part 2.A: The intersection… · p. 4
- Inside the conflict zone, the controller stores each scheduled vehicle’s subzone times and searches successive idle intervals for a collision-free crossing.If an idle interval cannot fit the complete crossing, it searches the next interval from the first subzone.Source: 2) Part 2.B: The intersection… · p. 4
- After the intersection, the controller checks whether the ego-vehicle’s braking distance would bring it too close to the vehicle ahead and increases its arrival time if needed.The check compares dbrake with dafter − dsafe-after.Source: 3) Part 2.C: Intersection controller… · pp. 4–5
- Fig. 3 depicts three scheduled vehicles and a controller scheduling V4; Fig. 4 contrasts a conflict with V2 against a feasible slot between V2 and V3.Source: 2) Part 2.B: The intersection… · p. 4, Figure 3 · p. 5
- The controller repeats its conflict checks and time calculations until no conflicts remain, then sends the adjusted subzone entry and exit times.The new intersection time cannot be earlier than t#.Source: B. Part 2: Evaluation and prescription… · p. 4, 3) Part 2.C: Intersection controller… · p. 5
C. Part 3: Mobility profiles settings and crossing scheduling
After receiving the controller’s prescription, the ego-vehicle checks whether it can meet the crossing times, confirms its plan, or returns with a new proposal or enters backup mode.
C. Part 3: Mobility profiles settings and crossing scheduling
- The ego-vehicle calculates a speed profile for the prescribed arrival time and confirms the plan with subzone entry and exit times.Source: C. Part 3: Mobility profiles settings… · p. 5
- If it cannot fulfill the prescription and will reach the intersection later, the ego-vehicle returns to the preliminary evaluation and proposes a new crossing.Source: C. Part 3: Mobility profiles settings… · p. 5
- A speed profile below threshold or failure to agree before reaching the proximity threshold triggers backup mode; otherwise, motion profiles extend through the conflict zone exit.Source: C. Part 3: Mobility profiles settings… · p. 5
- Fig. 4 contrasts V4’s conflict with V2 at its minimum arrival time with a schedule between V2 and V3.Source: Figure 4 · p. 5
- The controller’s procedure ends on ego-vehicle confirmation; otherwise, it waits for a new proposal.Source: C. Part 3: Mobility profiles settings… · p. 5
D. Backup mode
Backup mode is a last resort for heavy congestion or when safe scheduling thresholds are reached; vehicles then use their sensors and normal road rules. The controller waits for all backup-mode vehicles to cross before restarting the algorithm.
D. Backup mode
- An EGO enters backup mode if no agreement is reached by L/2 from the control zone or its calculated speed profile falls below 3 m/s.These thresholds are used in the implementation.Source: D. Backup mode · p. 5
- In backup mode, the CAV resumes normal operation, relying on its sensors and ordinary road rules rather than controller indications.Source: D. Backup mode · p. 5
- The controller waits for every backup-mode vehicle to cross the intersection before re-initializing the algorithm.A vehicle in backup mode notifies the controller after crossing the control zone.Source: D. Backup mode · p. 5
V. SIMULATION RESULTS
The results are based on SUMO simulations controlled through a Python script using TraCI. The main simulation parameters and settings are listed in Table I.
V. SIMULATION RESULTS
- Simulations use SUMO and a Python script communicating through TraCI to monitor vehicle positions and control behavior in the control zone and intersection.Vehicles are controlled from control-zone entry until they leave the intersection, unless they enter backup mode.Source: §V · p. 5
- Table I summarizes the main simulation parameters and settings.Source: Table I · p. 6
A. Simulation settings and outputs
The simulation models a four-road intersection with probabilistically generated vehicle arrivals and route choices, and applies bounded random communication delays. FIFS is assessed against four benchmark configurations.
1) Intersection definition:
- Right turns occupy one conflict subzone, straight movements two, and left turns three in the modeled four-road intersection.Each road is 2L long, with L=100 m, and arrivals follow a Poisson process with equal rates across directions; turn choices are uniform.Source: 1) Intersection definition · p. 6, Figure 5 · p. 6
2) Communication model:
- EGO-to-controller and controller-to-EGO transmission delays are modeled as uniform random values from 20 to 100 ms.The model assumes losses are detected and messages can be retransmitted until received.Source: 2) Communication model · p. 6
3) Benchmarks:
- Four benchmarks are horizontal-road priority, a traffic light, FIFO with ideal communication, and FIFS with ideal communication.The traffic light uses 35-second green phases and 3-second yellow phases; ideal communication means perfect vehicle positions and instantaneous decisions.Source: 3) Benchmarks · p. 6
4) Output Metrics:
The evaluation tracks travel time and vehicle CO2 emissions during travel.
4) Output Metrics:
- Travel time measures the seconds from a vehicle's entry into the simulation until it exits.Source: 4) Output Metrics · p. 6
- Emissions measure vehicle CO2 produced during travel time using SUMO's default settings.Source: 4) Output Metrics · p. 6
B. Discussion
Across traffic densities, FIFS with the designed communication protocol outperforms the non-CAV priority and traffic-light benchmarks and FIFO, though high congestion can trigger backup mode and non-optimal performance. The paper concludes that its V2I-based FIFS approach improves travel time and emissions at unregulated intersections.
B. Discussion
- At 0.05 vehicles/s per direction, controlled CAVs using FIFO or FIFS avoid intersection stops and perform best; FIFS and FIFO results are very similar.Traffic lights perform worst in this relaxed-traffic case because they cause unnecessary stops and waiting.Source: B. Discussion · p. 6
- At 0.15 vehicles/s per direction, FIFS travel time never exceeds 50 s and CO2 per vehicle never exceeds 200 g, including with realistic V2I delays.The FIFO approach has high travel times because vehicles enter subzones only after preceding vehicles are scheduled.Source: B. Discussion · p. 6
- With realistic V2I delay, FIFS significantly outperforms priority, traffic-light, and FIFO benchmarks; at 0.20 vehicles/s per direction, backup mode can make performance non-optimal.FIFS with ideal communication shows optimum performance at every density, including average results and worst cases.Source: B. Discussion · p. 7
VI. CONCLUSION
- The proposed FIFS algorithm and communication protocol use V2I between vehicles and a MEC-served controller to cross unregulated intersections without stopping.The conclusion reports advantages in travel time and emissions over priority, traffic-light, and FIFO management.Source: §VI · p. 7
- The evaluated setting excludes non-connected vehicles and vulnerable road users, and uses centralized V2I rather than distributed V2V communication.Field testing and these broader settings are left for future work.Source: §VI · p. 7