V2N-Based Algorithm and Communication Protocol for Autonomous Non-Stop Intersections

This paper presents "Moveover," a novel V2N-based algorithm and communication protocol that enables autonomous vehicles to cross intersections without stopping by delegating trajectory optimization to individual vehicles while using a local controller for collision prevention, demonstrating significant improvements in travel time and emissions under various network conditions and intersection layouts.

The Gist

Imagine a busy intersection not as a chaotic four-way stop where cars take turns waiting for a red light, but as a high-speed dance floor. In the current system, everyone stops, waits for a signal, and then rushes forward, often causing a "stop-and-go" wave that wastes time and fuel.

This paper introduces a new system called Moveover. Think of it as a smart, invisible traffic conductor that lets cars glide through intersections without ever hitting the brakes, provided the road isn't completely jammed.

Here is how it works, broken down into simple concepts:

1. The Problem: The "Stop-and-Go" Traffic Jam

Currently, intersections are the most dangerous and inefficient parts of our roads. Cars stop at red lights, even when no one is coming from the other side. This wastes gas, creates pollution, and increases the risk of accidents. While self-driving cars (CAVs) are the future, many existing ideas for managing them are too complicated (requiring a super-computer to solve math problems for every car) or rely on perfect internet connections that don't exist in the real world.

2. The Solution: The "Moveover" Dance

The Moveover system changes the rules of the road. Instead of a central computer telling every car exactly what to do, it uses a collaborative negotiation between the car and a local "traffic manager."

• The Car (The Dancer): Each self-driving car knows its own body. A heavy truck moves differently than a tiny sports car. In Moveover, the car calculates its own perfect path and speed based on its own capabilities. It doesn't ask the computer, "How do I drive?" It says, "Here is my plan to get through safely."
• The Controller (The Bouncer): The intersection has a local computer (like a bouncer at a club) that keeps a schedule. It doesn't tell the car how to drive; it just checks if the car's plan overlaps with anyone else's.
• The Negotiation (The Handshake): As a car approaches the intersection, it enters a "negotiation zone" (a stretch of road before the intersection). It sends its plan to the bouncer.
  • If the path is clear, the bouncer says, "Go ahead!"
  • If the path is blocked by another car, the bouncer says, "Wait 2 seconds, then go."
  • The car adjusts its speed slightly and tries again.
  • Once agreed, the car glides through the intersection without stopping.

3. The "Conflict Zones": The Dance Floor Rules

To prevent crashes, the intersection is divided into invisible "Conflict Zones." Think of these as single-person dance circles.

• Only one car can occupy a specific circle at a specific time.
• The system ensures that Car A's circle and Car B's circle never overlap in time.
• If Car A is in the middle of the intersection, Car B must wait until Car A leaves that specific circle before entering.

4. Dealing with Real-World Glitches (The Internet Problem)

A major hurdle for these systems is that real-world internet (4G and 5G) isn't perfect. It has delays (lag) and sometimes drops messages.

• The Analogy: Imagine trying to coordinate a dance over a walkie-talkie with static. If the message takes too long to arrive, the dancers might crash.
• The Fix: Moveover is designed to be robust. It was tested with the delays typical of 4G and 5G networks. Even with the "static," the system works. If the negotiation takes too long or the internet is too slow, the system has a Backup Mode: the cars simply slow down and follow standard safety rules (like stopping at a yield sign) until the connection is restored.

5. Why It's Better (The Results)

The researchers tested this in a computer simulation with different types of intersections (crossroads, roundabouts, and busy city maps).

• Speed: Cars got through intersections 25% faster on average.
• Pollution: Because cars didn't have to stop and accelerate repeatedly, they emitted 25% less CO2.
• Capacity: The intersection could handle more cars per hour without turning into a gridlock.

The Big Picture

Think of Moveover as turning a chaotic, stop-and-go traffic light system into a smooth, flowing river. Instead of cars waiting in line, they weave through the intersection like water flowing around rocks, guided by a smart system that knows exactly where every drop of water (car) is going.

It's a system that respects the unique speed of every vehicle, uses the internet to coordinate a perfect dance, and keeps everyone moving safely and efficiently.

Technical Summary

Here is a detailed technical summary of the paper "V2N-Based Algorithm and Communication Protocol for Autonomous Non-Stop Intersections" by Farina et al.

1. Problem Statement

Intersections are critical bottlenecks for road safety and traffic efficiency, accounting for a significant percentage of fatal crashes and congestion. While Connected and Autonomous Vehicles (CAVs) offer a solution, existing approaches suffer from three main limitations:

• Computational Complexity: Many solutions rely on centralized controllers using heavy optimization tools (e.g., Mixed-Integer Linear Programming) or Reinforcement Learning, which struggle to scale in high-density traffic and lack explainability.
• Lack of Vehicle Dynamics Integration: Most algorithms focus on scheduling (who goes when) but do not explicitly optimize the vehicle's kinematic profile (speed and acceleration), missing opportunities to reduce emissions and energy consumption.
• Idealized Communication Assumptions: Many proposals assume perfect, zero-latency communication. In reality, network delays (4G/5G) and packet losses can degrade safety and efficiency, often causing systems to fail when delays exceed specific thresholds (e.g., >247 ms).

2. Methodology: The Moveover Algorithm

The authors propose Moveover, a novel algorithm and communication protocol designed for Vehicle-to-Network (V2N) interactions. The core philosophy is decentralized trajectory planning combined with centralized conflict management.

Core Principles

• Decentralized Trajectory Generation: Unlike centralized approaches where the controller calculates paths for all vehicles, Moveover delegates the calculation of the mobility profile (trajectory and speed) to the individual vehicle (the EGO). The EGO calculates its own optimal path based on its specific kinematic constraints (acceleration, braking, turning limits) and intended direction.
• Centralized Conflict Resolution: A local intersection controller (hosted on a Multi-Access Edge Computing server) manages a Scheduling Table. It does not plan paths but validates them. It ensures safety by enforcing deterministic conflict zone reservations.
• Conflict and Negotiation Zones:
  • Conflict Zones: Specific road segments within the intersection where collisions could occur. Only one vehicle can occupy a zone at a time.
  • Negotiation Zones: Areas before the intersection where the EGO and Controller exchange messages to agree on the mobility profile.

The Communication Protocol

The protocol operates via a request-response mechanism between the EGO and the Controller:

1. Proposal: The EGO enters the negotiation zone and sends a proposed mobility profile (sequence of positions and timestamps) to the controller.
2. Validation: The controller checks the proposal against the Scheduling Table (reservations of other vehicles) for conflicts:
   • Before intersection: Ensures safe distance from the preceding vehicle in the same lane.
   • Within intersection: Ensures no overlap with reserved conflict zones.
   • After intersection: Ensures safe spacing from the vehicle exiting the same lane ahead.
3. Response:
   • Acceptance: If compatible, the controller implicitly accepts the profile and records the time intervals for the conflict zones.
   • Rejection/Adjustment: If conflicts exist, the controller returns the earliest available time windows for the conflict zones. The EGO recalculates a new profile adhering to these windows.
4. Backup Mode: If negotiation fails (e.g., due to excessive delay or inability to find a valid profile), the system switches to a "backup mode" where vehicles revert to autonomous driving with legacy priority rules (e.g., stop-and-go).

Handling Non-Ideal Networks

The system explicitly models communication delays (0–10 ms for 5G, 20–50 ms for 4G) and queueing at the controller. The negotiation length (distance before the intersection) is dynamically calculated based on network latency to ensure the protocol completes before the vehicle reaches the intersection.

3. Key Contributions

1. Novel Algorithm (Moveover): A fully explainable, deterministic algorithm that scales to large intersections by offloading trajectory optimization to the vehicle and reserving the controller for conflict checking.
2. Kinematic-Aware Control: Unlike previous works, Moveover explicitly optimizes speed profiles to minimize stops and smooth acceleration, directly targeting emission reduction.
3. Robustness Evaluation: The first comprehensive evaluation of an autonomous intersection algorithm under realistic non-ideal network conditions (4G and 5G delays) across multiple intersection types (single-lane, multi-lane, roundabouts).
4. Standardization Alignment: The protocol is designed to align with emerging SAE and ETSI standards for Maneuver Coordination Services (MCS), utilizing concepts like Target Road Resources (TRR).
5. Open Source: The authors provide open-source code for the algorithm, controller, and simulation scenarios to ensure reproducibility.

4. Results

The algorithm was evaluated using the SUMO traffic simulator across four intersection types and a real-world urban map (Verona, Italy).

• Performance vs. Benchmarks: Moveover significantly outperformed traditional Traffic Lights, Priority Rules, and FIFO (First-In-First-Out) algorithms.
  • Travel Time: Reduced travel times by approximately 25% in multi-intersection scenarios compared to priority rules.
  • Emissions: Achieved substantial reductions in CO2 emissions due to the elimination of stop-and-go maneuvers.
• Scalability & Capacity:
  • Ideal/5G: Moveover increased intersection capacity by 0.08–0.15 veh/s compared to the best benchmarks (Traffic Lights or FIFO). In a 4-way two-lane scenario, it supported up to 1.2 veh/s.
  • 4G Impact: While 4G introduces latency, Moveover remained robust. However, in high-density 4-way two-lane scenarios, 4G delays caused the negotiation zone to extend, triggering backup mode at lower densities (capacity dropped to 0.68 veh/s vs. 1.2 veh/s for 5G).
• Message Overhead: The negotiation process is highly efficient. The vast majority of negotiations (over 90%) required only 2 to 4 messages (proposal + confirmation). Even in worst-case scenarios, message exchange rarely exceeded 8 messages.
• Real-World Validation: On a real urban map with four intersections, Moveover reduced average travel time from 51.3s to 37.7s and CO2 emissions from 0.16kg to 0.12kg.

5. Significance

This paper bridges the gap between theoretical autonomous intersection management and practical deployment.

• Practicality: By acknowledging and modeling 4G/5G latency, it provides a realistic roadmap for implementation in current and near-future networks.
• Efficiency: It demonstrates that non-stop intersections are feasible without heavy centralized computation, making the system scalable and cost-effective.
• Environmental Impact: The focus on kinematic optimization proves that autonomous intersection management can be a powerful tool for reducing urban pollution, not just improving traffic flow.
• Safety: The deterministic, reservation-based approach ensures safety even when communication is imperfect, with a clear fallback mechanism (backup mode) to prevent gridlock.

In conclusion, Moveover represents a significant step toward the realization of efficient, safe, and environmentally friendly autonomous urban mobility.