Abstract:In this paper, we propose a novel road side unit (RSU)-assisted cooperative sensing scheme for connected autonomous vehicles (CAVs), with the objective to reduce completion time of sensing tasks. Specifically, LiDAR sensing data of both RSU and CAVs are selectively fused to improve sensing accuracy, and computing resources therein are cooperatively utilized to process tasks in real time. To this end, for each task, we decide whether to compute it at the CAV or at the RSU and allocate resources accordingly. We first formulate a joint task placement and resource allocation problem for minimizing the total task completion time while satisfying sensing accuracy constraint. We then decouple the problem into two subproblems and propose a two-layer algorithm to solve them. The outer layer first makes task placement decision based on the Gibbs sampling theory, while the inner layer makes spectrum and computing resource allocation decisions via greedy-based and convex optimization subroutines, respectively. Simulation results based on the autonomous driving simulator CARLA demonstrate the effectiveness of the proposed scheme in reducing total task completion time, comparing to benchmark schemes.

What problem does this paper attempt to address?

### Problems the paper attempts to solve This paper aims to solve the problem of overly long completion time faced by Connected Autonomous Vehicles (CAVs) when performing perception tasks. Specifically, by introducing a Road - Side Unit (RSU) - assisted cooperative perception scheme, the paper hopes to reduce the total completion time of perception tasks while ensuring that the perception accuracy meets the requirements. ### Background and challenges 1. **System heterogeneity** - CAVs in different locations have different Regions of Interest (RoI), and the quality of data collected by different CAVs also varies due to different viewing angles. - The computing power of RSU and CAVs differs by dozens of times. This system heterogeneity has a significant impact on the effectiveness of cooperative perception. An inappropriate cooperation scheme may lead to an increase in task completion time and a decline in perception accuracy. 2. **Coupled decisions** - In edge - assisted cooperative perception, the location of task processing (i.e., task placement decision) and the amount of allocated spectrum and computing resources (i.e., resource allocation decision) are coupled, which increases the difficulty of problem - solving. - Therefore, it is necessary to design low - complexity algorithms to make task placement and resource allocation decisions reasonably. ### Solutions 1. **Propose a new cooperative perception scheme** - Utilize the computing power and perception data provided by RSU to assist each CAV in completing perception tasks. - Improve perception accuracy by fusing LiDAR data of RSU and CAVs. 2. **Optimization problem modeling** - Model the joint optimization problem of task placement and resource allocation as a mixed - integer nonlinear optimization problem, with the goal of minimizing the total task completion time while satisfying the perception accuracy constraints. 3. **Two - layer algorithm design** - **Outer - layer algorithm**: Make task placement decisions based on Gibbs Sampling Theory. - **Inner - layer algorithm**: Allocate spectrum resources and computing resources respectively through Greedy - Based Algorithm and Convex Optimization Methods. ### Simulation results - **Convergence performance**: The algorithm converges the fastest when \(\tau = 10^{-4}\) and can converge to the global optimal solution. - **Perception accuracy**: With the assistance of RSU, the perception accuracy of each CAV has been improved and exceeds the threshold of 0.85. For example, the accuracy of CAV 1 has been improved by 38.7%. - **Task completion time**: The proposed scheme shows the lowest total task completion time under different computing resource conditions. Especially when the CAV computing resources are limited, the task completion time is significantly reduced. ### Conclusion This paper designs a CAVs cooperative perception scheme assisted by RSU. By efficiently using the computing and perception capabilities provided by RSU, it significantly reduces the total completion time of all tasks and meets the perception accuracy requirements of CAVs. Future work will study a mathematical model that can accurately characterize the impact of data quality on perception accuracy.

Edge-Assisted Accelerated Cooperative Sensing for CAVs: Task Placement and Resource Allocation

Accuracy-Aware Cooperative Sensing and Computing for Connected Autonomous Vehicles

A Vehicular Edge Computing-Based Architecture and Task Scheduling Scheme for Cooperative Perception in Autonomous Driving

Communication and Computing Resource Optimization for Connected Autonomous Driving

Cooperative Localisation for Multi-Rsu Vehicular Networks Based on Predictive Beamforming

Balancing Computation Speed and Quality: A Decentralized Motion Planning Method for Cooperative Lane Changes of Connected and Automated Vehicles.

Hybrid Sensing Data Fusion of Cooperative Perception for Autonomous Driving with Augmented Vehicular Reality

Cooperative Environmental Perception Task Offloading for Connected and Autonomous Vehicles

Delay Optimization in Vehicular Edge Computing with Sensing Information Fusion and Heterogeneous Tasks

Cooperative Connected Autonomous Vehicles (CAV): Research, Applications and Challenges

Reinforcement Learning-Based Resource Allocation for Multiple Vehicles with Communication-Assisted Sensing Mechanism

EdgeCooper: Network-Aware Cooperative LiDAR Perception for Enhanced Vehicular Awareness

Cooperation Method of Connected and Automated Vehicles at Unsignalized Intersections: Lane Changing and Arrival Scheduling

Multi-objective Cooperative Scheduling of CAVs at Non-Signalized Intersection

Optimizing the Placement of Roadside LiDARs for Autonomous Driving

Reconfigurable Intelligent Computational Surfaces for MEC-Assisted Autonomous Driving Networks: Design Optimization and Analysis

Enabling Road Detection Tasks via Collaborative Smart Vehicles

Cooperative and Comprehensive Multi-Task Surveillance Sensing and Interaction System Empowered by Edge Artificial Intelligence

Time-division ISAC Enabled Connected Automated Vehicles Cooperation Algorithm Design and Performance Evaluation

Deep Reinforcement Learning-Based Task Offloading for Vehicular Edge Computing With Flexible RSU-RSU Cooperation

CoOR: Collaborative Task Offloading and Service Caching Replacement for Vehicular Edge Computing Networks