Multi-objective Optimal Roadside Units Deployment in Urban Vehicular Networks (2402.18581v1)

Abstract: The significance of transportation efficiency, safety, and related services is increasing in urban vehicular networks. Within such networks, roadside units (RSUs) serve as intermediates in facilitating communication. Therefore, the deployment of RSUs is of utmost importance in ensuring the quality of communication services. However, the optimization objectives, such as time delay and deployment cost, are commonly developed from diverse perspectives. As a result, it is possible that conflicts may arise among the objectives. Furthermore, in urban environments, the presence of various obstacles, such as buildings, gardens, lakes, and other infrastructure, poses challenges for the deployment of RSUs. Hence, the deployment encounters significant difficulties due to the existence of multiple objectives, constraints imposed by obstacles, and the necessity to explore a large-scale optimization space. To address this issue, two versions of multi-objective optimization algorithms are proposed in this paper. By utilizing a multi-population strategy and an adaptive exploration technique, the methods efficiently explore a large-scale decision-variable space. In order to mitigate the issue of an overcrowded deployment of RSUs, a calibrating mechanism is adopted to adjust RSU density during the optimization procedures. The proposed methods also take care of data offloading between vehicles and RSUs by setting up an iterative best response sequence game (IBRSG). By comparing the proposed algorithms with several state-of-the-art algorithms, the results demonstrate that our strategies perform better in both high-density and low-density urban scenarios. The results also indicate that the proposed solutions substantially improve the efficiency of vehicular networks.

• The paper introduces AM-NSGA-III and AM-NSGA-III-c, algorithms that optimize RSU deployment by balancing total delay reduction and cost minimization.
• It employs a multi-population strategy and adaptive parameter tuning with constraint-handling techniques to address urban obstacles and variable traffic densities.
• Experiments using real-world traffic data demonstrate that the proposed methods outperform standard algorithms, offering effective trade-offs between RSU density and communication delay.

This paper addresses the critical problem of optimally deploying Roadside Units (RSUs) in urban vehicular networks to enhance communication efficiency and minimize infrastructure costs. The complex urban environment, characterized by obstacles and varying traffic densities, coupled with the conflicting nature of optimization objectives (e.g., minimizing delay vs. minimizing cost), presents significant challenges for RSU placement.

The problem is formulated as a multi-objective optimization problem with constraints. The three primary objectives are:

1. Minimize total communication delay: Aiming to reduce the overall latency experienced by all vehicles across all time periods.
2. Minimize maximum delay in latency-sensitive areas: Focusing on ensuring low latency in specific critical zones like accident-prone spots or busy intersections.
3. Minimize the number of deployed RSUs: To control the investment and maintenance costs.

The constraints considered include:

1. Obstacle avoidance: RSUs cannot be deployed in locations occupied by obstacles (buildings, gardens, etc.).
2. Minimum distance between RSUs: To prevent signal interference and ensure network stability, RSUs must be deployed with a minimum separation distance ($D_{min}$).

To solve this constrained multi-objective, large-scale optimization problem, the authors propose two enhanced versions of the NSGA-III algorithm: AM-NSGA-III and AM-NSGA-III-c. The core improvements and strategies incorporated are:

• Multi-population strategy: The overall population is divided into sub-populations that evolve independently but exchange superior solutions through a migration operator. This enhances the exploration of the large search space.
• Adaptive balancing of exploration and exploitation: Crossover and mutation rates are dynamically adjusted within each sub-population based on its performance (improvement in best fitness). Rates promoting exploration increase if fitness plateaus, while rates promoting exploitation increase if fitness improves.
• Constraint handling: An $\epsilon$-level comparison rule is employed to compare solutions that violate constraints. This method allows for a relaxed tolerance ($\epsilon$) for constraint violations during the optimization process, gradually reducing the tolerance over generations, which helps the algorithm navigate infeasible regions towards feasible ones.
• Offspring calibrating strategy (AM-NSGA-III-c only): A mechanism is introduced to adjust newly generated RSU deployment solutions to satisfy the minimum distance constraint. If RSUs are too close, the one covering a lower traffic volume is removed. This specifically addresses the RSU density constraint during the optimization process and aims to reduce the number of RSUs while maintaining coverage for high-traffic areas. AM-NSGA-III does not include this calibration.
• Iterative Best Response Sequential Game (IBRSG) for Data Offloading: To accurately evaluate the communication delay objectives, the paper proposes a data offloading strategy where vehicles iteratively determine their best RSU connection based on minimizing total latency for all vehicles. This game-theoretic approach models the interaction between vehicles and RSUs for data transfer.

The proposed algorithms were evaluated using real-world traffic data from Chengdu City, China, simulating RSU deployment in both high-density and low-density urban scenarios. Their performance was compared against the standard NSGA-III and MOEA/D using metrics such as the Number of Pareto-optimal Solutions (NPS), Number of Feasible Solutions (NFS), Inverted Generational Distance (IGD), Hypervolume (HV), and S-Metric.

The experimental results demonstrate that both AM-NSGA-III and AM-NSGA-III-c significantly outperform MOEA/D and standard NSGA-III in finding feasible and high-quality Pareto-optimal solutions. AM-NSGA-III-c, with its offspring calibrating strategy, showed a greater ability to find feasible solutions, especially in challenging high-density scenarios and when the number of latency-sensitive areas increased. The calibration helps manage the RSU density constraint effectively. While AM-NSGA-III-c tends to yield solutions with fewer RSUs (lower cost), AM-NSGA-III often produces solutions with more RSUs, leading to potentially lower latency. This highlights a trade-off between cost and delay, offering decision-makers different options on the Pareto front.

The IBRSG data offloading strategy was also evaluated independently against several other methods (Location-based, Signal strength-based, Random, GA-based, MCDM-based). IBRSG demonstrated a good balance between minimizing total latency and achieving load balance among RSUs, with competitive computation time.

In conclusion, the paper presents practical and effective algorithms for the multi-objective RSU deployment problem in complex urban environments. The proposed AM-NSGA-III variants, particularly AM-NSGA-III-c, are shown to be capable of handling constraints and large search spaces, generating a diverse set of high-quality, feasible deployment strategies. The IBRSG strategy provides a robust mechanism for evaluating communication performance within these deployments. Future work includes exploring the integration of edge computing and machine learning to further optimize vehicular network performance.