Optimization of Vehicular Networks in Smart Cities: From Agile Optimization to Learnheuristics and Simheuristics

Abstract

Vehicular ad hoc networks (VANETs) are a fundamental component of intelligent transportation systems in smart cities. With the support of open and real-time data, these networks of inter-connected vehicles constitute an `Internet of vehicles’ with the potential to significantly enhance citizens’ mobility and last-mile delivery in urban, peri-urban, and metropolitan areas. However, the proper coordination and logistics of VANETs raise a number of optimization challenges that need to be solved. After reviewing the state of the art on the concepts of VANET optimization and open data in smart cities, this paper discusses some of the most relevant optimization challenges in this area. Since most of the optimization problems are related to the need for real-time solutions or to the consideration of uncertainty and dynamic environments, the paper also discusses how some VANET challenges can be addressed with the use of agile optimization algorithms and the combination of metaheuristics with simulation and machine learning methods. The paper also offers a numerical analysis that measures the impact of using these optimization techniques in some related problems. Our numerical analysis, based on real data from Open Data Barcelona, demonstrates that the constructive heuristic outperforms the random scenario in the CDP combined with vehicular networks, resulting in maximizing the minimum distance between facilities while meeting capacity requirements with the fewest facilities.

Optimization Mobility in Sensor Actor Networks: A Comprehensive Framework

“…The interactions between the various components of the network framework are intended to validate and enhance mobility protocols, thereby augmenting the efficiency and adaptability of sensoractor communication and coordination within the virtual environment. The table enumerates the elements that are necessary for sensor-actor network validation and optimisation in the context of mobility [11]. The main processing unit is an Arduino Uno R3 (U1), the actuation is handled by a Micro Servo (SERVO1), and the ambient data is collected by a Temperature Sensor (U3) similar to the TMP36.…”

Machine learning based centralized dynamic clustering algorithm for energy efficient routing in vehicular ad hoc networks

“…4. Finally, NetSim performs two types of VANET applications: (1) data uploading is called V2R communication, and (2) data uploading and downloading is called V2R + R2V communication.…”

“…A unique method of communication amongst vehicles is needed as they become more conscious of their surroundings and move closer to being completely independent. The VANET is an important component of the intelligent transportation system (ITS) 1 and smart city, 2 which is a rapidly growing research field. Wireless access in vehicle environments (WAVE) defined by IEEE 1609 and dedicated short range communications (DSRC) by IEEE 802.11P 3 are major standard protocols of ITS that enable Vehicle‐to‐Vehicle (V2V), 4 Vehicle‐to‐Infrastructure (V2I), and Vehicle‐to‐Everything (V2X) communications in VANET.…”

Smart Cities and City Logistics

“…The problem is even more pronounced because it must be solved in real time and an uncertain and dynamic environment. In the paper (Peyman, et al, 2023 Jan), one such simulation was performed on the example of Barcelona using data downloaded from Open Data Barcelona. The results showed that the constructive heuristic outperforms the random scenario in the customer data platform combined with vehicular networks.…”

“…The results showed that the constructive heuristic outperforms the random scenario in the customer data platform combined with vehicular networks. Better results are achieved in shortening the distance between facilities with meeting capacity requirements using the fewest facilities (Peyman, et al, 2023 Jan).…”