networks to enhance smart parking systems, Improving energy efficiency and reducing search times for available spaces.

The The Internet of Things (IoT) is growing incredibly, and innumerable smart devices are being connected to the internet. Such massive interconnection potential generates numerous opportunities and challenges. Providing high-quality services in networks raises a multitude of difficult challenges. One of them is to determine an optimal path that honors a given set of constraints while allowing the usage of network resources. This implies the need to impose an additional optimality. Thus, the feasibility problem includes a requirement. At present, transport relies mainly on fossil fuels like coal and oil. All of these come with serious effects on the environment. These fuel burn-ups emit greenhouse gases that cause increases in world temperatures, thereby influencing the climate. These contribute to many environmental and climatic issues, melting glaciers, rising sea levels, and extreme climatic phenomena. In a bid to mitigate these adverse effects, scientists are conducting research into sustainable modes of transport to reduce the use of fossil fuels while also emitting less carbon at the same time. Also, the most important challenge of this trend is the energy consumption level because these devices always communicate wirelessly. Another challenge associated with IoT devices is that they are heterogeneous, meaning their capabilities, energy levels, and processing powers are not uniform.

All these challenges must be overcome with utmost importance for sustainable growth and effective operation of IoT systems. This paper uses an algorithm to solve the parking guidance problem in smart cities. The problem at hand addresses finding the shortest route, about distance, for a driver from the driver's present location to a parking space under certain conditions, including distance, traffic conditions, and time constraints. The authors address the challenge with the development of a heuristic algorithm based on Ant Colony Optimization (ACO). The ACO is the metaheuristic describing how a set of artificial ants simulates the behavior of real ants in their search for the shortest route that connects the nest to the food.

Regarding parking guidance, virtual ants propagate through the network and mark the best paths along roads with pheromones, such that gradually increasing levels of optimal routes, guiding subsequent virtual ants to choose such paths. The proposed algorithm will consider the driver's destination and optimize the route so that the vehicle reaches the parking space without traveling less time and distance than necessary to the final destination. This algorithm provides a powerful and efficient solution to the parking guidance problem with the help of ACO and improves mobility in the urban setting while reducing congestion in traffic. Section 4 evaluates the existing literature on the area under investigation after explaining the research challenges. The ACO algorithm is defined in Section 5. Section 6 describes the mathematical model to solve the parking routing problem based on the ACO Algorithm. The perspectives and conclusion are given in the last two sections.

This paper suggests a smart parking system that uses Wireless Sensor Networks (WSN), AI, and machine learning (ML) to make parking vehicles more efficient and secure. Using an SVM-based algorithm, The system addresses congestion, energy efficiency, and routing optimization. The Firefly algorithm (FFA) optimizes SVM parameters for efficient route discovery. The Energy-Aware Trust-Based Algorithm (EATSRA) detects malicious nodes by evaluating node trust scores. Nodes with low trust values are removed from the route to maintain secure communication [1].

The proposed MHSEER protocol enhances industrial operations in WSN-IIoT by ensuring secure, energy-efficient, and reliable data transfer while minimizing resources. It uses a clustering process to reduce overhead and extend network life. The protocol regularly adjusts cluster sizes to conserve energy and avoid reliance on Cluster Head (CH) nodes. It operates in two phases: selecting the next hop using a metaheuristic approach and integrating a secure protocol with CEM encryption for data transmission. Thus, MHSEER optimizes routing, enhances energy usage, and reduces packet delays through efficient next-hop decisions [2]. This study addresses the limitations of EVs by proposing an energy consumption model, the Amplified Ant Colony Algorithm (A²CO), which balances energy consumption with other heuristic factors (speed, SoC, and distance) in an objective function. The model is tested on the transportation map of Chandigarh, India finding that EV energy consumption is inversely proportional to speed and that the number of artificial ants affects the optimization outcome, with 20 ants yielding the best results [3]. This paper introduces a navigation and reservation-based technique that relies on low-cost devices to transmit IoT data for searching available parking spaces in smart cities. This is used to minimize the distance towards closer parking spots using a genetic algorithm for optimization, which provides what is also known as the evosuite method, which helps with issues with the nearest available parking slot. The result is more accurate, and parking management works [4]. This paper introduces a navigation and reservation-based technique that relies on low-cost devices to transmit IoT data for searching available parking spaces in smart cities. This is used to minimize the distance towards closer parking spots using a genetic algorithm for optimization. It provides what is also known as the evosuite method, which helps with issues with the nearest available parking slot. The result is more accurate, and parking management works [5]. In this paper, a new parking navigation system is developed to search for the nearest available empty slots using the IoT and a large-scale distributed Ant Colony Algorithm (ACO). The system considers several factors, such as the time needed to reach an indoor parking lot, road conditions, the number of available spots, and the vehicle's fuel level. The feature would allow drivers to quickly find the nearest parking and receive directions on how to reach that location. The idea is to enhance parking efficiency by facilitating the routing process [6].

This paper introduces an energy-efficient routing algorithm for WSNs called IHSBEER, which improves the Harmony Search (HS) algorithm (a meta-heuristic) by enhancing memory encoding and using the roulette wheel method to speed up convergence. This enables the new algorithm to outperform existing ant-based algorithms regarding energy savings and network lifetime achieved, as it improves energy efficiency and path length [7]. This paper explored the best way to place WSNs for fire monitoring in a smart car park by minimizing the number of Sensor nodes (Sn), relay nodes, and network size. Since this problem is complex, an algorithm has been developed that combines chaos theory with the Whale Optimization Algorithm (WOA) to get better results in less time than the standard WOA. This algorithm produced results better in computational time and solution quality [8]. A new method of routing data in IoT networks so that the nodes can move around has been introduced. This new method structures the network into a tree form, like the family tree, to optimize the path taken by data from the sensors to a moving receiver. The new method has proven to be much more reliable and energy-efficient than the older methods. This will ensure that the data is delivered to the receiver without any loss and distribute the energy consumption evenly among devices. This new method has been proven superior to existing methods through tests. They plan to extend this method to work with multiple moving receivers, adding even greater versatility in the future [9].

Vehicle traffic management in urban regions faces such stern challenges because of dependence on personal vehicles. Different research is being carried out about these issues using innovative technologies. Rizwan et al., introduce a real-time traffic management system through IoT-based detection sensors designed to analyze traffic flow. Fernandez-Ares follows the movement of traffic and pedestrians through Wi-Fi and Bluetooth signals. Sendra et al. developed a sensor network related to LoRa-based networking systems for monitoring pollution and managing traffic circulation. Kazmi et al. developed a VITAL-OS IoT platform that may manage data streams to detect traffic noise for rerouting. Kök et al. applied LSTM deep learning models to predict air quality in smart cities. Jabbarpour et al. proposed an IoT application named Intelligent Guardrails, which can exploit vehicular networks for traffic management. These technologies aim to reduce the loss caused by traffic due to congestion while ensuring environmental protection [10].

Abualigah et al. [11] designed metaheuristic algorithms, such as TLBO, based on the classroom learning process between teachers and students to optimize functions with multiple variables. The algorithm takes a randomly generated population as an initial solution and evolves iteratively until a valid solution is obtained. Singh et al. offer a visual Big Data analytics framework for identifying automatically bikers without helmets in city traffic. Next, Pawłowicz et al. designed a traffic management system by inter-relating 5G communication, RFID-based parking monitoring, cloud services, and learning for supervision. Rathore et al. have proposed smart digital city systems that operate on IoT devices and Big Data analytics to assemble real-time city data and knowledge extraction [12]. A technique for optimization based on the amount of energy used in IoT develops optimal paths with efficient routes used to energize an IoT network. This paper develops a novel algorithm called the Chaos Fuzzy Grasshopper Optimization Algorithm, FLGOA, developed from the principle of Lorenz chaos theory applied during initialization and the fuzzy approach applied to parameter adaptation. The performance of the proposed technique is assessed by three criteria: remaining energy, network lifetime, and coverage rate. The tests were performed using two scenarios: efficiency over time and efficiency per number of nodes. The results confirm that the proposed algorithm FLGOA outperforms the base methods. For example, FLGOA shows a 9% longer network life than FGOA, 12% longer in comparison with GOA, and 16% longer than GSO. Besides that, FLGOA provides 16% better-remaining energy than the other methods, 21% better than GOA, and 22% better than GSO. In addition, the coverage rate for FLGOA is 12-16% higher than for base methods [13]. Energy efficiency and optimal load equilibrium between sensing devices are fundamental challenges due to the limitation of energy resources in IoT devices. Considering the cruciality of prolonging IoT networks and the NP-hard nature of energy optimization in heterogeneous and distributed IoT infrastructures, this research represents a hybrid energy-aware protocol for data routing in IoT. It is deployed based on ML and metaheuristic HTOA (Heat Transfer Optimizer). The proposed protocol adopts a four-phase approach for fuzzy clustering, predicting energy consumption based on Support Vector Regression (SVR) and time series techniques, selecting optimal CHs considering energy and centralization factors, and routing and data transmission from each sensor to and from CH to the sink using ACO algorithm.

The simulation outputs revealed that our scheme performs more efficiently in terms of energy consumption, End-to-End Delay (EtED), load balancing, overhead, and network lifetime in all designed scenarios compared to other methods. Heuristics are computational methods to find feasible solutions for complex optimization problems that often implement randomness or local search. Metaheuristics, like the GA and PSO, rely on two key components, intensification, and diversification, whose balance becomes very important for the quality of the solution. GA starts with an initial population of solutions. At each iteration, it chooses the fittest solutions based on a fitness function so there is no premature convergence problem. It uses recombination and mutation to produce new solutions until it finally converges to an optimal solution. PSO is a nature-inspired algorithm in which the behavior of swarming animals, like birds, fish, etc., is mimicked. It is rather efficient for global optimization. PSO is applied to solve complex problems in scientific research and engineering. It uses principles from swarm intelligence to improve the quality of the solution iteratively. Today, one of the commonly used techniques in the evolving computation for solving optimization problems in GA and PSO is common [14].

Heuristics are computational methods to find feasible solutions for complex optimization problems that often implement randomness or local search. Metaheuristics, like the GA and PSO, rely on two key components, intensification, and diversification, whose balance becomes very important for the quality of the solution. GA starts with an initial population of solutions. At each iteration, it chooses the fittest solutions based on a fitness function so there is no pre-mature convergence problem. It uses recombination and mutation to produce new solutions until it finally converges to an optimal solution. PSO is a nature-inspired algorithm in which the behavior of swarming animals, like birds, fish, etc., is mimicked. It is rather efficient for global optimization. PSO is applied to solve complex problems in scientific research and engineering. It uses principles from swarm intelligence to improve the quality of the solution iteratively. Today, a commonly used technique in the evolving computation for solving optimization problems in GA and PSO is common [15]. To enhance security, the sensor network is equipped with high throughput. While the importance of both Source Node (SN) and Base Station (BS) location privacy and security is acknowledged, recent research has predominantly focused on location privacy. To address security challenges in WSN, we propose an optimized meta-heuristic clustering-based privacy key-agreement routing technique. In the suggested system, a gateway-based network is constructed to devise a key arrangement protocol that promotes privacy during communication. The proposed routing strategy involves forming clusters of Sn, facilitating the efficient selection of CHs that prioritize nodes with the least modification. This effectively addresses the EC problem. A comprehensive performance evaluation is conducted, considering improvements in energy efficiency, Packet Delivery Ratio (PDR), throughput, EtED ( E^2 delay), and EC [16]. Energy efficiency nowadays has become a main issue in WSNs. Hierarchical clustering with a multipath routing protocol technique is important for improving packet overhead, network lifetime, QOS, and power consumption. Many such methods are suggested to Improve the Energy efficiency of the whole WSN region. The Ad-hoc On-Demand Distance Vector (AODV) routing protocol is very suitable, as it is more scalable and less overhead. This AODV protocol has two operations to find and maintain routes, i.e., Path discovery and path maintenance. Using the Clustering approach, the data packet is shared among members of different clusters with the help of the CH, which ultimately saves energy. Hence, this approach uses a Hierarchical clustering algorithm and a Hybrid Genetic Algorithm (GA) with a Particle Swarm Optimization (PSO) algorithm.The GA and PSO algorithm creates a hierarchy of CHs. The energy-saving scheme increases with the increase in the cluster present in WSN.

Therefore, hierarchical clustering with hybrid GA and PSO (HC-HGAPSO) methodology performed better regarding throughput, network lifetime, and residual energy [17]. Parking is scarce in congested metropolitan cities. The parking management system is the solution for the emerging parking issues in these modern cities. This paper proposes a smart parking management system based on WSNs that can monitor the parking slot's occupancy status and communicate the availability to the end users. The energy-aware trust-based routing algorithm calculates the trust score of each participating node in the route. The FFA algorithm and artificial neural network classifier are used, and the comparative results have proven the proposed scheme to be better in QoS parameters [18].

The issue of finding parking space is a serious concern in some locations, especially shopping complexes, hospitals, buildings, malls, and other premises that require large parking spaces for vehicles. Parking worsens on special occasions like festivals, fiestas, and weekends. Conventional techniques that install sensors in parking zones for parking vehicles have become costly. Therefore, a reliable method that works for a long time is required to manage traffic congestion while searching for parking spaces. This paper proposes a novel method for parking vehicles using metaheuristic approaches [19].

Their work emphasizes algorithms and optimization techniques, often by a metaheuristic and other approaches for improved energy efficiency.Wireless communication using the IoT is always under pressure as far as energy consumption and heterogeneity are concerned.Complex ecosystems in wireless communication in IoT make efficient data transmission difficult.Traditional ad-hoc routing protocols, including AODV, LSR, and OSPF, depend too much on energy inefficiency and scalability.Few routing protocols consider determining factors like residual energy, load, and traffic flow.The configuration of parameters in IoT routing is difficult and often NP-complete. EV routing routing has not been extensively researched but is a growing field. Conventional VRP-type algorithms can be inapplicable to EVs for several reasons. Many search algorithms do not take into account energy consumption, for instance. The EV battery has very few control issues and temperature constraints when removing and applying power to the battery. The battery under high and low temperatures will reduce capacity without damage, to the face more on the EV system developers.

In addition, the simulation model used in this paper is simple but has enough ability to describe the actual situation. As the research will further develop into a large-scale publication that uses the Geographic Information System (GIS), some meta-heuristic methods such as PSO, the ant colony algorithm, or Tabu Search, that take into account the number of the regulations and constraints involving such as large-scale case for example, are suggested [20]. The development of smart cities focuses on improving urban life quality through smart infrastructure and technology, including optimizing transportation and reducing environmental impacts. A key challenge is the Travelling Salesman Problem (TSP), which seeks the shortest route that visits a set of points and returns to the starting point. In smart city contexts, IoT devices provide vast amounts of data that can enhance route optimization. To improve performance, this work explores solving the TSP using a modified Teacher Learner Based Optimization (TLBO) algorithm implemented on a parallel GPU architecture (CUDA). Smart cities integrate IoT, RFID, and LoRaWAN technologies to enhance mobility and traffic management. For example, sensors track vehicle movement and environmental data, while systems like VITAL-OS manage IoT data streams. These technologies support real-time traffic control and data-driven decision-making, improving urban mobility and reducing fuel consumption [21]. This article presents a two-level hierarchical chain routing scheme, PEG_ABC, based on the Bee Colony Algorithm (BCA) to optimize energy consumption and reduce transmission delay in IoT-based WSNs. PEG_ABC addresses the challenges of long transmission delays in chain communication by considering factors like sensor energy, Euclidean distances, and BS proximity in the fitness function. At the first level, it selects optimal chain leaders. At the second level, it forms end-to-end paths for data transmission between leaders, further adjusting the fitness function to minimize delays and maximize energy efficiency. Simulation results demonstrate improved network lifetime, reduced transmission delays, and better energy utilization than existing methods.

The paper also explores the optimal number of clusters for hierarchical chain communication. Inspired by bee foraging behavior, the Bee Colony Algorithm is employed to find the most energy-efficient paths and reduce network load. This approach is applied in various IoT domains, including smart healthcare, smart cities, and industrial systems. The proposed method improves upon traditional PEGASIS and ant colony-based schemes regarding network performance. This paper addresses energy-efficient routing in Software-Defined Wireless Sensor Networks (SD-WSN), focusing on optimizing CH selection using a hybrid nature-inspired algorithm called HGWO-BC. The hybrid approach combines Grey Wolf Optimization (GWO) and PSO to improve convergence and performance. The algorithm optimizes network lifetime by considering factors like intra-cluster distance, CN selection, and node energy.

HGWO-BC aims for balanced energy consumption and scalable clustering, ensuring efficient routing in dense, battery-constrained WSNs. Extensive simulations demonstrate its effectiveness in maximizing network lifetime compared to traditional protocols like LEACH. The SD-WSN model includes a centralized controller that manages resources and monitoring tasks, enhancing real-time adaptability. The paper emphasizes the need for efficient WSN architectures to handle the growing complexity of IoT applications [22].

This paper presents a hybrid data routing algorithm for energy-efficient WSNs in IoT applications, combining swarm optimization with an energy-efficient heuristic approach. The Particle Swarm Optimized Residual Energy-based Stable Election Protocol (PRESEP) improves network lifetime, stabilizes cluster formation, and minimizes energy consumption by optimizing CH selection. It outperforms previous algorithms regarding network scalability, alive nodes, and reduced energy use. WSNs, often deployed in remote areas, face challenges due to limited power sources and energy scarcity. The paper also addresses issues of data redundancy and overload in WSNs, which are crucial for efficient IoT operations [23].

The architecture (Fig. 1) is composed of a set of elements linked by a cloud platform. These elements are equipped with intelligent technologies that enable the collection and transmission of data to the cloud for processing.

Hence, intelligent sensors should be incorporated in all car parks, providing information on the status and number of available spaces. Each road needs to have a certain number of sensors to get a real-time update on the traffic situation.

A specialized agent will relay the information concerning each service station to the cloud. Considering all these data, the system should be capable of giving the driver the best path to follow in the shortest distance and with the least congestion. Information about each service station will be transmitted to the cloud through a specific agent. A. Search for the Occurrence of Physical Objects to Address Infrastructure Phenomena Based on the various requests made by the driver, it is necessary to strategically place several sensors in every indoor parking area so that proper monitoring is done to provide a measurement of parking availability in real-time. Information gathered by various sensors is sent to such an agent, who will then send the information to a cloud-based system through the MQTT communication protocol.

MQTT [18] (Message Queue Telemetry Transport) Protocol is a publish-subscribe-based "lightweight" messaging protocol that is used on top of the TCP/IP protocol. B. The description of the parking routing system The model of the parking routing system provides access to data and computing capabilities located in the cloud. The clouds offer data storage and computing resources for the parking routing system. When in the cloud, requesting the driver, the system will always catch the Coordinates of his real-time position. To retrieve the traffic flow in a specific area we employ Road-Side Units (RSUs). Each RSU receives data from the neighbors and estimates the junction traffic with the current count and potential vehicles from surrounding intersections. Constantly adjusting the communication overlapping distance of each cell allows RSU's to acquire approaching traffic information ahead. We will also deploy Google traffic services. Once all the appropriate information is available, we will invoke the structured algorithm that enables us to determine the optimal underground parking location that has available spaces and is reachable from the driver's current position. The solution must cater to the driver's preferences regarding distance, traffic, and time needed to accomplish the task.

In this paper, we present a new strategy for parking routing system design using the IoT with a distributed ant colony algorithm to enhance the quality of the solutions proposed. This solution is designed to help drivers locate the nearest available parking spaces. It will take into consideration a set of constraints such as arriving at indoor parking as soon as possible, road conditions, remaining spaces in each lot, and the quantity of fuel in the vehicle; the proposed distributed architecture includes a set of distributed intelligent agents; each of them has a predefined behavior, these agents interact with each other to determine the best as well as shortest paths between the driver's position and indoor parking with available spaces. In this research, our system becomes one of the valuable solutions for drivers finding the nearest parking and giving the best path to go toward that parking according to the ACO technique dedicated to solving problems related to the vehicle routing system.