An intelligent parking management system using RFID technology based on user preferences

Due to the tremendous progress in the automotive industry, the growth of the urban population, the number of cars is increasing and this creates parking challenges. Intelligent parking management systems offer an optimal solution for finding empty parking spaces so that drivers can quickly find their car parking spaces. To solve these issues, it is necessary to design an intelligent parking system, besides providing comfort to drivers, which is also economically viable. In this paper, an intelligent multi-storey car parking management system called MODM-RPCP is proposed with the help of RFID technology and user preferences review, which effectively solves car parking issues. The MODM-RPCP approach is a multi-objective decision-making method that reduces the problem of average booking time and response time of the central parking management server. The simulation results show that the MODM-RPCP reduces the average booking time by more than 19.2 and 27.1%, and decreases the response time of the central parking management server more than 20.1 and 29.78% compared to MOGWOLA and ODPP approaches.


Introduction
Due to the increasing number of cars, traffic congestion is a worrying problem around the world and it is increasing day by day. Almost every day, about 40% of car traffic congestion in all countries of the world is due to the search for car parking space, and it takes an average of 10 to 12 min for the driver to find the parking space for the car. Such issues lead to traffic congestion, increased waiting times for drivers and wasted car energy due to parking searches. The car parking problem needs an optimal solution to save driver time, reduce pollution and economic losses (Lou et al. 2020;Juwita et al. 2020).
Smart parking systems can be solved with the help of innovative solutions, by integrating different resources to upgrade facilities and parking management. These parking systems can provide real-time updates to users about the nearest available parking spaces. It can also provide a smart parking management system for booking and checking empty spaces remotely. These parking systems consist of low-cost sensors, real-time data collection, and automated mobile payment systems for booking (Jung 2020). With RFID after intelligent parking identification, additional features such as fast car recovery, parking adjustment, parking gate management can also be used. Smart parking can be modeled as a parking gate. In each parking slot, a sensor is placed to detect the presence or absence of a vehicle that creates an access map to guide parking and other services. Such a system can also be considered a problem in managing multiple parking lots because it has to manage multiple parking lots distributed in different internal and external areas. To design intelligent parking systems, connecting the sensor measurement to a physical location is essential (Perković et al. 2020).
Instead of deploying all cars with GPS capability, it is preferable to have only a few cars on the network to identify their exact location via GPS. These cars are referred to as anchor or reference cars. Other network nodes will be able to show their position close to anchor nodes by measuring the received signal strength (RSS) and arrival time. These methods have made significant progress in computational accuracy and timing. Figure 1 shows the proposed method architecture where different drivers are communicating with the parking management server in to determine the empty parking spaces.

Problem statement
The main problem of this paper is managing and finding space in a smart parking lot that solves the two important issues of Average booking time and Response time of central parking management server in multi-storey car park. Previous methods have the problem of high response time in finding empty spaces in the parking lot, and many methods have high complexity in solving these issues. To solve these issues, an approach called MODM-RPCP has been proposed that solves both Average booking time and Response time of central parking with high efficiency.

Main contribution
This study presents a smart multilevel parking management system using RFID and checking users' priorities, which can efficiently solve parking issues. The proposed method called MODM-RPCP is a multi-objective decision-making method that reduces the parking problem.
• The main focus of the proposed approach based on drivers' priorities includes reducing average booking time and response time. • In this paper, an intelligent multi-storey car parking management system called MODM-RPCP is proposed with the help of RFID technology and user preferences review, which effectively solves car parking issues. • The MODM-RPCP approach is a multi-objective decision-making method that reduces the problem of average booking time and response time of the central parking management server. • This system manages parking spaces using RFID tags and CRPST parking metrics to reduce search time.
The rest of the paper is organized as follows: Sect. 2 presents related work. Section 3 brings the proposed MODM-RPCP solution. The parameters used for assessing the performance are studied and simulation outcomes are deliberated in Sect. 4. Finally, conclusion of this research is discussed in Sect. 5.

Related work
Cyprus International Airport decided to use the Hikvision smart parking system to minimize the issues caused by parking cars and protect them. They left the access control system, the smart airport parking system, and monitoring other parts of the airport to the closed-circuit camera system of Hikvision. In this solution of Hikvision, the barriers installed in the car park entrance are controlled smartly using sensors and special radar. Efficiency of this model has other advantages besides easy installation and setup Fig. 1 Multi-objective decisionmaking method to reduce the problem of car parking architecture which include not being affected by environmental factors like lighting, dust, and rain. Besides the automatic and smart barrier control capability, this system can leave some of the control manually to the operator. By using smart algorithms and modern technologies in the security and closed-circuit field, the Hikvision smart system can present valuable statistical and analytical information to the managers so that, if necessary, they can make important decisions using this information and increase security and customer satisfaction levels (Zong et al. 2019).
In some studies that Zung et al. conducted for understanding the decisions in car parks by modeling the structural equation, they have created a structural equation for analyzing parking decisions. The data used in this study was obtained from Information Park in Beijing. The relationships between the three parking decisions were studied. The results show a two-way correlation between street park and duration. These findings can be used for developing some measures for regulating interaction Analysis of parking in the car park and the parking mechanism for balancing parking time-time distribution and also formulating parking management policies (Maravall and Lope 2007).
Maravel et al. have solved the problem of automatic parking using a rear-wheel drive vehicle using a biometric model based on the direct connection between the perception of the vehicle and actions. This problem has been inspired using the external approach where the vehicle controller does not need to know the car communique and dynamics. Also, it does not require previous knowledge of the environment's map. The main point in the proposed approach is the definition of performance indicators that happen to automatic parking and actions are injected into the car robot controller in real-time. This solution is in the form of a multi-objective dynamic optimization problem and is highly analytical. Using the genetic algorithm, they have obtained a straightforward and effective solution (Gao et al. 2019).
In the studies Zhao et al. have conducted on analyzing the activity-based trip chain in the parking fee network program, they incorporated the chaining behavior of activity-based trip in analyzing network stability and an integrated model has been presented for describing the passenger's behavior, which is a combination of Beckman link congestion terms and types two logit demand function. The convexity conditions and equivalence of the model have been discussed. Based on the integrated model, a twolayer model has been designed for maximizing social welfare by suitable parking costs. Also, an extensive network for eliminating services and trips in the main network has been developed. Then, the Simulated annealing (SA) method was used for solving the proposed two-layer model. Numerical examples have been presented for studying the availability of the model and the effects of parking fee scheme on passengers' behavior and social welfare, which indicate that this model is effective in describing the trip chaining behavior in the network (Zhang et al. 2018).
To solve the traffic pressure caused by container trucks in ports, Chang et al. proposed an underground container logistics service (UCLS) between Shanghai terminal and northwest logistics park. To ensure the connection between the system and the airport, designing underground parking is proposed. Underground parking is a buffer used for loading and unloading underground cars (UGVs). A nonlinear ordinary planning model (MNIP) has been designed for UGVs and outdoor cranes to minimize the overall cost of the cranes in line with UGVs and terminals. Then, the optimization model has been implemented using MATLAB software (Ornstein et al. 2019).
Oren Steen et al. have designed a parking management system that comprises a central dataset in communication with the server, at least one user device, at least one merchant console, and one parking control device in a network. The central database is presented for receiving and storing data from several parking systems. A process has been presented for analyzing the information received by the central database. A dynamic data engine has been presented for analyzing the data received from several parking systems and generating dynamic data. A targeted advertising engine has been presented for analyzing users' data and creating a targeted advertisement. Dynamic pricing information is given to the user's device so that the user can reserve the parking space from one of the parking systems. Targeted promotion is given to the user's device so that the user can choose an advertisement from a merchant (Segwick 2002).
Works carried out in Iran related to smart car parks: a. designing smart parking management and guidance system and its role in securing and increasing road capacity according to paper (Yang and Ma 2002). To obtain information, first they have referred to the reality of the society and by interviews, questionnaires, and case studies addressing the issues and issues present regarding car parks. This paper is conducted using the descriptive method and 70 citizens (drivers). In the following, we address some of these questions and tables.
According to Table 1, the results of this study regarding the satisfaction levels of drivers from parking services around the city were at the low and medium level and only 6 percent have high satisfaction.
Table 2, shows that locating a parking location by asking the people of the region is the highest at 48 percent. Drivers trying to find a parking spot on the street are second with 30 percent and routing systems with 3 percent have had the lowest usage by drivers.
According to the results of Table 3, this question has been raised about how important the factors affecting a parking spot are therefore, the drivers were asked to score the factors mentioned in the questionnaire according to their effect on choosing a parking spot from one to ten. These results show that some parameters affecting parking selection play a more important role while other factors are less important.
And in the final question where drivers were asked about the highest services to the customers, they ranked mentioned factors from one to four and the results are presented in Table 4 which shows the importance of providing better parking services concerning market work with the highest share at 38 percent.
Works carried out abroad related to smart car parks: a. Parking Guidance and Information System (PGIS): As a part of smart transportation systems, it can be helpful in solving traffic issues. The state of traffic in big city areas has become severe due to looking for parking spots. Advanced parking guidance and information system are considered one of the most effective traffic management approaches which can control and affect the usage of cars, especially in crowded and busy parts of the city. Beijing was chosen to host the 2008 Olympics and the government of China has acknowledged that these games would be the most splendid in Olympics history. Before holding these games, it was evaluated in an overall analysis that one of the factors that would affect the Olympic games was the traffic system and Beijing officials decided to provide a comfortable, safe, and accurate traffic system to reduce the adverse effects of this matter on Olympic games. Despite the fact that nowadays many traffic indicators of China like road network congestion, number of licensed cars, and number of daily passengers are at the top of worldwide rankings, infrastructure facilities of Beijing are behind the worldwide trend with regards to traffic management (Yang and Ma 2002).
Traffic properties of Olympics are defined as follows: • Short-term events • High congestion • High demand One of the tools used by Beijing officials in managing Olympics traffic was controlling and scheduling the traffic in Olympics village. Since 1997, the Chinese Academy of Sciences has started two massive projects to research and develop regarding traffic issues. The initial name of this project was ''Urban Traffic Flow Guidance System.'' Jilin University was responsible for carrying out one of these projects. Due to this reason, a research group was formed to design and build a system for the information and management of car parks. This system was named ''Urban Traffic Flow Guidance System'' (Lan and Shih 2014). The scheme for integrating parking guidance with traffic flow guidance in smart transportation systems is shown in Fig. 2.
Comparison of targeted car parking is shown in Fig. 3.

a. Smart Driver Location System for Smart Parking:
Often, finding a parking spot is tiresome for the drivers and the car park itself is expensive in all the major cities of the world. In Saharan et al. (2020), a crowdsourced solution has been proposed that gathers the real information of available car parks using the sensors in smartphones. This system is designed based on cell phones which can follow the driver's route until they want to leave the car park. In this paper, it has been focused on the efficiency and accuracy of using mobile phones for depicting the driver's walking route, which is carried out using the pedestrian dead reckoning (PDR) method installed on the belly and can measure the driver's moving distance with high accuracy. Also, an algorithm synchronous with the map has been designed to measure route errors while the driver is indoors (interior environment). It has carried out this deed by utilizing existing floor maps of buildings. The results have shown that it can guess user's walking distance with an approximate accuracy of 98 percent, which along with location errors is about 0.48 m. In this paper, it has been focused on how to detect the exiting car park activity. This idea is simple because if the phone detects that the driver is approaching where their car is parked, it seems like the driver wants to leave the spot and the parking spot will be available soon.
In Saharan et al. (2020), a driver, who has parked recently, can provide a message regarding when he or she wants to leave the spot and this information might be sold to another driver who wants to pay (using virtual money, like Bitcoin). Once the buyer reaches the parking spot when it is close to the departure time of the seller, he can occupy the spot after the seller leaves. Therefore, drivers only exchange information regarding available car parks. This action is carried out automatically in this paper, i.e., the act of registering and removing parking spots is done automatically. The main focus of this paper is on how the walking path of the driver, which is the key to this method, can be efficiently and accurately depicted. According to this paper (Saharan et al. 2020) and all their previous research, previous systems all worked manually and could not automatically carry out the park registration and removal operation the users needed to apply this operation manually. In order for his system to carry out this operation automatically, it needs to monitor the user's behavior.
In Khalkhali and Hosseinian (2020) the authors proposed a method that uses a machine learning method to predict parking occupancy, which in turn is used to deduce occupied driving prices for the entry of cars. Parking data on Seattle City Street has been used to train, test and compare different models of machine learning. This is the first time a parking occupancy forecasting system has been used to generate an occupancy-based parking price for a Seattle Street parking system.
In Buldakov et al. (2020), the authors proposed a multifunctional gray wolf optimization technique with the aim of minimizing the localization error. For telemetry and geometric constraints, two objective functions are considered. In our research for the optimal localization of wireless sensor nodes with IoT capability to determine their positions in smart parking with the aim of developing a model based on the optimization of multifunctional gray wolf. Objective functions include distance and topological constraints. Using the multi-purpose gray wolf optimization (MOGWO), the Pareto optimal solution is obtained to determine the optimal solution. The goal of localization is  An intelligent parking management system using RFID technology based on user preferences 13873 to achieve efficiency and reduce the number of anchor nodes. In Najmi et al. (2021) a parking management system is proposed. The system is based on Kingsford for the city of Sydney. This paper introduces a factor-based simulation model to analyze the effects of different parking policies in a dynamic realistic framework. In addition, a new behavioral pricing formula has been integrated into the simulation model, which dynamically seeks to maximize the utility of all factors in the system by considering their travel behaviors. This model has been implemented for Kingsford city center to examine the effects of different parking policies and demand scenarios on parking use and system performance.
In Sudhakar et al. (2021), a smart parking system has been proposed with the development of a mobile application that is useful for the user to access detailed information about the parking space and its efficient management in the parking lot. The proposed intelligent parking system in Sudhakar et al. (2021) uses the image processing technique to identify the license plate and also provides automatic door opening and closing operations whenever it detects a vehicle at the parking lot entrance. The mobile application also provides information on parking safety and security features such as fire alarms and gas leaks. Raspberry Pi is used to control and process all system operations. The liquid crystal display (LCD) is located at the entrance to the parking lot to show the current available parking space. Infrared (IR) sensors are used to detect the presence of a vehicle at the entrance to the parking lot. From the image taken, the license plate characters can be identified and then the Raspberry Pi sends a signal to the servo motor to open the gate for a specified period of time. The user then parks the vehicle at the existing parking lot. When the user wants to leave the space and move the vehicle, date and time information is recorded, which is mostly used for bill processing.
In Jabbar et al. (2021), an IoT Raspberry Pi-based parking management system (IoT-PiPMS) has been developed to help staff/students easily access available parking spaces with simultaneous visibility and GPS coordinates, all using a telephone app. Smart, find. The system introduced in Jabbar et al. (2021) consists of an embedded Raspberry Pi 4 B ? (RPi) computer, a Pi camera module, a GPS sensor, and ultrasonic sensors. To collect and process input data in IoT-PiPMS, RPi 4 B ? uses sensors / cameras, and data is uploaded via Wi-Fi to the Blynk IoT server. Ultrasonic sensors and LEDs with Pi camera support are used to detect the occupancy of parking spaces to ensure the accuracy of the data.

The proposed MODM-RPCP solution
As the number of cars increases, finding a parking space becomes challenging. The drivers usually do not know if there is a space to park their car or not. Also, finding a proper space in the large multilevel parking facilities for users, especially those who park for a short time, is complex and a waste of time. Therefore, the parking issue has become one of the main issues in urban transportation management, because the urban spatial resources are Fig. 3 Comparison of targeted car parking duration before and after utilizing the PGIS system in one of Beijing's regions (Lan and Shih 2014) limited and parking cost is high. Due to limited parking space, many cars spend much time in the streets to find a proper parking slot, and wait in long queues to park or retrieve their car. Smart multilevel parking facilities are constructed at the center of most large cities to handle these limitations and develop an intelligent approach to inform the drivers to select parking facilities. This study presents a smart multilevel parking management system using RFID and checking users' priorities, which can efficiently solve parking issues. The proposed method called MODM-RPCP is a multi-objective decision-making method that solves the parking problem.

The general architecture of the MODM-RPCP
A short-range wireless communication technology capable of detecting radio frequency is called RFID technology, which can read or write related data via radio signals without any mechanical or optical communication. It also can identify specific targets. The proposed MODM-RPCP method introduces a multilevel parking space management method based on priorities, including parking management, and reducing search time. The management system of the parking space information also implements the parking guide management and parking costs in each parking. This system manages parking space using RFID tags and parking search time reduction criteria called CRPST.
The proposed MODM-RPCP method comprises the following sections: decision-making, input control, and output control. The multilevel parking space management structure is shown in Fig. 4.

Phase 1: decision-making
In this phase, a model is introduced for optimal allocation of the cars to the spaces existing in the intelligent multilevel parking aiming to reduce the waiting time, cost, and energy loss. Some symbols are used in this section, described in the following.
These symbols include the input cars, parking floors, parking locations on each floor, as shown in Table 5.

Introducing the parameters
The parameters of the proposed method include time interval, duration (how long a car parks), and time to space index, as given in Table 6.
Decision-making, including definitions like allocating a vehicle to a parking floor, and selecting a parking space in the corresponding floor is introduced as in Table 7.

Decision-making model for allocating a vehicle to the parking
The decision-making model for allocating a vehicle to parking spaces is a nonlinear objective (cost) function, including integer variables aiming to reduce the total time of allocating the cars to the specified parking slot. This function comprises elements like the total number of assigned cars, and the time interval of the assigned parking slot from the entrance/exit.
In the proposed model, constraints (2) and (3) ensure that a vehicle assigned to floor P F and location F L has a smaller CRPST than the cars assigned to floor P 0 F and F 0 L . Because floor P F and location F L are closer to the entrance/ exit than P 0 F and F 0 L . Since one vehicle can be assigned to only one location in one parking level, the constraint (4) is used. On the other hand, constraint (5) represents that a parking location F L can be assigned to one parking floor. Finally, the cars on floor P F and location F L should be balanced as represented in constraint (6). The constraints (7) and (8) apply the dual (binary) constraints to the decision variables. As seen in the model, in the first constraint, the vehicle with smaller CRPST is assigned to a lower level. This criterion that is represented by constraint (9) is obtained by dividing the time that the vehicle V parks at the parking by the space required to park the vehicle V. In this equation, ST V is the time that the vehicle V stops at the parking and RS V is the space required to park the vehicle V.
According to Eq. (9), if the duration demand of the vehicle is reduced, its CRPST is decreased, and the vehicle with shorter requested duration at the parking has a higher priority for parking in floors and locations closer to the exit/entrance.

Phase 2: entrance to the parking using MODM-RPCP
In the proposed MODM-RPCP method, the general schematic of the parking is shown in Fig. 5. The hardware employed in the input control includes antenna, tag reader, automatic card issuance, automatic rail, decision-making terminal, and positioning, which is aware of the vacant space of each parking facility.
The RFID tag reader includes an antenna, tags, and readers. The reader device reads the label information, which contains a unique number, when the label enters the cover of the electromagnetic wave emitted by the antenna. Then, using the decision-making terminal and based on the previous discussion, the location and floors of the parking, which is a serial number similar to the tag is assigned to the input vehicle. First, the input cars receive an entrance card via the automatic card issuance device, if all parking slots are full, the device does not issue any card and announces that the parking is full; if there are vacant spaces in the parking, the vehicle enters. Each specific vehicle receives an RFID tag from the card issuance device, which includes information of the vehicle, personal information of the   driver, including account number and duration. Then, the parking information stored in the database is examined, and a vacant space in correspondence with the individual's request is assigned to the vehicle. A vehicle is controlled via RFID when it enters the coverage area of the EM waves emitted by the antenna. If the tag is valid, and a space is assigned to the vehicle, the input rail goes up, and the vehicle enters the parking without stops to reach the assigned location; if the tag is invalid, the rail does not go up.
As mentioned in the decision-making section, if the user wants to park for a shorter time, it is assigned to a closer space in the lower levels using the decision-making system. When the vehicle enters, its entrance time is automatically recorded in the database, then the rail goes up automatically to let the vehicle in, and when the vehicle enters the rail goes down.
Since the parking is multilevel and has many spaces, after allocating a space to a user, the user should be directed to find the space faster. In this method, the public guidance used in the conventional parking management cannot be used. Therefore, in the proposed method, a fastpositioning system with a user display system is used. This system helps the user find its parking space fast. This system is comprised of the previous hardware facilities, including control, reader, RFID tags, antenna.
The parking floors are plotted on the planar electronic map; the users can use this map to find the parking space easily. When the user reaches the parking space of interest, enters its parking card received automatically at the entrance to the installed terminal. Then, the user carries the tag with themselves to find the parking space considering the terminal reminder. The positioning system of the multilevel parking is shown in Fig. 6.
As shown in Fig. 6, each parking space on each floor is represented by a unique ID. The bold cells represent the occupied spaces, the empty cells represent the empty spaces, and the parking space of interest is represented in green, which blinks. The graphical electronic map shown in Fig. 6 is relatively simple, and it is only used as an example. In real conditions, the map is 2D and plotted based on the parking space distribution. The parking positioning system and the representation terminal are installed at the entrance of all parking facilities; only some spaces are shown in the figure. This terminal simplifies finding the parking space and making queries. The process performed at the entrance and finding the assigned parking space is shown in Fig. 7. 3.5 Phase 3: fast exit from the parking in MODM-RPCP method As fast entrance to the parking is important and reduces waste of time, fast exit from the parking also plays an essential role in reducing waste of time. The required devices for exit control are as follows: A common computer with the entrance system and control database, card reader and antenna, card restoration device, automatic rail, parking cost display based on duration, and the paid bill printer.
Since the parking space is very large and the parking spaces are similar, the users might forget their parking space. However, they can use the parking card and the electronic map terminal to find the information about the parking space. When a user enters the area covered by the EM waves, the system automatically reads and records the parking card information. After finding the vehicle, when the user leaves the parking, the parking card should be returned to the card restoration device. Then, the system automatically reads the card information, records the duration, and calculates the costs, and the costs are deposited from the users' account using RFID tags when exiting the parking. After paying the bill, the rail goes up automatically, and the user exits. The flowchart of fast exit control is shown in Fig. 8.

Evaluating the performance
This section evaluates the qualitative performance in the form of numerical results to validate the performance of the proposed MODM-RPCP method. To demonstrate a feasibility study, the performance analysis of MODM-RPCP has been divided into two parts: 1. Average booking time, 2. Response time of central parking management server.

Performance metrics
The proposed MODM-RPCP method has been simulated and its performance evaluated in Network Simulator version 2 (NS-3) running on Linux Ubuntu 18.04 LTS. The results were compared with both methods [ODPP (Khalkhali and Hosseinian 2020) and MOGWOLA (Buldakov et al. 2020)].

Simulation results
All three methods are evaluated according to Table 8 under three scenarios. Table 8 displays the significant parameters used in the simulation. In this section, the performance of our proposed approach is evaluated using NS-3 on Linux Ubuntu 18.04 LTS as the simulation tool, and the results are discussed further. Table 9 displays the significant parameters used in the simulation (Ghorpade et al. 2020;Najmi et al. 2021). Table 10 , 11, 12, 13, 14, 15 compares the performance of MODM-RPCP solution vs ODPP and MOGWOLA methods in terms of average booking time, and response time. Figure 9 shows the simulation results of average booking time in three days of the week for the proposed method, ODPP, and MOGWOLA. This metric is shorter for the proposed method than the two other methods in all three days. In the ODPP method, the machine learning-based methods are used to predict the parking occupancy, and the learning is carried out using the collected data, while the An intelligent parking management system using RFID technology based on user preferences 13879 users might have different requests for which the machine is not learned. In MOGWOLA, the gray wolf optimization is used to reduce the computation error, and as a result, the node is localized and positioned faster. Thus, MOGWOLA outperforms ODPP. However, this method also does not consider the user requests and does not discriminate between the users with different durations (how long a user parks). In the proposed MODM-RPCP method, the system makes decisions based on the users' requests and its duration and allocates a proper parking space to the user. As shown in Fig. 9, in all three days, at the first hours of the morning, the average booking time increases because the offices are open and the number of users that need a parking space increases, and it decreases at noon. Then, it increases again in the afternoon as the stores open and the requests increase. However, the critical point is that in all day hours for all three simulated days, the proposed method outperforms the other two methods and has a shorter average booking time. The main reason is that the decisionmaking system decides based on each user's information and its duration. Response time of central parking management server. As shown in Fig. 10, the server's response time in different day hours in three days of the week for MODM-RPCP is shorter than ODPP and MOGWOLA. Because in the proposed MODM-RPCP method, the decision-making system selects a parking for each user based on its requirement and duration, and the users with shorter duration are assigned to spaces close to the entrance/exit. On the other hand, in the proposed MODM-RPCP method, a positioning terminal is used in each parking, that the user can use it with the RFID tag on the card and the electronic map to find its parking slot fast. Also, in the proposed method, all payments are made automatically using the RFID tag which reduces the response time. As seen in Fig. 10, from 9 to noon, the response time is increasing, because the offices are open and the requests for parking facilities is high; thus, the number of requests to servers is high. From 12 to 16, as the offices are closed, the response time decreases, and it increases again in the afternoon as the stores and markets open, and the requests increase. The response time decreases again at the night's end when the requests decrease. The critical point is that in all day hours, the proposed method's response time is shorter than the   two other methods because of employing the decisionmaking system. Therefore, the proposed MODM-RPCP method can allocate the best parking space using the decision-making system and considering the users' priorities. Also, using the RFID technology with the decision-making system provides the possibility of fast entrance/exit to/from the     An intelligent parking management system using RFID technology based on user preferences 13881 multilevel parking without wasting time and reducing energy consumption.

Conclusion
This paper proposes a way to find empty space in the parking lot using RFID technology. In the first phase, the proposed MODM-RPCP method, using the decision-making system, and based on the preferences of users, assigned the best space for parking the car. Then, in the second phase, using RFID technology along with the decision-making system, it provides fast entry and exit from the large multi-storey car park without wasting time and reducing energy consumption. The proposed MODM-RPCP was effective in terms of parking space, according to the proposed optimal solution in all three different scenarios. In addition, the simulation results show that the MODM-RPCP improves the average booking time, and response time of the central parking management server, significantly. The proposed method calculated the optimal solution in less time. It also ensured faster vehicle placement in the empty space and improved network performance.
A: Saturday B: Sunday C: Monday Funding This study has received no funding from any organizations.
Availability of data and material The data that support the findings of this study are available from the corresponding author upon reasonable request.
Code availability All code for data analysis associated with the current submission is available from the corresponding author upon reasonable request.

Declarations
Conflict of interest All of the authors declare that they have no conflict of interest.
Ethical approval This article does not contain any studies with human participants or animals performed by any of the authors.