A Review of Gateway Selection and Gateway Placement in Wireless Mesh Networks

1. Introduction

Wireless mesh networks have garnered significant attention in recent times due to their potential to offer flexible and cost-effective solutions to various communication challenges. The selection of the gateway in networks of this nature is of utmost importance, as it directly influences the overall performance and efficiency of the network [1,2]. To begin, gateway selection is critical to maintaining stable and efficient communication between the various mesh network nodes [3,4]. Nodes in separate subnetworks may communicate efficiently with one another when gateways are strategically placed to allow for direct communication between them. As a consequence, the network's coverage expands, signal interference decreases, and dependability rises [5,6,7].

Second, the efficiency with which a network manages and distributes traffic is affected by the gateway selection methods in use. The entire network performance may be improved by choosing gateways depending on criteria including available bandwidth, traffic load, and network congestion. Using efficient gateway selection algorithms, traffic can be spread evenly over the network, congestion hotspots can be avoided, and resources may be used to their full potential. This, in turn, improves network efficiency by increasing throughput and decreasing latency [8,9].

In order to guarantee the scalability and flexibility of wireless mesh networks, it is essential that the right gateways be chosen. Selecting gateways that can handle more traffic and more nodes is crucial as the network develops and evolves. The network's scalability may be planned and maximized by thinking about things like gateway capacity, processing power, and topology [10,11].

Gateway selection in wireless mesh networks is crucial because it facilitates seamless connection, maximizes network performance, and permits scalability. Data transmission efficiency, increased network coverage, and better overall performance are all attainable in wireless mesh networks via the use of intelligent gateway selection methods and tactics. Gateway selection is an area of ongoing study and development because of the increasing relevance of wireless mesh networks and their growing number of use cases [12,13,14].

2. Basic Concepts

2.1. Definition of Wireless Mesh Networks

A wireless mesh network (Figure 1) is a self-organizing and decentralized communication network made up of individual nodes. In a wireless mesh network, nodes work together to construct the network architecture, as opposed to in a typical network where devices connect to a central access point or router. Data packets are relayed from one network node to another, thereby increasing the range and coverage of the network as a whole [15,16].

Typically, wireless transceivers are installed in each node of a wireless mesh network so that they may exchange data with other nodes in close proximity. These nodes work together to dynamically construct and maintain network connections, guaranteeing that the network will remain up regardless of whether or not specific nodes are functioning properly or are added to or deleted from the network [17,18].

Benefits of wireless mesh networks include scalability, adaptability, and reliability. The network may be simply extended by adding more nodes, since each node can talk to many of its neighbors directly. since of its distributed architecture, mesh networks are less susceptible to disruptions and outages since data may be redirected to working nodes in the network [19,20].

Many different types of settings may benefit from wireless mesh networks, from homes and businesses to "smart cities," factories, and emergency situations. They provide for secure and efficient communication, facilitating the linking and sharing of equipment across great distances and in harsh or ever-changing conditions [21].

The following are some distinguishing characteristics of wireless mesh networks:

Self-Healing: Wireless mesh networks may automatically adjust to new network conditions and restore itself once nodes fail or the topology is altered. Neighboring nodes may dynamically redirect traffic over different channels when a node goes down, keeping the network up with minimal interruptions [22].

Wireless mesh networks are self-organizing in the sense that their nodes work together to build and maintain the network's architecture without any central authority. Rather of manually configuring each node to communicate with its neighbors, self-organization methods allow them to do so automatically [23,24].

Data in a wireless mesh network may be sent from node to node, or "hop," to hop, until it reaches its final destination. Network coverage may be increased and nodes that are not in close proximity to each other can still communicate thanks to multi-hop communication [25,26].

Wireless mesh networks provide ad hoc connectivity because they can be set up rapidly, even in places with minimal or no existing infrastructure. Since nodes may construct a network via ad hoc connections on the fly, this architecture is useful in circumstances when a permanent network infrastructure would be difficult, such as during disaster recovery efforts or during short-term deployments [27,28].

Because of the simplicity with which more nodes may be added to a wireless mesh network, its coverage area and throughput can be quickly expanded. As the network expands, it will be able to manage more traffic and more devices because to the mesh network's decentralized design and its ability to make optimal use of its resources [29,30].

Community networks, smart homes, outdoor wireless networks, wireless sensor networks, and IoT deployments are just a few of the many use cases for wireless mesh networks. In situations where it would be difficult or costly to construct a typical infrastructure-based network, such as in the case of wireless communication, they provide a flexible and adaptive option [31,32].

gateway placement in WMN

In a Wireless Mesh Network (WMN), the placement of gateways plays a crucial role in the network's performance and efficiency. Gateways serve as the bridge between the WMN and other networks, such as the Internet or other external networks [30].

The placement of gateways in a WMN should consider several factors, including [24,25,26,27]:

Coverage: Gateways should be strategically placed to ensure optimal coverage throughout the network. Placing gateways at central locations can help minimize the distance between nodes and improve network connectivity.

Traffic Load: Gateways should be placed considering the expected traffic load in different areas of the WMN. Areas with high user density or heavy data traffic may require additional gateways to handle the load efficiently and avoid congestion.

Redundancy and Fault Tolerance: To ensure network reliability, it is important to have redundant gateways in the WMN. Redundancy helps to avoid single points of failure and ensures that if one gateway fails, others can take over its responsibilities.

Interference and Signal Strength: Gateways should be placed in locations where they can receive strong signals from mesh nodes without significant interference. Factors such as physical obstructions, radio frequency interference, and signal attenuation should be considered to determine optimal gateway placement.

Network Management: The placement of gateways should also consider the ease of network management and maintenance. Placing gateways in accessible locations simplifies tasks such as configuration, monitoring, and troubleshooting.

It's important to note that the optimal gateway placement in a WMN can vary depending on the specific network requirements, topology, and deployment scenario. Conducting a thorough network analysis, including site surveys and simulations, can help determine the most suitable gateway placement strategy for a particular WMN deployment.

2.2. Introduction to Gateways and Their Role in Mesh Networks

In order to connect to and exchange data with other networks, wireless mesh networks need gateways. Data traffic flows via gateways, giving users instantaneous access to the internet and other wired and wireless networks. Gateways provide this interconnection across networks, allowing mesh network nodes to communicate with nodes and services outside of the mesh itself [33,34].

In wireless mesh networks, gateways play a significant role in establishing connections to the internet. They're crucial because they connect the mesh network to the wider web, letting devices on the network use the web, communicate with distant servers, and take use of cloud services. The gateways in a mesh network are the hubs via which all the devices may access the internet and its plethora of resources [35,36].

In wireless mesh networks, gateways also translate addresses, a very important function. They are responsible for translating addresses from the mesh network's internal addressing system to those of external networks [37,38]. Network Address Translation (NAT) is one mechanism used by gateways to facilitate communication between mesh network devices and devices in external networks, which may implement a different addressing scheme. Mesh nodes are able to successfully connect with devices outside the network thanks to this address translation feature [39,40].

When it comes to the safety of wireless mesh networks, gateways play a vital role. They often include firewall features, which provide a defensive wall against unwanted intrusion and assaults. When data enters or leaves a mesh network, it is filtered and controlled by gateways, which also impose security regulations. Gateways improve the security of a mesh network by continually monitoring and regulating network traffic, therefore protecting it from possible attacks [41,42].

When it comes to managing traffic inside a mesh network, gateways are in charge of sending information to its intended recipients. Factors including network status, traffic volume, and QoS needs all go into their routing determinations [43,44]. To alleviate congestion and maximize network performance, gateways use load balancing strategies. Gateways improve the overall performance and dependability of a wireless mesh network by intelligently handling traffic [45,46].

In addition, gateways are crucial in the administration and tracking of networks. They are responsible for gathering data about the network, checking on the status of devices, and allowing for the control and administration of network parameters. Gateways provide network administrators with powerful means of monitoring and controlling the mesh network. They make it easier to do things like diagnose problems, boost performance, and fix issues in the wireless mesh network [47,48].

In the context of wireless mesh networks, gateways assume a crucial function in the establishment of connections and facilitation of exchanges with other networks [49,50]. Mesh network nodes utilize mediators to facilitate communication with the broader World Wide Web or conventional wired networks. The essential characteristics and operations of gateways in mesh networks encompass the subsequent aspects [51,52]:

Gateways serve as the pivotal nodes within a wireless mesh network, enabling seamless communication with other networks. The connectivity of devices within a mesh network to external networks and resources is facilitated through the acquisition of access to services and resources offered by other networks [53,54].

Gateways are responsible for ensuring that the mesh network is able to establish connectivity with the internet. The nodes serve as a central point for the transmission of data between the mesh network and the broader internet. The devices within a mesh network have the capability to utilize internet services, engage in communication with other devices within the network, and access remote servers for the purpose of storing and retrieving data, all facilitated by the network connection [55,56].

Gateways perform address translation tasks, such as Network Address Translation (NAT), to facilitate the connectivity of mesh network devices with devices in external networks that employ diverse addressing systems. As a result, the devices within the mesh network are capable of seamless communication with devices located on the internet or other networks, as evidenced by sources [57,58].

Gateways are crucial in ensuring the security of mesh networks and enforcing security protocols. To govern the transmission of data within and outside the mesh network, several of these networks possess firewall functionalities. The implementation of this measure serves to safeguard the network against potential security breaches such as hacking and malicious attacks, as indicated by the cited sources [59,60].

Data packets inside a mesh network are routed and managed by gateways to guarantee efficient and dependable transmission. Factors including network status, traffic volume, and QoS needs all go into their routing determinations. To avoid congestion and maximize network performance, gateways also oversee traffic distribution and load balancing [61,62].

Gateways also allow for control and monitoring of the mesh network, which brings us to point number six. They are responsible for gathering data about the network, checking on the status of devices, and allowing for the control and administration of network parameters. This paves the way for administrators to keep tabs on the mesh network, fix any problems that arise, and keep it running smoothly [63,64].

Gateways may be used to divide a wireless mesh network into smaller networks, each with its own gateway (see also: segmenting the network, below). Because of this, administrators can more easily monitor, protect, and direct network traffic [65,66].

Bandwidth allocation and management are tasks performed by gateways in a mesh network. To guarantee the best possible performance for mission-critical apps and hardware, they might prioritize traffic, implement Quality of Service (QoS) standards, and set and enforce bandwidth limitations [67,68].

Gateways may also translate protocols between internal mesh network protocols and those of other networks. As a result, a wide variety of network configurations may be supported, and devices that use different protocols can communicate with one another without any hitches [69,70,71].

Gateways also make it possible to connect wireless mesh networks to other types of networks, whether they're wired or wireless. They provide for hybrid network installations and more connection choices by letting multiple network technologies like Wi-Fi, Ethernet, and cellular to coexist [72].

To guarantee high availability and fault tolerance, gateways may be set up with redundancy and failover techniques. In the case of a gateway failure, traffic may be immediately diverted via the remaining gateways in the event of their deployment [73,74].

Gateways govern network traffic and keep the mesh network safe by enforcing network regulations and access control methods. Sensitive information and prevent unwanted access by using authentication, encryption, and authorization procedures [75].

Gateways allow for wireless mesh network monitoring and troubleshooting, which is explained in point 13. They track network activity, gather statistics on network performance, and offer diagnostic tools for locating and fixing network faults [76,77].

Gateways aid with the scalability and adaptability of wireless mesh networks, which is a key feature of the technology. Additional gateways may be added to the network as it expands to handle more traffic and cover more ground. Gateways also provide scalability, making it easier to add new devices to an existing network or implement cutting-edge networking technology [77,78].

In conclusion, gateways link the wireless mesh network to other networks and perform addressing, routing, security, and administration tasks. In a wireless mesh network, they are essential for facilitating conversation, maintaining connection to the internet, and controlling data flow.

2.3. Introduction to Gateway Selection Criteria

Gateway selection in wireless mesh networks involves considering various criteria to determine the most suitable gateways for efficient routing and optimized network performance. The following are common gateway selection criteria [79,80,81,82,83,84]:

• Link Quality: The quality of the link between nodes and potential gateways is a crucial criterion. Factors such as signal strength, signal-to-noise ratio, packet loss rate, and link stability are evaluated. Gateways with stronger and more reliable links are preferred to ensure robust and stable connections.
• Network Metrics: Several network-level metrics are considered during gateway selection:

  a.

  Network Congestion: The level of congestion in the network and at candidate gateways is assessed. Gateways with lower congestion levels are favored to avoid bottlenecks and ensure smooth data transmission.

  b.

  Available Bandwidth: The bandwidth capacity of candidate gateways is taken into account. Gateways with higher available bandwidth are preferred, especially for applications requiring high data rates.

  c.

  Hop Count: The number of hops required to reach a gateway is considered. Gateways with a lower hop count can minimize latency and reduce routing overhead.

• Quality of Service (QoS) Requirements: Specific applications or services may have unique QoS requirements, such as low latency, high throughput, or reliable connections. Gateways that can meet these requirements are prioritized during gateway selection. Differentiated QoS policies and mechanisms can be applied to ensure the desired level of service for different types of traffic.
• Network Topology and Coverage: The network topology and coverage area are crucial factors. Gateways should be strategically placed to ensure adequate coverage and connectivity to all parts of the network. The distribution of gateways should optimize network reachability, minimize transmission distances, and ensure efficient network operation.
• Security Considerations: The security aspects of gateways are taken into account during selection. Gateways should have robust security mechanisms, such as encryption, authentication, and access control, to protect the network from unauthorized access and ensure data confidentiality and integrity.
• Redundancy and Fault Tolerance: Redundant gateways can be deployed to enhance network reliability and fault tolerance. Selection criteria may consider the presence of backup gateways and their ability to seamlessly handle traffic in case of gateway failures. Redundancy helps ensure continuous network operation and reduces the impact of single points of failure.
• Energy Efficiency: The optimization of energy consumption is a crucial factor, particularly in wireless mesh networks that operate under limited resources. Gateways exhibiting lower energy consumption or higher energy efficiency are prioritized to extend the longevity of the network and reduce the energy consumption of individual nodes.
• Scalability and Manageability: Gateways should be scalable and manageable in large-scale mesh networks. The selection criteria may consider factors such as the scalability of gateway management, ease of configuration and maintenance, and compatibility with network management protocols.
• Gateway Capacity: The capacity of the gateway to handle the expected traffic load is an essential criterion. It considers factors such as processing power, memory, and storage capacity of the gateway. Gateways with higher capacity can effectively handle a larger volume of traffic without performance degradation.
• Cost and Deployment Constraints: The cost of deploying and maintaining gateways is a practical consideration. The selection criteria may take into account the cost of the gateway hardware, installation, configuration, and ongoing operational expenses. Additionally, physical constraints, such as the availability of power supply and suitable locations for gateway placement, can also influence gateway selection.
• Traffic Pattern Analysis: Analyzing the traffic patterns and characteristics of the wireless mesh network can be used as a criterion for gateway selection. By considering the traffic flow, volume, and communication patterns, gateways can be strategically selected to optimize routing efficiency and reduce network congestion.
• Application Requirements: Different applications within the wireless mesh network may have specific requirements that need to be considered during gateway selection. For example, real-time applications such as voice or video streaming may require low latency and high bandwidth, while data transfer applications may prioritize reliable connections and efficient throughput.
• Network Stability and Resilience: Gateway selection criteria can include evaluating the stability and resilience of potential gateways. Gateways that have a history of stable performance, minimal downtime, and resilience to network disruptions are preferred to ensure continuous network operation and minimize service interruptions.
• Policy-Based Selection: Gateway selection can be influenced by policy-based rules and preferences. Administrators can define policies based on factors such as cost, performance, security, or specific routing requirements. These policies guide the selection process, allowing gateways to be chosen based on predefined rules.
• Compatibility and Interoperability: The compatibility and interoperability of gateways with existing network infrastructure and protocols are important criteria. Gateways should support the necessary communication protocols, standards, and interfaces to seamlessly integrate with the wireless mesh network and external networks.
• Vendor Reliability and Support: The reliability and support provided by gateway vendors can be considered during selection. Reputation, track record, and vendor support capabilities can play a role in ensuring that the selected gateways are backed by reliable manufacturers and have access to timely technical assistance if needed.

It's worth noting that the significance and weight assigned to each criterion may vary depending on the specific requirements and priorities of the wireless mesh network. Network administrators and researchers can adapt and customize gateway selection criteria based on the unique characteristics and goals of the network deployment.

3. Gateway Selection

3.1. Gateway Selection Algorithms

In wireless mesh networks, a number of different gateway selection techniques may be utilized. These algorithms attempt to rank potential gateways according to a set of characteristics. Some popular algorithms for choosing gateways are listed below [85,86]:

Weighted Sum Algorithm (WSA) : Link quality, network congestion, available bandwidth, and hop count are only few of the variables that the Weighted Sum Algorithm gives importance to. Based on these criteria, the algorithm computes a weighted total for each gateway and selects the gateway with the largest sum as the preferred gateway [86].

Here are a few research papers and articles that discuss the application of the Weighted Sum Algorithm (WSA) for gateway selection in wireless mesh networks [87,88]:

For multi-radio wireless mesh networks, Zhang et al. (2015) offer a method for selecting gateways that is based on the WSA. Link quality, available bandwidth, and the number of hops are only few of the parameters that are taken into account by the algorithm, which is then given a weight. The study simulates the algorithm's operation and compares its results to those of alternative gateway selection strategies. In their 2018 paper, Sharma et al. describe a fuzzy logic WSA-based gateway selection technique. The technique uses fuzzy inference algorithms (Figure 2) to deal with vague or incomplete data while deciding which gateway to use.

The study utilizes simulations to demonstrate the effectiveness of the proposed algorithm and assesses its performance in relation to network throughput and latency. Gupta et al. (2016) have conducted research on the subject of energy-efficient gateway selection in wireless mesh networks, utilizing the Wireless Sensor Actor (WSA) technology. The algorithm considers multiple indicators, of which energy usage is merely one. The study presents a novel energy model and evaluates the efficacy of the algorithm in terms of network durability and power usage. The results demonstrate that the energy-conscious Wireless Sensor Architecture (WSA) effectively prolongs the lifespan of the network. Tuan et al. (2017) incorporated the Wireless Sensor Actuator (WSA) as one of the numerous instances in their examination of gateway selection algorithms for wireless mesh networks. The article provides an in-depth analysis of the advantages, limitations, and adaptability of the WSA approach. The essay presents an introduction to the WSA and draws a comparison with other gateway methods while highlighting its adaptable and multifaceted nature. Below, you can see Table 1, which relates to ‘Gateway in Wireless Mesh Networks (WMNs) and Weighted Sum Algorithm’.

3.2. Multi-Criteria Determination Making (MCDM)

Gateways are assessed and prioritized based on diverse criteria, utilizing multi-criteria decision-making (MCDM) algorithms. The algorithms take into account the relative significance of specific criteria to prioritize the gateways.

In recent years, there has been an increase in the level of attention given to the utilization of fuzzy judgements for the purpose of gateway selection. A prevalent use case involves the identification of the optimal candidate for the position of cluster head (CH) within a network cluster, thereby facilitating the network's sustained operation over an extended duration. The field of decision-making encompasses various schemes, including Multiple Attribute Decision Making (MADM) [15], The present study discusses several clustering algorithms that aim to enhance energy efficiency in distributed systems. Among these algorithms is the Energy-Efficient Distributed Clustering Algorithm based on Fuzzy Scheme (EEDCF), which utilizes the fuzzy Takagi-Sugeno-Kang (TSK) model to select cluster heads. Additionally, the Adaptive Network based on Fuzzy Inference System (ANFIS) is presented, which employs a fuzzy neural network to optimize clustering. Lastly, the Density of Nodes approach is discussed, which utilizes the Mamdani method of fuzzy inference to select a set of candidate nodes.

The utilization of imprecise evaluations is a notable application in the determination of an optimal routing pathway through the implementation of multihop connections, which involves traversing from one node to another. This is exemplified by the relay node selection scheme founded on fuzzy inference algorithms (RNSFIA) [19]. The Relay Node Selection Framework for Industrial Applications (RNSFIA) employs a fuzzy inference methodology to determine the optimal relay node. This decision-making process considers various parameters, including the inter-node distance, residual energy, and communication level. The simultaneous enhancement of both network lifetime and throughput is achieved by RNSFIA in the context of MOD-LEACH [20]. Multi-criteria decision making (MCDM) is a routing technique that utilizes imprecise judgments to achieve optimal results [21]. This approach considers various factors, including hop count, packet transmission frequency, and residual energy, and assigns weights to them.

The utilization of hierarchical procedures constitutes the basis of the second strategy, which involves making comparisons across various criteria. The selection of relay nodes in body area networks (WBANs) involves the use of weights among multiple candidate nodes through the Analytical Hierarchy Process (AHP) [22]. Meanwhile, AHP MCDM [23] utilizes a two-phase clustering approach that entails determining the nodes' location through the sink position and criteria like the number of neighbors, centrality, and residual energy. Additionally, the Analytical Network Process (ANP) [24] based on MCDM selects the optimal CH node based on criteria such as... Some researchers (e.g., [25]) utilize a fuzzy approach to incorporate both Analytic Network Process (ANP) and Analytic Hierarchy Process (AHP) in the selection of cluster network CH. Numerous techniques are currently being scrutinized due to the significant emphasis placed on enhancing energy efficiency within the realm of Underwater Wireless Sensor Networks (UWSN). The study revealed that the utilization of the FAHP MCDM methodology with input parameters such as hop count, distance to the sink, and a number of neighbors yielded superior outcomes compared to the existing cutting-edge techniques. The only difference between FAHP and AHP lies in the approach employed to measure qualitative feedback. The Analytic Hierarchy Process (AHP) method is utilized to determine the relative significance of criteria in qualitative assessments. Conversely, the Multiple Criteria Decision Making (MCDM) approach solely permits the use of imprecise numerical values as input. Consequently, the development of fuzzy AHP (FAHP) was proposed as a substitute for AHP, owing to its inadequacy in managing judgment ambiguity. A decision maker is unable to select any value within the range of 4 to 6, but must instead specifically choose the value of 5. In contrast to AHP methodologies, FAHP has the potential to accommodate imprecision by utilizing fuzzy integers. In the context of wireless mesh networks, gateways are selected for the purpose of facilitating network communication. Li and colleagues (2014) propose a hybrid Multiple Criteria Decision Making (MCDM) approach. The process of prioritizing selection criteria and ranking potential gateways is achieved through the utilization of a combination of two decision-making methodologies, namely the Analytic Hierarchy Process (AHP) and the Approach for Order Preference by Similarity to Ideal Solution (TOPSIS). The research endeavors to replicate the proposed methodology and assess its effectiveness in comparison to prior gateway selection algorithms, with the aim of drawing inferences about its efficacy. The topic of gateway selection in wireless mesh networks has been examined by Kumar et al[13] through the implementation of the TOPSIS and VIKOR (VlseKriterijumska Optimizacija I Kompromisno Resenje) methodologies. The TOPSIS method is commonly acknowledged for its capacity to assess the proximity of a solution to optimality, while the VIKOR method is frequently deliberated for its effectiveness in selecting the most suitable gateway, taking into account the compromise rating. Simulations are employed to evaluate the efficacy of proposed methodologies. The incorporation of Multiple Criteria Decision Making (MCDM) techniques into the framework has included notable approaches such as the Analytic Hierarchy Process (AHP), Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS), and Simple Additive Weighting (SAW). The study employs rigorous simulations to evaluate the performance of the framework and scrutinize its ability to select the optimal gateway based on various factors.

Singh and colleagues [51] performed a thorough examination of multiple MCDM techniques utilized in the selection of wireless mesh network gateways. The present study examines various techniques, including AHP, TOPSIS, and VIKOR, among others, that are commonly employed in the Multiple Criteria Decision Making (MCDM) framework for the purpose of gateway selection. The manuscript assesses the techniques in terms of accuracy, intricacy, and adaptability. The following Table 2 provides details on Gateway in Wireless Mesh Networks (WMNs) and MCDM.

3.3. Load Balancing Algorithms

Thirdly, load balancing algorithms disperse traffic over various gateways to lessen the impact of congestion on the network and make better use of available resources (Figure 3). Considerations including gateway usage, traffic volume, and available computing power are baked into these algorithms. The Round-Robin, Weighted-Round-Robin, and Least-Connection algorithms all fall within this category. Gateway selection in wireless mesh networks relies heavily on load balancing techniques. These algorithms increase network efficiency, resource usage, and dependability by intelligently distributing traffic. Round-robin, weighted round-robin, least connection, and intelligent dynamic load balancing are just a few examples of load balancing algorithms. Each has its advantages and disadvantages, and is best used in certain situations. Load balancing algorithms will be crucial in the development of reliable and effective WMNs as they become more commonplace [63,64,65,66,67].

The goal of load balancing algorithms is to maximize network efficiency by spreading traffic over many gateways. In order to identify the most appropriate gateway, these algorithms take into consideration data on connection quality, node capacity, network congestion, and traffic patterns. There have been several suggested and implemented load balancing techniques for wireless mesh networks [68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94].

This method rotates the incoming connection requests among the available gateways. It prevents any one gateway from being overworked by spreading the load over all of them. However, it does not take into account the capabilities of specific gateways or the state of the network, which might result in subpar performance [95,96].

This method improves the round-robin strategy by giving various gateways varied weights according on their capacity or network quality. In order to maximize the efficiency with which network resources are used, traffic is prioritized toward gateways with greater weights. However, it does not take into account current delays or congestion in the network [97,98].

This method chooses the gateway with the fewest number of open connections. The goal is to spread the traffic load uniformly across all gateways. This method is useful when gateways have widely varying connection counts, but it may overlook other elements that are crucial to performance [99].

This cutting-edge method integrates real-time monitoring of network variables including connection quality, traffic load, and congestion levels to provide intelligent dynamic load balancing. By taking them into account in real time, the gateway selection is optimized for performance and resource usage. Models that can anticipate network circumstances and make informed gateway selection choices may be trained using machine learning methods [100,101].

There are several advantages to using load balancing algorithms for gateway selection in wireless mesh networks. It boosts network efficiency, prevents congestion, increases dependability, and makes the most of available resources. It eliminates bottlenecks and optimizes performance by balancing traffic across many gateways. Also, load balancing algorithms are adaptable and scalable because they can adjust to new or changing network circumstances. Load balancing techniques pose certain difficulties to implement in WMNs. The overhead of a mesh network increases when nodes must communicate and report on network status in real time. Moreover, thought must be given while choosing the indicators and criteria to use in making decisions. In addition, load balancing techniques should be built to withstand disruption from assaults or malicious nodes [102,103].

Table 3 below illustrates information related to Gateways in Wireless Mesh Networks (WMNs) and load balancing algorithms.

3.4. Game Theory

To use game theory to gateway selection, one must first represent the interaction between nodes and gates as a game. By factoring in the strategic interactions between nodes and gateways, algorithms like the Nash Bargaining Solution and the Stackelberg Game may be utilized to make the most optimal gateway selection options. In WMNs, gateways act as the entry points connecting the mesh network to external networks, such as the Internet. The selection of gateways becomes challenging due to factors like varying link quality, traffic patterns, and network congestion. The goal is to choose gateways that maximize network performance while considering factors like load balancing, link quality, and available resources. Game theory algorithms offer a powerful framework to address these challenges [104,105].

Game theory provides mathematical models to study strategic interactions between multiple decision-making entities, called players. By applying game theory algorithms to gateway selection in WMNs, we can consider the interactions between mesh nodes and their decision-making processes. The following game theory algorithms are commonly employed in gateway selection [106,107]:

Non-cooperative Games: In non-cooperative games, each mesh node acts independently to maximize its own utility. This approach models gateway selection as a strategic decision made by each node to optimize its own performance metrics, such as minimizing latency or maximizing throughput. Nodes evaluate available gateways based on local information, including signal strength, congestion levels, and their own resource capacities. Examples of non-cooperative game models include the Nash equilibrium and the Stackelberg game [108,109].

Cooperative Games: Cooperative game theory focuses on the collaboration among mesh nodes to achieve a collective goal. In the context of gateway selection, cooperative game theory can be applied to form coalitions of mesh nodes that collectively optimize network performance. Cooperative game models consider the joint selection of gateways to maximize global metrics such as network throughput or fairness. Solutions like the core, Shapley value, and bargaining solutions can be employed to determine the allocation of gateways among the nodes [110,111].

Applying game theory algorithms in gateway selection for WMNs offers several benefits. Firstly, it provides a framework to model the strategic decision-making of mesh nodes, enabling them to make optimal choices considering their local information. Secondly, game theory algorithms can lead to efficient and fair gateway allocation, which enhances overall network performance and resource utilization. Furthermore, game theory-based approaches can adapt to dynamic network conditions and provide robustness against failures or changing topologies [112].

However, there are challenges in implementing game theory algorithms in WMNs. One challenge is the need for accurate and up-to-date information about network conditions and available resources, which may require additional overhead and communication among mesh nodes. Additionally, the design of appropriate utility functions and the choice of strategic interactions can impact the effectiveness of the algorithm. Furthermore, the scalability and computational complexity of game theory algorithms need to be considered to ensure practical implementation [64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98].

Game theory algorithms offer a powerful framework for gateway selection in wireless mesh networks. By incorporating strategic decision-making and considering the interactions between mesh nodes, these algorithms can optimize network performance, resource utilization, and fairness. The application of game theory algorithms in WMNs enables intelligent gateway selection, adapting to dynamic network conditions and enhancing overall connectivity.

Table 4 provides details on Gateway in Wireless Mesh Networks (WMNs) and game theory.

3.5. Reinforcement Learning Algorithms

Gateway selection choices may be learned and adapted depending on rewards and penalties received from the network using reinforcement learning algorithms like Q-learning and Markov Decision Processes. These algorithms allow gateways to make real-time adjustments to their selection depending on information about the state of the network and the results of previous attempts [99,100,101].

In WMNs, gateways act as the bridge between the mesh network and external networks, such as the internet. The selection of gateways becomes a complex task due to factors such as varying link quality, traffic patterns, and network congestion. The goal is to choose gateways that maximize network performance while considering factors like load balancing, link quality, and available resources. Reinforcement learning algorithms provide a promising solution to address these challenges [102,103].

Reinforcement learning is a branch of machine learning that focuses on training agents to make sequential decisions based on interactions with their environment. In the context of gateway selection in WMNs, reinforcement learning algorithms enable mesh nodes to learn and adapt their gateway selection policies over time. The following reinforcement learning algorithms are commonly applied [104,105,106,107]:

3.6. Q-Learning

Q-Learning is a popular algorithm in reinforcement learning. In gateway selection, each mesh node can be considered as an agent, and the selection of gateways is treated as a sequential decision-making process. Nodes learn from their experiences by maintaining a Q-table, which stores the expected rewards for choosing different gateways in different network states. By exploring and exploiting this Q-table, nodes can make intelligent decisions regarding gateway selection [108,109].

Deep Reinforcement Learning: Deep Reinforcement Learning (DRL) combines reinforcement learning with deep neural networks to handle high-dimensional and complex problems. In gateway selection, DRL algorithms, such as Deep Q-Networks (DQN) or Proximal Policy Optimization (PPO), enable nodes to learn directly from raw input data, such as signal strength, link quality, and traffic patterns. DRL algorithms can capture complex patterns and dependencies in the network environment, leading to more sophisticated gateway selection strategies [110,111].

Applying reinforcement learning algorithms in gateway selection for WMNs offers several benefits. Firstly, these algorithms enable nodes to adapt their gateway selection policies based on real-time network conditions, resulting in improved network performance and resource utilization. Secondly, reinforcement learning algorithms can handle the dynamic and uncertain nature of WMNs, allowing nodes to learn and make decisions in dynamic environments. Furthermore, these algorithms can adapt to changing network topologies and user demands [112,113,114].

However, there are challenges in implementing reinforcement learning algorithms in WMNs. One challenge is the need for significant computational resources, especially for DRL algorithms, which require training and updating deep neural networks. Additionally, the training process of reinforcement learning algorithms may require a considerable amount of data, which may be difficult to collect in real-world scenarios. Furthermore, ensuring the fairness and stability of gateway selection policies across different nodes in the network remains an open research challenge [118,119].

3.7. Genetic Algorithms

Genetic algorithms use evolutionary principles to optimize gateway selection. These algorithms create a population of potential solutions (gateways) and evolve them through successive generations, applying genetic operators such as selection, crossover, and mutation to find the fittest gateways based on fitness functions. Gateways serve as the entry points connecting the mesh network to external networks, such as the Internet. In WMNs, gateway selection becomes challenging due to factors such as varying link quality, traffic patterns, and network congestion. The objective is to select gateways that maximize network performance while considering load balancing, link quality, and available resources. Genetic algorithms provide a robust framework to tackle these challenges [120].

Genetic algorithms are optimization techniques inspired by the process of natural selection and evolution. They involve iteratively evolving a population of candidate solutions through successive generations. In the context of gateway selection in WMNs, genetic algorithms can be applied as follows:

• Representation: The first step in applying genetic algorithms is to define the representation of the gateway selection problem. Each individual in the population represents a potential gateway selection policy. The chromosome of an individual can be encoded as a binary string, with each gene indicating the selection or non-selection of a particular gateway [121,122].
• Fitness Evaluation: The fitness of each individual in the population is evaluated based on predefined metrics, such as network throughput, latency, or load balancing. Fitness evaluation involves simulating the network and measuring the performance of the selected gateways based on the encoded policy [123,124].
• Genetic Operators: Genetic algorithms employ genetic operators, including selection, crossover, and mutation, to evolve the population and generate new generations of individuals. Selection biases the selection of individuals with higher fitness, crossover combines the genetic material of two individuals to produce offspring, and mutation introduces random changes to maintain diversity in the population [124,125].
• Evolution and Convergence: The population evolves through multiple generations, with fitter individuals being more likely to survive and pass on their genetic material. Over time, the population converges towards solutions that exhibit better performance in terms of network metrics [124,125].

The application of genetic algorithms in gateway selection for WMNs offers several benefits. Firstly, genetic algorithms provide a global search capability, exploring a large search space of potential gateway selection policies. This enables the identification of near-optimal or optimal solutions that may not be achievable through traditional algorithms. Secondly, genetic algorithms can adapt to changing network conditions and requirements, ensuring robustness and adaptability in dynamic WMNs [118,119].

However, there are challenges in implementing genetic algorithms in WMNs. The convergence speed of genetic algorithms can be influenced by the size of the search space and the complexity of fitness evaluation. Additionally, the choice of appropriate fitness metrics and the design of suitable genetic operators are crucial for obtaining effective results. Furthermore, the computational overhead of executing multiple simulations for fitness evaluation can be demanding, especially in large-scale networks [116,118].

Genetic algorithms provide a promising approach to gateway selection in wireless mesh networks. By leveraging evolutionary principles, these algorithms can explore and optimize gateway selection policies in dynamic and complex WMNs. The application of genetic algorithms in WMNs enables efficient network performance, resource utilization, and adaptability to changing conditions.

Below, there is Table 5 which illustrates information about Gateway in Wireless Mesh Networks (WMNs) and Genetic algorithms.

3.8. Machine Learning-Based Algorithms

Machine learning algorithms, such as decision trees, support vector machines, or neural networks, can be trained using historical network data to predict the optimal gateway selection decisions [57,58]. These algorithms learn patterns and relationships in the data and make gateway selection decisions based on the learned models. Gateways serve as the interface between the mesh network and external networks, such as the Internet. In WMNs, gateway selection becomes challenging due to factors such as varying link quality, traffic patterns, and network congestion [59,60]. The objective is to select gateways that maximize network performance while considering load balancing, link quality, and available resources. Machine learning techniques provide a promising solution to address these challenges [61,62].

Machine learning encompasses a variety of algorithms and techniques that enable systems to learn from data and make predictions or decisions without explicit programming. In the context of gateway selection in WMNs, the following machine-learning techniques can be applied [63,64,65]:

3.9. Supervised Learning

Supervised learning algorithms learn from labeled training data to predict outcomes or make decisions. In gateway selection, labeled historical data can be used to train models that can predict the best gateway selection based on features such as signal strength, congestion levels, and resource availability. Algorithms like decision trees, random forests, and support vector machines can be employed to learn gateway selection policies [66,67].

Reinforcement Learning: Reinforcement learning algorithms enable agents to learn through trial and error interactions with their environment. In gateway selection, mesh nodes can be considered as agents that learn optimal gateway selection policies based on rewards and penalties received for their actions. Reinforcement learning algorithms, such as Q-Learning or Deep Q-Networks (DQN), can be utilized to learn and optimize gateway selection policies based on real-time network conditions [69,70,71].

Ensemble Learning: Ensemble learning combines multiple machine learning models to improve prediction accuracy and generalization. In gateway selection, an ensemble of models can be trained using different machine learning algorithms or subsets of data. By aggregating the predictions of these models, a more robust and accurate gateway selection policy can be obtained [77,78].

The application of machine learning techniques in gateway selection for WMNs offers several benefits. Firstly, machine learning algorithms can capture complex patterns and dependencies in network data, enabling more accurate and effective gateway selection. Secondly, machine learning techniques can adapt to dynamic network conditions, allowing nodes to learn and update their gateway selection policies in real-time. Furthermore, machine learning algorithms provide the ability to leverage large volumes of data, enabling more informed decision-making [79,80,81].

However, there are challenges in implementing machine learning in WMNs. One challenge is the need for sufficient and representative training data that accurately reflects the network conditions and performance metrics. Collecting and labeling such data can be resource-intensive and may require substantial time and effort. Additionally, the design of appropriate features and the choice of suitable machine learning algorithms play a crucial role in achieving optimal gateway selection. Furthermore, the computational overhead of training and deploying machine learning models should be considered for practical implementation [82,83,84].

Machine learning techniques offer a powerful approach to gateway selection in wireless mesh networks. By leveraging the capabilities of supervised learning, reinforcement learning, and ensemble learning, these techniques enable intelligent and adaptive gateway selection policies. The application of machine learning in WMNs leads to improved network performance, resource utilization, and adaptability to dynamic network conditions [85,86].

The needs, features, and deployment situation of a wireless mesh network will determine which gateway selection method is best. Based on the network's objectives and limits, researchers and administrators may choose or develop the optimal algorithm.

Below, you can see Table 6, which relates to Gateway in Wireless Mesh Networks (WMNs) and Machine Learning.

4. Statistical and Predictive-Based Gateway Selection Algorithms

Statistical and predictive-based gateway selection algorithms are approaches that leverage statistical analysis and predictive modeling to make informed decisions regarding gateway selection in wireless mesh networks (WMNs). These algorithms utilize historical data, network metrics, and statistical techniques to predict the performance of potential gateways and select the most suitable one. In this essay, we will explore the concepts and applications of statistical and predictive-based gateway selection algorithms in WMNs [87,88].

4.1. Statistical Analysis:

Statistical analysis plays a crucial role in understanding the characteristics and behavior of WMNs. By analyzing historical data and network metrics, statistical techniques can provide valuable insights into the performance of gateways. Some statistical-based algorithms for gateway selection include [89,90]:

a) Statistical Thresholding: This approach involves setting performance thresholds based on statistical analysis of network metrics such as latency, throughput, or link quality. Gateways that exceed or fall below these thresholds are selected or rejected accordingly [91,92].

b) Probability-Based Selection: By utilizing statistical distributions and probabilities, gateways can be selected based on their likelihood of achieving desired performance levels. Probability-based algorithms consider the probability density functions of network metrics to determine the optimal gateway [93,94].

4.2. Predictive Modeling

Predictive modeling employs machine learning and data mining techniques to build models that can predict future performance based on historical data. These models can aid in gateway selection by estimating the performance of potential gateways. Some predictive-based algorithms for gateway selection include [95,96]:

a) Regression Analysis: Regression models can be used to establish relationships between network metrics and performance parameters. These models can then predict the performance of gateways based on the observed network metrics. For example, a regression model may predict the throughput of a gateway based on signal strength and traffic load [97,98].

b) Time-Series Forecasting: Time-series forecasting models can predict future performance based on historical data patterns. By analyzing temporal trends in network metrics, these models can estimate the future behavior of gateways. Time-series forecasting algorithms are particularly useful for predicting performance fluctuations and adapting gateway selection accordingly [99,100,101].

c) Machine Learning Algorithms: Machine learning algorithms, such as decision trees, support vector machines, or neural networks, can be trained on historical data to learn patterns and relationships between network metrics and gateway performance. These models can then be used to predict the performance of potential gateways based on real-time network conditions [102,103].

Statistical and predictive-based gateway selection algorithms offer several benefits in WMNs. Firstly, these algorithms provide a data-driven approach to gateway selection, utilizing historical data and statistical techniques to make informed decisions. Secondly, they enable proactive decision-making by predicting future performance based on current and historical network metrics. Furthermore, these algorithms can adapt to changing network conditions and enhance overall network performance [103,104].

However, there are challenges in implementing statistical and predictive-based gateway selection algorithms. One challenge is the availability and quality of historical data, as well as the need for continuous data collection to maintain accurate models [105]. Additionally, the choice of appropriate statistical techniques, regression models, or machine learning algorithms requires careful consideration and evaluation. Furthermore, the computational complexity of predictive models and the training process of machine learning algorithms should be taken into account for practical implementation [106,107,108].

Statistical and predictive-based gateway selection algorithms offer valuable insights and predictions for selecting optimal gateways in wireless mesh networks. By leveraging statistical analysis, regression models, time-series forecasting, and machine learning techniques, these algorithms enable proactive decision-making and enhance network performance. The application of statistical and predictive-based algorithms in WMNs provides intelligent gateway selection strategies based on historical data and predictive modeling, leading to efficient resource utilization and improved network connectivity [109,110].

4.3. Issues and Challenges

When it comes to gateway selection in Wireless Mesh Networks (WMNs), several issues and challenges need to be considered. WMNs are complex network infrastructures that rely on gateways to connect the mesh network with external networks or the internet. Here are some key issues and challenges in gateway selection for WMNs [57,58,59]:

➢

Scalability: One of the primary challenges in gateway selection for WMNs is ensuring scalability. As the network grows and the number of mesh nodes increases, the gateway must have the capacity to handle the growing traffic and provide efficient routing between the mesh network and external networks. Selecting a gateway that can scale with the network's growth is crucial to avoid performance bottlenecks and ensure smooth operations.

➢

Network Performance: Gateway selection plays a vital role in maintaining network performance in WMNs. The gateway should have sufficient processing power, memory, and bandwidth to handle the network's traffic demands. Inadequate gateway performance can lead to delays, packet loss, or degraded network performance, impacting the overall user experience.

➢

Quality of Service (QoS): WMNs often support diverse applications with varying QoS requirements, such as real-time multimedia streaming, voice communication, or data transfer. Gateway selection should consider the ability to prioritize and manage different types of traffic to ensure appropriate QoS levels. The gateway must support QoS mechanisms that prioritize critical traffic and allocate network resources accordingly.

➢

Interoperability: WMNs can comprise nodes from different vendors or operate on different wireless standards. Ensuring interoperability between the gateway and the mesh nodes is essential for seamless communication and network integration. The selected gateway should support the necessary wireless standards and protocols used by the mesh nodes to facilitate interoperability.

➢

Security: Security is a critical concern in gateway selection for WMNs. Gateways serve as the entry points between the mesh network and external networks, making them potential targets for attacks. The selected gateway should have robust security features, such as strong authentication, encryption, intrusion detection, and secure routing protocols, to protect against unauthorized access, data breaches, or network intrusions.

➢

Power Consumption and Energy Efficiency: In WMNs, nodes may be powered by limited energy sources, such as batteries or solar panels. The gateway should be energy-efficient to minimize power consumption and maximize the network's overall lifespan. Selecting a gateway with low power requirements and power-saving features can contribute to the sustainability and longevity of the WMN.

➢

Deployment and Management: Gateway selection should also consider factors related to deployment and management. The gateway should be easy to install, conFigure , and manage, as it plays a central role in network operations. Additionally, remote management capabilities, centralized monitoring, and configuration management tools are beneficial for efficient administration and maintenance of the gateway.

➢

Cost: Cost is always a consideration in gateway selection. The selected gateway should provide a balance between cost and the required functionality. While it is important to invest in a reliable and secure gateway, organizations need to consider their budgetary constraints and evaluate the cost-effectiveness of the chosen solution.

Addressing these issues and challenges in gateway selection for WMNs requires careful evaluation, considering the specific requirements of the network, scalability needs, performance expectations, security measures, and budgetary considerations. Collaborating with experienced vendors and conducting thorough testing and validation can help organizations make informed decisions and select the most suitable gateway for their WMN deployment.

4.4. Security Issues in Gateway Selection

When selecting gateways in Wireless Mesh Networks (WMNs), there are specific security issues that need to be considered. WMNs are wireless networks where multiple nodes communicate with each other to provide extended coverage and connectivity. Here are some security issues related to gateway selection in WMNs [66,67,68,69,70]:

➢

Authentication and Authorization: Gateways in WMNs should enforce strong authentication and authorization mechanisms to ensure that only authorized nodes can connect to the network. Weak or compromised authentication can lead to unauthorized access and potential security breaches.

➢

Secure Key Management: WMNs often use encryption to protect the confidentiality of data transmitted between nodes. Gateways play a crucial role in key management for secure communication. It is important to select gateways that implement robust key management protocols to ensure secure generation, distribution, and revocation of encryption keys.

➢

Secure Routing: Gateways in WMNs are responsible for routing traffic between the mesh network and external networks or the internet. Secure routing protocols should be implemented to protect against attacks such as spoofing, tampering, or hijacking of routing information. Gateways should employ secure routing protocols, such as the Secure Hybrid Wireless Mesh Protocol (SHWMRP), to ensure the integrity and authenticity of routing information.

➢

Denial-of-Service (DoS) Attacks: Gateways in WMNs are potential targets for DoS attacks, which can disrupt network operations by overwhelming the gateway with excessive traffic or exploiting vulnerabilities. Gateways should have mechanisms in place to detect and mitigate DoS attacks, such as rate limiting, traffic shaping, or intrusion prevention systems (IPS).

➢

Physical Security: Gateways in WMNs may be deployed in outdoor or publicly accessible areas, making them vulnerable to physical attacks. Physical security measures, such as tamper-resistant enclosures, video surveillance, or access control mechanisms, should be considered to protect the gateways from unauthorized access or tampering.

➢

Firmware and Software Security: Gateways in WMNs run firmware or software that may have vulnerabilities. It is crucial to select gateways from trusted vendors that regularly release security patches and updates to address any identified vulnerabilities. Additionally, gateways should have secure update mechanisms to ensure that firmware or software updates are obtained from authenticated and trusted sources.

➢

Monitoring and Intrusion Detection: Gateways should have monitoring and intrusion detection capabilities to detect and respond to security incidents promptly. This includes monitoring network traffic, detecting anomalies, and employing intrusion detection systems (IDS) or intrusion prevention systems (IPS) to identify and mitigate potential threats.

➢

Compliance Considerations: Depending on the specific industry or regulatory requirements, gateways in WMNs may need to comply with specific security standards, such as the General Data Protection Regulation (GDPR) or the National Institute of Standards and Technology (NIST) guidelines. It is essential to select gateways that meet the required compliance standards.

➢

Vendor Trustworthiness: The trustworthiness of gateway vendors is crucial in ensuring the security of WMNs. It is important to choose vendors with a strong reputation in security, timely security updates, and a commitment to addressing vulnerabilities promptly.

By addressing these security issues during gateway selection in WMNs, organizations can enhance the overall security of their wireless mesh networks and protect against potential threats and attacks. Implementing a combination of security measures, such as strong authentication, secure routing protocols, encryption, monitoring, and regular updates, can significantly mitigate security risks in WMNs.

5. Conclusion and Recommendations

gateway selection in Wireless Mesh Networks (WMNs) is a critical task that requires careful consideration of various factors to ensure optimal network performance, security, scalability, and resource utilization. Selecting the right gateways plays a pivotal role in establishing seamless connectivity between the mesh network and external networks or the internet.

Throughout this discussion, we have highlighted several key points and challenges in gateway selection in WMNs. These include scalability, network performance, quality of service (QoS), security, interoperability, power consumption, deployment and management, and cost. Each of these factors requires in-depth research and consideration to make informed decisions when selecting gateways.

Furthermore, we have identified several areas for future research in gateway selection. These include dynamic gateway selection, machine learning-based approaches, multi-objective optimization, security-driven strategies, green gateway selection, edge computing integration, IoT integration, network virtualization and SDN, and real-world deployment studies. By exploring these areas, researchers can contribute to advancing the field and addressing the evolving challenges in gateway selection for WMNs.

Ultimately, effective gateway selection in WMNs contributes to the overall performance, reliability, and security of the network. By considering the unique requirements of WMNs and addressing the challenges and future research areas identified, organizations can make informed decisions and deploy gateways that meet their specific needs, enhance network connectivity, and enable efficient communication within the wireless mesh network and beyond.

5.1. Key Points and Areas for Further Research

Gateway selection in Wireless Mesh Networks (WMNs) is an important area of research that continues to evolve with advancements in networking technologies. Here are key points and areas for further research in gateway selection in WMNs:

1. Gateway Placement Algorithms: Research can focus on developing efficient algorithms for gateway placement in WMNs. These algorithms should consider factors such as network topology, traffic patterns, QoS requirements, and energy efficiency to determine optimal gateway locations. The goal is to minimize network congestion, improve performance, and ensure effective coverage.

2. Load Balancing and Traffic Management: Investigate load balancing techniques to distribute traffic across multiple gateways in WMNs. Research can explore dynamic load balancing algorithms that adapt to changing network conditions and traffic patterns. Effective traffic management strategies can optimize resource utilization, enhance network performance, and ensure fair distribution of network resources.

3. Security and Trustworthiness: Further research is needed to enhance the security and trustworthiness aspects of gateway selection in WMNs. This includes developing robust authentication and encryption mechanisms, secure key management protocols, intrusion detection and prevention techniques, and secure routing algorithms. Addressing emerging security threats and vulnerabilities specific to WMNs will be crucial in securing the gateways and the overall network.

4. Integration with 5G and Beyond: Investigate the integration of WMNs with emerging wireless technologies like 5G and beyond. Research can explore the challenges and opportunities in gateway selection when deploying WMNs alongside next-generation wireless networks. This includes exploring the interworking between WMNs and 5G infrastructure, optimizing handover mechanisms, and leveraging network slicing capabilities.

5. Mobility Support: Study the implications of mobility in gateway selection for mobile WMNs. Research can focus on seamless handover mechanisms, efficient routing protocols, and dynamic gateway selection strategies to support the mobility of mesh nodes while maintaining uninterrupted connectivity and QoS.

6. Energy Efficiency and Sustainability: Explore energy-efficient gateway selection techniques in WMNs to prolong the network's lifespan and reduce environmental impact. This includes investigating sleep scheduling algorithms, energy-aware routing protocols, and power management strategies to optimize energy consumption in gateways while meeting network requirements.

7. Machine Learning and Artificial Intelligence: Investigate the application of machine learning and artificial intelligence techniques in gateway selection for WMNs. Research can explore the use of AI algorithms for dynamic gateway selection, predictive traffic analysis, anomaly detection, and automated network optimization to improve network performance and management efficiency.

8. Hybrid WMNs: Research the challenges and opportunities in gateway selection for hybrid WMNs that integrate different wireless technologies, such as Wi-Fi, cellular networks, or satellite communication. Investigate the optimal selection of gateways in these heterogeneous environments to achieve seamless connectivity and efficient network integration.

9. Resource Allocation and QoS Provisioning: Further research can focus on resource allocation strategies for WMNs to ensure effective utilization of network resources and efficient QoS provisioning. This includes investigating dynamic bandwidth allocation, priority-based scheduling, and congestion control mechanisms to optimize network performance and meet application-specific requirements.

By exploring these key points and areas for further research, the field of gateway selection in WMNs can advance, leading to more efficient, secure, and reliable wireless mesh network deployments.

6. Recommendations for Future Research

Here are some specific recommendations for future research in the area of gateway selection in Wireless Mesh Networks (WMNs):

1. Dynamic Gateway Selection: Investigate dynamic gateway selection algorithms that adapt to changing network conditions and traffic patterns. Consider factors such as real-time network performance, load balancing, and energy efficiency to dynamically select the most suitable gateway for optimal network performance.

2. Machine Learning-based Gateway Selection: Investigate creating smart gateway selection models with machine learning strategies like deep learning and reinforcement learning. Based on factors including traffic volume, connection quality, and user preferences, these models may make intelligent gateway selection judgments using collected network data.

3. Multi-objective Optimization: Conduct research on multi-objective optimization algorithms for gateway selection in WMNs. Consider multiple conflicting objectives, such as minimizing latency, maximizing throughput, and reducing energy consumption. Develop optimization models that can find the best compromise solutions to meet diverse network requirements.

4. Security-driven Gateway Selection: Focus on enhancing the security aspects of gateway selection in WMNs. Investigate advanced security mechanisms, such as anomaly detection, intrusion prevention, and threat intelligence, to integrate security considerations into the gateway selection process. Develop secure gateway selection frameworks that can mitigate security risks and ensure the integrity and confidentiality of data.

5. Green Gateway Selection: Address the energy efficiency challenges in gateway selection by developing eco-friendly approaches. Investigate energy-aware routing and gateway selection algorithms that consider the energy consumption of both mesh nodes and gateways. Explore energy harvesting techniques and renewable energy sources for powering gateways in WMNs.

6. Edge Computing-enabled Gateway Selection: Study the integration of edge computing capabilities in gateway selection for WMNs. Investigate the placement of gateway nodes with edge computing resources to enable low-latency data processing, reduce network congestion, and enhance application performance. Develop algorithms that consider both computational and networking aspects in gateway selection.

7. Integration with Internet of Things (IoT): Explore the integration of WMNs with IoT devices and applications. Investigate gateway selection strategies that can handle the unique requirements and massive scale of IoT deployments. Develop efficient gateway selection algorithms that consider IoT device density, data volume, and communication patterns to ensure seamless connectivity and efficient data processing.

8. Network Virtualization and Software-Defined Networking (SDN): Investigate the application of network virtualization and SDN principles in gateway selection for WMNs. Explore the dynamic allocation of virtual gateways and programmable routing mechanisms for improved network flexibility, scalability, and resource management. Develop novel gateway selection frameworks based on virtualized network architectures.

9. Real-world Deployment Studies: Conduct comprehensive empirical studies and field trials to validate the effectiveness and performance of different gateway selection approaches in real-world WMN deployments. Evaluate the scalability, reliability, security, and performance impact of various gateway selection algorithms under different network conditions and deployment scenarios.

By pursuing these recommendations, researchers can contribute to advancing the field of gateway selection in WMNs, leading to more efficient, secure, and resilient wireless mesh network deployments.

Ethical Approval: Not Applicable.

Competing interests: The authors declare no competing interests.

Authors' contributions: Write the paper and prepare tables: M.N; Review: M.N,M.J.

Availability of data and materials: The availability of data and materials is not applicable for this study as there were no data collected or materials used.