Enterprise Networking Optimization: A Review of Challenges, Solutions, and Technological Interventions

Abstract

Enterprise networking optimization has become crucial recently due to increasing demand for a secure, adaptable, reliable, and interoperable network infrastructure. Novel techniques to optimize network security and toimprove scalability and efficiency are constantly being developed by network enablers, particularly in more challenging multi-cloud and edge scenarios. This paper, therefore, presents a comprehensive review of the traditional and most recent developments in enterprise networking. We structure the paper with particular emphasis on the adoption of state of-the-art technologies, such as software-defined wide area network(SD-WAN), secure access service edge (SASE) architecture, and network automation, driven by artificial intelligence (AI). The review also identifies various challenges associated with the adoption of the aforementioned technologies. These include operational complexity, cybersecurity threats, and trade-offs between cost-effectiveness and high performance requirements. Furthermore, the paper examines how different organizations are addressing a plethora of challenges by exploiting these technological innovations to drive robust and agile business interconnectivity. The review is concluded with an outline of possible solutions and future prospects, capable of promoting digital transformation and enhancing seamless connectivity within the enterprise networking environment.

1. Introduction

Innovations in cloud computing, edge computing, artificial intelligence, and improved cybersecurity have fundamentally changed the enterprise networking landscape in the modern era [1,2,3]. Today, enterprise networks are the foundation of digital transformation. They provide scalability, speed, and flexibility needed to meet the needs of contemporary business environments. Some of these are increased data throughput, hybrid and remote work environments, and easy access to varied applications in the cloud [4,5,6]. In order to satisfy these requirements, significant advancements in zero trust network access (ZTNA), software-defined networking (SDN), and enhanced network automation have become essential [7,8]. These interventions enable enterprises to lower costs, enhance security features, and optimize organizational activities. An enterprise network represents an organization’s local internet that allows file sharing, system access, and communication between employees and devices [9,10,11]. Through effective communication, users can analyze the performance of the connecting environment that propels business operations. Businesses, organizations, devices, and applications should maintain constant communication to facilitate messaging, calls, emails, and all forms of office operations. Such communication can exist between the staff, spreading across various local and international branches [12,13]. Generally, an enterprise network consists of hardware and software units, including different structures that facilitate end-to-end service delivery [14]. A well-managed network supports thousands of devices and allows many users to work reliably in hundreds of sites worldwide at any time. The structure of these networks relies on intranets and extranets, intrinsically built on both local area network (LAN) and wide area network (WAN) technologies [15]. Security issues on the networks are then addressed through efficient management of internal and external traffic flow patterns [16,17].

Enterprise networks utilize advanced networking and security technologies that allow users and devices to communicate securely and remotely from the office [18]. To perform optimally, there is need to incorporate high-speed and reliable information technology (IT) infrastructure. This helps to reduce communication protocols, alleviate bottlenecks, optimize uptime, and enable smooth operations [19,20]. In addition, hundreds of networking devices, such as switches and routers, are deployed to interconnect servers and end-user workstations. The interconnection can either be through wired (copper), fiber-optic, or wireless technologies [21]. Hence, seamless communication and device interoperability can be facilitated across a wide range of devices, such as smartphones, computers, tablets, and printers, with different operating systems (Android, Apple, Linux, and Windows). To provide guaranteed security, network connectivity in geographically heterogeneous locations is encrypted using virtual private networks (VPNs) [22]. The purpose of VPN encryption in enterprise environments is to restrict connectivity to specific users, devices, and facilities, unlike the internet.

Apart from a connection provided to a user via a single router, an enterprise network hosts a series of connectivity between different devices and the internet. In large networks, an autonomous system number (ASN) may be assigned to provide seamless connection [23]. The earlier method employed in enterprise networking was to connect users and devices to centralized data centers located within the same premises. These data centers served as data repositories and platforms for running applications. In this manner, users and devices were granted access to the head office via LAN [24,25,26]. Thus, the LAN in each office usually formed a connected link to other offices through a large enterprise WAN. This was usually implemented via dedicated multiprotocol label switching (MPLS) links [27,28,29,30], as shown in Figure 1.

However, the current reality of network demands in terms of security has rendered traditional network connectivity unsuitable for the fast-paced needs of modern organizations. Many applications described earlier are currently being combined as a form of cloud migration to address the new security challenges of businesses and organizations [30,31]. Therefore, this paper seeks to provide a comprehensive survey of various state-of-the-art technologies that have contributed to the evolutionary trends in enterprise networking. To the best of our knowledge, there is no existing literature that provides a detailed review of enterprise network evolution in the context of challenges, solutions, and new trends as presented in this paper. Furthermore, it is essential to present an updated overview of enterprise networking, considering its importance in supporting businesses to run highly secure and reliable networking systems.

Our specific contributions in this paper are summarized as follows:

1.

Summary of traditional and state-of-the-art enterprise networking technologies and the impact on organizational transformation.

2.

Discussion of the challenges and opportunities of enterprise networks, with special focus on security challenges and their limitations towards achieving seamless and reliable network connectivity.

3.

Presentation of new trends and potential future research directions in the deployment of enterprise networks for next-generation business solutions.

The rest of the paper is organized as follows. Section 2 introduces the main composition of enterprise networking architecture, which consists of access, distribution, and core layers, in addition to the corresponding sub-structures. Section 3 presents the benefits of an optimized enterprise networking system to herald the significance of functionalities discussed in the subsequent sections. Section 4 discusses various network and security protocols that are aggregated to support proper functioning of enterprise networks. Major state-of-the-art technological innovations in enterprise networking are discussed as new trends and focus for future research directions in Section 5. Finally, the paper is concluded in Section 6. An overview of the organization and structure of the paper is shown in Figure 2.

2. Enterprise Networking Architecture

Within an organization, data, applications and communication protocols interact through the underlying layout of the enterprise network structure and architecture. Enterprise networks were initially planned as centralized, hierarchical, on-premises data centers [32,33]. This model has developed into recent architectures which have since evolved to accommodate distributed and cloud-based systems [34,35,36]. An enterprise network is structured and designed to serve as bedrock for effective resource management, data interchange, and security in a corporate setting. Enterprise networks are organized around essential elements and frameworks that facilitate local and remote connectivity, promoting smooth operations in multi-cloud and hybrid configurations [37,38,39,40]. Nowadays, the multi-tiered architecture is a popular structure in enterprise networking [41]. It divides the network into three distinct layers: access, distribution, and core [42]. This provides improved security and allows free flow of traffic by leveraging the capability of software-defined networking (SDN) technologies. The SDN framework separates the control plane from the data plane, so that the network can be managed efficiently [43,44]. An overview of the common architectural components in enterprise networking is discussed in the following, with Figure 3 showing a typical structure.

2.1. Access Layer

This layer provides connectivity to end users through the routers, switches and wireless access points. It is the closest layer to the end users. The access layer relies entirely on network segmentation and access controls as two security features frequently deployed to protect the network periphery [45]. At this point, the access layer connects to the distribution layer.

2.2. Distribution Layer

The distribution layer is an intermediary layer where data aggregation takes place from access to core layer. Hence, it can also be referred to alternatively as the aggregation layer. Switches and routing features are usually incorporated for policy governance, traffic management, security measure implementation, and quality of service provisioning [45].

2.3. Core Layer

The core layer can be regarded as the network backbone which connects several cascaded layers of the network. Fault isolation is typically carried out in the core layer. It is also noteworthy that the core can function as a rented leased line provided by an internet service provider (ISP) to large organizations in order to reduce costs [46]. Any little downtime can impact the network and therefore result in massive loss of revenue. Thus, the core layer is expected to provide high levels of reliability, redundancy, and reduced latency.

Distribution of the network into these three-layered hierarchical structures greatly helps to manage and organize the network efficiently. While the aforementioned structure is maintained, there exists further classification of the network into enterprise campus, enterprise edge and Service Provider Edge. The sub-division is altogether termed eEnterprise composite network model (ECNM) [46], as depicted in Figure 4. The ECNM network sub-structure is very important due to its advantage in offering logical, physical, and hierarchical functional design.

2.4. The Enterprise Campus

The enterprise sub-classification is a campus infrastructure that encompasses the network management facility and server farm or server cluster. Both layer 2 and layer 3 switches are used to provide appropriate network connectivity in the building access branch [47]. Thus, the adjoining building distribution compartment, which uses layer 3 devices, is connected via the trunk and virtual LAN (VLAN) links to the building access traffic aggregation. Therefore, implementation of access control, quality of service (QoS), and routing are all executed by this module.

The campus core segment ensures ultra-fast and high speed connectivity between the three blocks [48], i.e., the building distribution section, data center farms, and the enterprise edge. It also reduces unnecessary redundancy in the devices and links. Hence, the goal of achieving adequate fault tolerance and fast convergence will be met by this segment. The function of network management is to ensure enhanced network performance for optimum service. The server farm is responsible for the provision of security, protection, and high speed connectivity for the network.

2.5. The Enterprise Edge

This module provides connection to the service provider edge with the aid of its component, viz., the E-commerce, WAN, VPN and internet [47]. Part of the services provided are security, QoS, and policy enforcement when connections are extended to remote areas.

In addition to balancing traffic, a well-designed network minimizes the extent of failure domains. The region of a network that is affected when a crucial device or service malfunctions is known as the failure domain [46]. The effect of a failure domain, however, depends on the function of the device that breaks first.

2.6. The Service Provider Edge

The services provided by this module are WAN, public switched telephone network (PSTN), and internet [48]. Some of the resources outside the campus are connected through the service provider, which also streamlines communication to ISP and WAN.

3. Benefits of an Optimized Enterprise Network System

Nowadays, enterprises are embracing an all-encompassing, open networking approach that connects applications and IT systems across many network domains [29,49]. This interconnectivity facilitates better performance, efficient operations, enhanced security provisioning, and consistency across the whole enterprise. There are numerous benefits that can be derived when these services are optimally exploited.

3.1. Drive Towards Efficiency and Productivity

The effectiveness and productivity of an organization’s numerous departments are greatly enhanced by a well-designed and reliable IT infrastructure [50]. Meeting project deadlines and the productivity of an employee are directly impacted by the seamless transfer of information, which in addition promotes quicker and more effective communications [51]. Apart from enhancing communication speed, an optimized enterprise network will guarantee improved client interactions through enhanced website functionality, online transactions, and interactive customer experiences.

3.2. Enhanced Security Provisioning

In order to detect security risks and cyber-attacks early, enterprise network optimization is crucial. With a well-developed network, a security vulnerability can be found quickly enough to stop extensive data compromise. Furthermore, in the event of hardware loss or breach of security, network optimization eliminates data manipulation, which promotes improved disaster management and speedy recovery [52,53].

Big businesses are striving for real-time detection of security threats for better monitoring and protection from attacks. An enterprise network serves as a main threat detector by implementing real-time security mechanisms to complement the roles of firewalls, secure internet gateways, and security applications [54]. Some of these functions are accomplished through profiling, access control, network monitoring, verification, and device management and identification.

3.3. Reduced IT Downtime

Incessant IT breakdowns, malfunctions, and downtime can cause businesses to suffer significant financial losses [50]. Organizations lose a lot of money with every outage or failure, and these unpleasant situations will occur more frequently in a company with obsolete and ineffective IT infrastructure. By optimizing the cloud-based IT infrastructure, a company can prevent considerable loss of revenue and in turn realize a huge increase in productivity [55]. Faster speed and greater network availability are outcomes of a strong and optimized network, which reduces downtime and employee frustration, while boosting productivity.

3.4. Cost Minimization

Reduced operational expenses are a direct result of an optimized enterprise network, since it drastically minimizes expenses for operations, services, software, hardware, and maintenance [56]. In addition, implementing cloud-based structures promotes efficient data management, improved corporate procedures, and smooth device interoperability [57]. Large enterprises may find it easier to benefit from network optimization, even though it is crucial for all kinds of businesses. Nonetheless, network optimization is very important for small organizations, as it will help them reduce cost, better manage revenue, and provide high-quality services.

3.5. Effective Cloud Integration

With the increasing development and delivery of data and applications across several public clouds, enterprise networks facilitate efficient communication between cloud-based applications and end users [58]. Enterprises anywhere can benefit from improved connectivity through cloud network integration, which will enable them to efficiently manage their operations and clients [59]. Additionally, it streamlines the real-time access to information and interaction between end users, internal staff, external stakeholders, and the network infrastructure.

4. Enterprise Network Functionality

The primary goal of enterprise networks for many years has always been to interconnect everyone and everything. All connections are routed to the self-hosted, on-premise centralized data centers, where programs are run and data are stored [60]. Users and devices can then be connected to the LAN in different branch offices to enable this access. The LAN in individual branch offices is connected to broader enterprise WAN using a dedicated multiprotocol label switching (MLSP) links [29,30].

Every layer facilitates an organized method of controlling network access and traffic, which eliminates bottlenecks, supports efficient data processing, and promotes effective management of security policies. All these processes are facilitated by different techniques in order to guarantee seamless interconnectivity.

4.1. Data Flow and Communication Protocols

Enterprise networks use transmission control protocol/internet protocol (TCP/IP) to transport data [61,62]. These protocols organize data into packets and route them from their source to destination. Switching and routing devices utilize dedicated protocols such as open shortest path first (OSPF), border gateway protocol (BGP), and ethernet to determine the optimal path for every packet. The aim is to guarantee a dependable data flow throughout the network [63].

4.2. Centralized Management with SDN

In the current enterprise networks, SDN becomes crucial since it separates physical hardware completely from network management [64]. Hence, a centralized control plane that can dynamically regulate the data flows throughout the network is supported by SDN. In addition, SDN enables managers to rapidly alter network resources, set security guidelines, and manage services in response to immediate demands via a centralized console. Thus, a tight integration of AI/ML technique via network orchestration can further strengthen SDN self-adaptation capabilities to satisfy IoT service conditions [65].

4.3. Embedded Security Architecture: Zero Trust and SASE

Nowadays, advanced security features have been integrated into enterprise networks to protect systems and data. Some of the important security mechanisms are zero trust network architecture (ZTNA) and secure access service edge (SASE).

4.3.1. Zero Trust Network Architecture

ZTNA is an enterprise cybersecurity framework that is based on ‘never trust, always verify’ principles. It is designed to reduce internal lateral movement and prevent data breaches. ZTNA protects data and services and minimizes unauthorized access by performing checks on network requests, based on an established access policy [7,8,66]. The type of access requests could range from enterprise applications, infrastructure components, devices, and virtual and cloud components. Hence, the ZTNA concept can broadly be seen as cybersecurity strategy for an organization that incorporates operational policies and network infrastructure (both virtual and physical). Based on the lateral movement of threats and zero trust assumptions, the network security can further be improved by incorporating micro-segmentation policy [67]. As shown in Figure 5, access will only be granted to an enterprise facility through a policy enforcement point (PEP) and a policy decision point (PDP) [8]. Thus, ZTN applies two basic functions of authentication and authorization through policy decision/enhancement point (PDP/PEP) to grant access to any resources.

4.3.2. Secure Access Service Edge

The SASE is an emerging paradigm of cybersecurity, a cloud-based design that moves typical on-premise security and network connectivity to the cloud for proximity to data [5,68]. It ensures secure access for branch offices and remote users alike. Additional security is guaranteed by integrating software-defined WAN (SD-WAN) capabilities with network security features such as ZTNA [69], firewall-as-a-service (FWaaS) [70], secure web gateways (SWG), and cloud access security broker (CASB) [69]. In addition, SASE creates the convergence and integration of security functions and cloud-based connectivity [71]. As illustrated in Figure 6, it is important to note that SASE can be viewed beyond a single solution paradigm. Rather, as a platform for dynamically deploying network, data, application, cloud, and security structures for adaptability. Due to adaptability of SASE architecture, there is no restriction on what can be implemented.

4.4. Cloud and Multi-Cloud Integration

In order to provide access to applications and services, enterprise networks frequently interact with both public and private cloud environments [37,72]. Secure application programming interface (API) gateways, dedicated links, or VPNs are commonly used to accomplish this task. With the help of multi-cloud approaches, or cloud systems, organizations can use services from several cloud providers to create a scalable and robust environment. A typical example is when a user or an organization has data stored on a private cloud, performs document sharing on, say, Google Cloud platform, and executea data analysis on another cloud application.

However, as businesses streamline their operations in multi-cloud environments, the need to ensure adequate security and resource management for optimal performance may arise. Thus, cloud providers can address security challenges by taking into account several factors. They include implementing identity and access management (IAM), controlling threats and vulnerabilities, and conducting comprehensive audit processes. These functions are embedded in federated access control, which hinges on creating relationships of trust between cloud and identity providers [73]. With federated access control, enterprises can be assured of adequate security without compromising network performance.

4.5. Network Monitoring and AI-Driven Analytics

Enterprise networks depend on AI-driven analytics and monitoring technologies to guarantee reliable performance and security [74]. Network managers are constantly alerted of any problems by network performance management (NPM) tools, which constantly track traffic patterns in order to identify abnormalities. Consequently, this process is improved by artificial intelligence (AI) and machine learning (ML) through identification of malicious patterns, activation of self-healing mechanisms [75,76,77], and prediction of bottlenecks within the network [1].

4.6. Network Virtualization and Cloud-Managed Networking

Network virtualization increases flexibility and scalability of enterprise networks. This is achieved through the deployment of several virtual networks on a shared physical infrastructure via VLANs and SD-WAN technologies [78]. Segmented traffic, especially for applications with different data and security needs, can be managed in this manner by various organizations. Furthermore, cloud-managed networking technologies simplify IT operations for distributed networks by allowing remote configuration and management.

Every technology, layer, and security approach in an enterprise network cooperates to guarantee movement of data freely and efficiently between devices and corresponding users [79]. In addition, cloud resources and branch and remote offices are all seamlessly incorporated into the core network to provide reliable, secure, and continuous access from any location. However, this continuous exchange of cloud resources across different platforms may lead to excessive consumption of energy. Thus, a green energy-saving computing solution, similar to the approach developed by [80], will prevent high energy consumption within the cloud-managed environment.

5. New Trends and Future Research Directions in Enterprise Networking

Enterprise networking has advanced in recent years as a result of emerging technologies and evolving business requirements. With the advent of cloud-native applications, overwhelming cybersecurity risks, and increasingly distributed work environments, new trends are committed to improving network agility, security, and management efficiency [81]. In the preceding sections, we have outlined how enterprise networking has evolved over the years as a promising tool that facilitates ease of communication in a business environment.

Also, we highlighted various benefits derived and challenges encountered from adopting these traditional techniques. Nonetheless, there are several other recent development strategies in enterprise networking. Notable amongst these are the adoption of state-of-the-art technologies such as software-defined wide area networks (SD-WAN) [82], network automation [83], AI-driven network management [1,74], hybrid and multi-cloud networking [37,72], and the increasing adoption of secure access service edge (SASE) and zero trust networking [5,7,66,68]. Table 1 summarizes the survey techniques and their classification into different performance-based interventions.

5.1. SD-WAN

SD-WAN has rapidly taken over as the standard WAN technology due to its increased flexibility, enhanced performance, and robust security [82,84,85]. Compared to traditional WANs, SD-WAN provides a layered architecture that is far simpler to run. It also offers a network design that transfers the management and control layers to the cloud via a centralized controller [86]. It is anticipated that there will be further adoptions and migrations from the conventional enterprise WAN technology to SD-WAN. Since SD-WAN does not rely on public lines, there is considerable reduction in cost due to utilization of public internet.

Utilizing SD-WAN optimizes the current IT infrastructure, increasing network capacity, decreasing latency, and improving network scalability. Organizations that have already embraced the technology will also concentrate on leveraging its extensive potentials. That is, by rationalizing their applications through the use of universal customer premises equipment (UCPE) and virtual network functions (VNFs) [85,87]. The network functions are separated into data, control, and application planes by SD-WAN (see Figure 7), in the same way as SDN.

The data plane transports application and user data, while the control plane performs system configuration, monitoring, and making routing decisions [82]. However, the application plane, commonly known as orchestration plane, leverages on the services provided by the control plane. Despite these benefits, SD-WAN technologies can be affected by limitations that impact network security and performance, especially in latency-sensitive applications. Some of them include configuration errors, software vulnerabilities, and interoperability and scalability problems. To address these challenges, ref. [88] proposed SD-WAN-based multi-objective networking improvement measures, such as new transport protocols, NFV, and ML for networking.

Table 1. Classification of survey techniques for enterprise networking optimization.

Category	Survey Techniques	Challenges	Solutions & Performance	Technological Interventions
Security-based Techniques	Intrusion Detection and Prevention (IDPS), VPNs	High false positive rates, low detection latency	AI-driven threat detection [53,54], automated mitigation, Improved WAN optimization [4,37,72]	Deep learning-based intrusion detection [6], early network monitoring tools [74]
	Zero Trust Network Architecture (ZTNA)	Implementation complexity, integration with legacy systems	Micro-segmentation [67], continuous authentication [8]	Cloud-native Zero Trust [81]
	Cybersecurity Mesh Architecture (CSMA)	Data inconsistency, real-time analysis difficulty	Data aggregation and tool integration [89], threat intelligence [90]	Threat-sharing platforms [91], federated learning for cybersecurity [39]
Cloud and SDN-based interventions	Cloud Networking	Security breaches, Cost management	Multi-cloud strategies [92], hybrid cloud security models [38]	Software-defined cloud networking, SASE [4,69]
	Software-Defined Networking (SDN)	Control plane vulnerabilities, compatibility issues	Open-source SDN controllers [93], policy-based automation [83]	AI-driven network orchestration [65], intent-based networking [94]
Cloud and SDN-based interventions	Software-Defined WAN (SD-WAN)	Complexity in deployment, security risks, high power consumption	AI-based traffic routing [95], AI-driven integrated security SASE [72], Energy-saving control [80], Ultra-low latency [88]	Cloud-native SD-WAN [84,85], Hybrid & Multi-cloud networking [37,38], AI-based optimization [75]
AI-Driven and Automation solutions	AI-based network Optimization	High computational costs, trustworthiness of AI decisions	Federated learning [39], cloud-based AI platforms [74]	Predictive analytics [75], real-time network self-healing [1,75]
	Edge Computing, Network automation	Difficult policy translation, scalability concerns	Adaptive control mechanisms [43]	AI-powered policy automation [96,97]
	Self-Healing Networks	Complexity in automation, fault prediction accuracy	Reinforcement learning models [77]	Autonomous networks & Self-configuring AI agents [76]

Table 1. Classification of survey techniques for enterprise networking optimization.

Category	Survey Techniques	Challenges	Solutions & Performance	Technological Interventions
Security-based Techniques	Intrusion Detection and Prevention (IDPS), VPNs	High false positive rates, low detection latency	AI-driven threat detection [53,54], automated mitigation, Improved WAN optimization [4,37,72]	Deep learning-based intrusion detection [6], early network monitoring tools [74]
	Zero Trust Network Architecture (ZTNA)	Implementation complexity, integration with legacy systems	Micro-segmentation [67], continuous authentication [8]	Cloud-native Zero Trust [81]
	Cybersecurity Mesh Architecture (CSMA)	Data inconsistency, real-time analysis difficulty	Data aggregation and tool integration [89], threat intelligence [90]	Threat-sharing platforms [91], federated learning for cybersecurity [39]
Cloud and SDN-based interventions	Cloud Networking	Security breaches, Cost management	Multi-cloud strategies [92], hybrid cloud security models [38]	Software-defined cloud networking, SASE [4,69]
	Software-Defined Networking (SDN)	Control plane vulnerabilities, compatibility issues	Open-source SDN controllers [93], policy-based automation [83]	AI-driven network orchestration [65], intent-based networking [94]
Cloud and SDN-based interventions	Software-Defined WAN (SD-WAN)	Complexity in deployment, security risks, high power consumption	AI-based traffic routing [95], AI-driven integrated security SASE [72], Energy-saving control [80], Ultra-low latency [88]	Cloud-native SD-WAN [84,85], Hybrid & Multi-cloud networking [37,38], AI-based optimization [75]
AI-Driven and Automation solutions	AI-based network Optimization	High computational costs, trustworthiness of AI decisions	Federated learning [39], cloud-based AI platforms [74]	Predictive analytics [75], real-time network self-healing [1,75]
	Edge Computing, Network automation	Difficult policy translation, scalability concerns	Adaptive control mechanisms [43]	AI-powered policy automation [96,97]
	Self-Healing Networks	Complexity in automation, fault prediction accuracy	Reinforcement learning models [77]	Autonomous networks & Self-configuring AI agents [76]

5.2. Network Automation

Since the emergence of SDN and NFV, the traditional static network has witnessed a significant transformation, opening the door for new network technologies. Another remarkable advancement is the automation and effective management of complicated networks through software for achieving optimal performance [83,98]. Large enterprises are using automated networks more frequently in an attempt to increase productivity and reduce operational expenses. Manual network management is becoming even less effective due to increasing global digitization, with challenges ranging from configuration or human error, to variations and IT outages. Thus, organizations will experience overall network efficiency with the use of AI and machine learning for network operations, providing the possibility of central or remote management [96,97].

Due to the manual nature of most network management tasks, performing regular maintenance on them has led to time consumption and wastage of resources. Network automation readily comes to the rescue, as these operations can be carried out more regularly and efficiently, thereby reducing the likelihood of network failure and downtime. Additionally, network automation provides easier ways to maintain network operations by applying configurations (in Figure 8) consistently with less effort across the infrastructure. However, with more connected devices and adoption of software-enabled management systems, the network is continuously exposed to a range of automation security threats. They include phishing attacks, insider threats, ransomware, and malware, among others. By identifying these vulnerabilities, more sophisticated approaches with threat hunting capabilities can be applied to complement other security interventions [99].

5.3. Hybrid and Multi-Cloud Networking

Several enterprise firms are seeking ways to integrate services from multiple cloud providers in order to achieve flexibility, resilience, and performance optimization [37,72]. This is apparent in the shift towards multi-cloud and hybrid networking. Hybrid networking promotes data sovereignty and low-latency access to essential applications by combining cloud services with on-premises infrastructure [57]. Networking solutions are developing as businesses implement multi-cloud strategies to offer smooth connectivity and control across various data centers and cloud providers. This will minimize latency and improve data security. Cloud computing, which integrates traditional technologies, enables computer functionalities to be accessed and managed remotely [92].

Therefore, cloud computing in hybrid, multi, public, or private connectivity modes permit computations and data processing to take place at data centers. Data centers possess significant amounts of data and computational capacity to transmit between different locations [100]. Consequently, clients only pay for storage, data transfer, or processing resources utilized at any instant time. This allows organizations to focus on other business goals without any concern about maintaining the computer hardware and infrastructure. Generally, as illustrated in Figure 9, cloud computing technologies can be deployed in the form of infrastructure-as-a-service (IaaS) [101], software-as-a-service (SaaS) [102], and platform-as-a-service (PaaS) [93] mode. These deployments use the respective infrastructure, software, and platform from the cloud to provide services in the form of rent to end users.

5.4. Network-as-a-Service (NaaS)

With network-as-a-service (NaaS), organizations can scale network services up or down according to their needs, attributed to its flexible, subscription-based networking paradigm [9,103]. Hence, NaaS streamlines networking by eliminating the need for capital expenditures and operational complexity. In addition, it separates infrastructure management, thereby enabling firms to rent network capabilities as required [104]. This pattern is consistent with a larger movement toward “as-a-service” models. For this reason, business firms can use specialized network services without the need for requisite knowledge, while concentrating on their core business operations. By utilizing enterprise 5G and cloud computing as the enabling technology, customized traffic flow can be created by NaaS to significantly improve the QoS requirements of diverse applications [95].

Also, the enterprise may easily incorporate new applications that require wider expansion, allowing for flexible organizational growth (see Figure 10). Hence, NaaS paradigm eliminates the need for the IT department to determine or have information about network activities. Rather, networks and IT share a common knowledge of network services by relying on the same set of open APIs [103]. This gives networking domains more freedom to decide how best to develop, guarantee, and control the visibility of a network function, such as firewall-as-a-service (FaaS). However, this also depends upon which options best meet the needs of clients.

5.5. Edge Computing and 5G Integration

Edge computing is becoming an essential component of enterprise networking due to the explosion of Internet of Things (IoT) devices and the demand for real-time data [91,105]. With the proliferation of data usage across various fields, the entire world is gradually drifting towards digitalization. Edge computing lowers latency and preserves bandwidth by processing data closer to its source. This will ultimately promote optimal performance of sectors such as finance, healthcare, and manufacturing that rely on quick responses.

The adoption of 5G networks further enhances edge computing, since it offers low-latency and high-speed connectivity [106,107]. This helps to facilitate quicker data transfer between devices, enabling decision-making and real-time analytics closer to the data source, as illustrated in Figure 11. While cloud technology proves to be very efficient in data processing, the bandwidth required to send data limits its application. Hence, edge computing readily emerges to remedy the data transfer bottleneck regularly encountered by cloud computing during computation [108,109]. However, it is noteworthy that edge computing also performs the function of a data producer in addition to data consumer role at the cloud.

The edge computing architecture is categorized in

Table 2. Key features and characteristics of edge computing architecture [109].

Edge Computing Architecture Layer
Characteristics	IoT	Edge	Fog	Cloud
Data	Source	Process	Process	Process
Storage Limits	Extremely Limited	Limited	Limited	Unlimited
Deployment	Distributed	Distributed	Distributed	Centralized
Components	Physical devices	Edge Nodes	Fog Nodes	Virtual Cloud res.
Response Time	No response time	The fastest	Fast	Slow
Location awareness	Aware	Aware	Aware	Aware
Nodes Count	The largest	Very large	Large	Small
Computational limits	Limited	Limited	Limited	Unlimited
Data source distance	The source	The nearest	Near	Far

based on the deployment types and roles played for the successful implementation of its objectives. In addition, some software platforms, such as virtualization, data analytics [19], and edge orchestration, help to simplify the complex nature of edge computing.

5.6. Cybersecurity Mesh Architecture (CSMA)

Cybersecurity means protecting the software, hardware, network, and data from unauthorized access and accidental attack or damage. Since its inception, the field of cybersecurity has developed to be dynamic and challenging [89,90]. Currently, CSMA includes a variety of tools and techniques, functioning to protect systems, networks, and data from a broad range of cyber threats. Figure 12 shows CSMA architecture and different aggregated layers. With every new security solution, there are more ways for attackers to penetrate past existing security features. Thus, CSMA proves to be of great advantage to large organizations by relying on data aggregation and tool integration approaches [110].

Each layer of CSMA platform can communicate with the layers above and below it by grouping existing security and analysis devices into stratified layers. As a result, a functional structure is created that enables communication and cooperation amongst highly distributed networks. Security activities are therefore made more efficient by the centralized dashboards and controls that receive all of this information. Additionally, CSMA features AI-supported automation and unified threat intelligence awareness to create a dynamic security system that can circumvent potential attackers [89]. Modern enterprise requires a network to connect its on-premises servers, cloud services, tools, and home and edge devices [111]. Therefore, network security revolves around protecting these geographically distributed resources.

In summary, much has been discussed about advancements in enterprise networking which are shaping the current landscape. Other technologies such as artificial intelligence (AI)-powered IBN architecture [94]; Wi-Fi 6; its successor, Wi-Fi 7; or IEEE 802.11be and beyond 5G (B5G) networks [112,113] are the latest technologies provisioned to deliver seamless wireless access for businesses. They possess the advantage of providing higher bandwidth and supporting more users per access point. These cutting-edge wireless solutions have been optimized to provide ultra-fast speed compared to wired access architecture. With Wi-Fi 7, more connected users and devices are supported without affecting performance, while allowing data to move securely over wireless networks.

6. Conclusions

This paper provided a comprehensive survey of state-of-the art technologies and trends in enterprise networking. The significant attraction gained by enterprise networking as gateway to optimized business solutions has necessitated the urgent need for appraisal of conventional network technologies. Hence, the architectural composition was first considered in order to give an idea of the strategic importance of enterprise networking in ensuring seamless connectivity between users and devices. We then presented various benefits of enterprise networking in terms of profit maximization and enhanced security and productivity. At the same time, we appraised dedicated converged techniques of network and security solutions aimed towards achieving advanced hybrid cloud environments.

Therefore, in achieving its status as a unique contribution to the literature, this paper not only presented a detailed overview of both traditional and contemporary trends in enterprise networking, but also provided specific direction for researchers to identify potential areas of improvement, especially in terms of security and performance.