A review of SDN‐enabled routing protocols for Named Data Networking

Named Data Networking is a novel network architecture for the future internet in which data is named rather than addressed. This enables routers to forward data more efficiently based on its name, outperforming IP‐based routing. Another networking architecture that divides the control plane and the data plane is software‐defined networking (SDN). In an SDN network, the control plane decides on routing while the data plane forwards data packets. This separation increases routing flexibility and scalability. Numerous techniques to improve routing can be achieved with NDN and SDN integration. This paper provides an in‐depth examination of routing approaches in NDN based on SDN, emphasizing design principles, algorithms, and performance measures. We begin by summarizing the NDN architecture and delving into its essential components. We next go into the core routing ideas in NDN and categorize and study several routing solutions based on Software Defined Networks. Finally, we highlight the need for scalable, effective, and secure routing systems that may satisfy the expanding requirements of the contemporary internet. We also suggest open research topics in NDN routing based on SDN. This review provides an extensive overview of current centralized routing approaches in NDN, including their limitations and future possibilities.


INTRODUCTION
Named Data Networking (NDN) 1 is a future internet architecture that replaces the traditional IP address-based packet delivery with a data-centric model.In NDN, data is named, requested, and routed based on these names.This approach eliminates the need for a specific location or path for data retrieval, making the network more robust, secure, and efficient.Software Defined Networking (SDN) 2 is a networking paradigm that separates control from data.This separation allows network administrators to manage network services by abstracting lower-level functionality.Integrating NDN and SDN can transform the way networks are designed and managed.By combining the programmability and flexibility of SDN with the data-centric approach of NDN, networks can be designed to be more efficient, secure, and easier to manage.This integration can lead to a more scalable network, 3 supporting the increasing demand for data and the growing number of connected devices.
Researchers in several works [4][5][6] have discussed routing in Named Data Networking (NDN).They did not go into enough detail or cover every aspect of NDN routing, particularly SDN-based routing.While 7,8 investigate studies on routing and forwarding in vehicle-named data networks, 9,10 focus specifically on energy consumption, wireless networks, and IoT.
This study investigates NDN routing based on SDN by examining its performance and addressing outstanding concerns.In NDN, SDN-based routing offers several advantages over other routing strategies, such as centralized control, dynamic adaptation, and network optimization.However, it also poses some challenges, such as scalability, security, and compatibility.
The major contributions of this study are as follows: • Provide the necessary background on basic NDN routing concepts, components, and challenges to set the context.
• Comprehensively survey and classify recent SDN-based NDN routing protocols.
• Analyzing the advantages of centralized routing for NDN enabled by SDN controllers.
• Compare different mechanisms and highlight relative strengths and limitations.
• Identifying salient open issues and research challenges, specifically in SDN-powered NDN routing.
Unlike existing studies that take a generalized look at routing in NDN, this study focuses exclusively on SDN-enabled routing protocols.Reviewing the state-of-the-art literature offers useful insights into the opportunities and tradeoffs presented by unifying complementary SDN and ICN technologies to transform NDN routing.This paper also guides future research directions to address open problems in this emerging domain.
The remainder of this essay is organized as follows: Section 2 provides an overview of NDN routing, including its principles, components, and mechanisms.Section 3 discusses centralized routing systems in NDN, their classification, and their advantages.Section 4 presents future suggestions and recommendations for NDN routing based on SDN.Section 5 concludes the paper.

NAMED DATA NETWORK BACKGROUND
Named Data Networking (NDN) revolutionizes network architecture by prioritizing content-centricity over host-centric models, challenging traditional communication norms.In response to the increasing demand for efficient content dissemination, NDN introduces a unique approach.Unlike conventional Internet Protocol (IP) networks that rely on location-based addressing, NDNs focus on the data directly, emphasizing its naming conventions.This transformative shift facilitates a more scalable and efficient content delivery mechanism.Within the framework of NDN, key components such as the Forwarding Information Base (FIB), Pending Interest Table (PIT), and Content Store (CS) play pivotal roles.The FIB directs packets toward their destinations, the PIT manages pending interest requests, and the CS caches frequently accessed content, enhancing retrieval efficiency.The NDN's routing and forwarding mechanisms ensure optimal data flow, 11 prioritizing content retrieval efficiency.Scalability within an NDN is enabled by its hierarchical naming structure and adaptive routing algorithms, which accommodate varying network sizes and complexities.Security in NDN is enhanced by cryptographic mechanisms, which ensure data integrity and confidentiality.This comprehensive exploration of NDN underscores its transformative potential, emphasizing its main components and technical intricacies in routing, scalability, and security protocols.

Routing and forwarding in Named Data Networking
The NDN's transformative design is based on departing from IP-based routing and forwarding.In NDN, routers make forwarding decisions by examining the names of content rather than source and destination addresses.Each piece of content is assigned a unique, hierarchical name, creating a structured and efficient mechanism for content retrieval.When a consumer initiates an interest packet for specific content, it forwards the routers to the content producer based on its name.Upon reaching the producer, data packets are sent back along the reverse path, satisfying the original request.
This content-centric routing not only streamlines the retrieval process but also eliminates the need to maintain extensive routing tables, improving network efficiency.

Security in Named Data Networking
Ensuring robust security measures is paramount in any networking paradigm, and NDN addresses this imperative through its content-centric design. 12In NDN, content is signed by producers, thereby introducing a robust mechanism for ensuring data integrity and authenticity.Consumers, upon receiving content, can verify its source and integrity through cryptographic means, mitigating the risks associated with various security threats.The granular access control features of NDN provide content producers with the ability to specify who can access their data, enhancing the overall security posture of the network.By enabling authentication and authorization at the content level, NDN introduces a nuanced approach to security that aligns with the dynamic nature of contemporary networks.
In conclusion, Named Data Networking emerges as a transformative force in networking, offering distinctive features in routing and forwarding, scalability, and security.The NDN framework's content-centric approach and efficient caching strategies make it a promising solution for addressing the challenges of modern network architectures.In today's world where content-driven communication is becoming increasingly prevalent, NDN is at the forefront of network operation and evolution, presenting a paradigm that redefines the essence of how networks function in the digital era.

Routing scalability in Named Data Networking
NDNs exhibit inherent advantages in terms of routing scalability compared with traditional IP-based networks. 11By implementing effective caching strategies, routers complement NDN's content-centric nature.Frequently requested content is stored in local caches, allowing routers to satisfy subsequent requests for the same content locally.This proactive caching mechanism significantly reduces the load on the upstream routers and contributes to the overall scalability of the network.As the volume of content increases, NDN's ability to efficiently handle diverse content demands without a substantial degradation in performance becomes a notable strength.The scalability of an NDN is intricately tied to its ability to leverage content caching as a strategic mechanism for optimizing resource utilization.

Traditional NDN routing against SDN based routing for NDN
Traditional Named Data Networking (NDN) relies on control planes that are distributed across routers and operate through dynamic routing protocols.This approach enables autonomy but lacks a global view of the network.In contrast, SDN-enabled NDN leverages centralized routing modules, typically present on SDN controllers, to make optimal routing decisions.This is possible because of the holistic network visibility and state provided by the controller, which uses programmable control logic to dynamically manipulate data plane forwarding rules via interfaces such as OpenFlow.This facilitates the creation of flexible routing behaviors customized to NDN traffic patterns and in-network caching mechanics.For example, transport-layer metrics, such as link delays, can be integrated with cache-related metrics, such as data retrieval distances.Such tight coordination of routing and caching is complex with autonomous NDN routers but is more feasible with SDN control.Hybrid SDN deployments maintain separation from the data plane, providing continued operation during controller failures.This contrasts with pure SDN, which renders the data plane non-functional without the controller.However, centralized control also introduces tradeoffs such as limited scalability, single points of failure, and computational bottlenecks.

NDN INTEGRATION WITH SDN
Two new paradigms, Named Data Networking (NDN) and Software-Defined Networking (SDN), aim to overcome the limitations of the current Internet architecture, 13 such as low performance, poor security, and lack of scalability.NDN is a content-centric approach that uses human-readable names to request and deliver data, regardless of where it is stored or hosted.This offers several advantages over the IP-based approach.For example, it reduces network load and improves content availability because of in-network caching, which allows any node to store and serve data.It also enables efficient data transfer, thanks to receiver-driven delivery that only sends data packets in response to interest packets from consumers, thereby avoiding unwanted or redundant data.Moreover, it enhances data security, thanks to built-in security that verifies the integrity, authenticity, and origin of the data by any node in the network, without relying on external mechanisms such as certificates or encryption.
The SDN approach separates the control plane from the data plane in network devices.A centralized controller handles the network intelligence and control functions, whereas devices only forward packets based on the controller's instructions.This offers several benefits over the distributed approach.For instance, it optimizes network performance, efficiency, and stability, thanks to a global network view that allows the controller to monitor the state and status of all devices in the network and use this information to improve the network behavior.It also creates a flexible, scalable, and agile network because of a dynamic configuration that allows the controller to adapt the network behavior in real-time based on changing conditions, requirements, and demands.
By integrating SDN and NDN, routing can be optimized for NDN via SDN to manage NDN's Forwarding Information Base (FIB).The SDN controller calculates optimal FIB entries based on global information and dynamically updates the FIB to adapt to changes and feedback.However, this integration also introduces new security challenges and risks that need to be addressed.For example, an attacker can compromise the SDN controller and manipulate the FIB entries of NDN nodes, disrupting content delivery.This compromises the availability, integrity, and confidentiality of NDN data.Another example is that the FIB entries of NDN nodes can be inconsistent with the SDN controller's view, causing routing errors and inefficiencies.This can degrade the performance and reliability of the NDN data.NDN networks are vulnerable to attacks that exploit the features of the NDN architecture.These attacks may include interest flooding, content poisoning, or naming hijacking and can affect the availability, integrity, and authenticity of data transmitted over the NDN network.It is worth noting that SDN might not be able to detect or prevent such attacks because they are unique to the NDN protocol and its semantics.
Therefore, security is a crucial aspect that needs to be considered when integrating SDN and NDN.The security mechanisms of both paradigms should be aligned and coordinated to ensure the protection of NDN data and SDN control.Moreover, it is imperative to develop new security solutions to address the unique challenges and requirements of the combined architecture.
In summary, both NDN and SDN transform networking by changing its fundamental aspects.NDN moves communication from host-based to content-based, while SDN moves control from distributed to centralized.Combining complementary strengths can lead to an improved Internet architecture that meets future needs and challenges.

ROUTING IN SDN FOR NDN
Researchers are trying to find new ways to solve problems in routing for Named Data Networking (NDN) by combining Software Defined Networks (SDN) and Named Data Networks (NDN). 6,14One solution they are exploring is called centralized routing.This type of routing is where a central location, called the "controller," makes all the decisions about how to send data packets.Each router follows the instructions given by the controller.The controller knows everything about the network and traffic, so it can make intelligent choices about how to route data.Centralized routing can give more control over network resources and security.

CRoS-NDN
Named Data Networking (NDN) is a unique architecture emphasizing content-centricity over host-centricity, aiming to tackle current Internet challenges.Despite its benefits, NDN encounters issues like FIB size explosion, slow convergence, and content mobility.In response, CRoS-NDN, 15 a controller-based routing strategy for NDN, separates control and data planes to enhance routing performance, utilizing SDN's centralized control capabilities.CRoS-NDN treats NDN routers as basic forwarding devices, centralizing routing logic in a controller for improved efficiency and convergence speed.This approach addresses challenges such as FIB size explosion and slow convergence, enabling fast content recovery without relying on prefix aggregation.The scheme supports content mobility and incorporates CDN-like features into NDN, enhancing routing efficiency.
CRoS-NDN operates in two phases: Bootstrap and Named-Data Routing.The Bootstrap phase establishes global network topology knowledge, while the Named-Data Routing phase ensures content localization and access.Procedures include Controller Discovery, Hello, Router Registration, Named-Data Registration, Route Request, and Route Installation.The scheme introduces a tunnel extension for content mobility and multiple controllers, creating virtual links.
Extensive simulations demonstrate the robust behavior of CRoS-NDN, with superior efficiency and delay performance compared to similar protocols.Analytical expressions establish lower-bound efficiency and upper-bound latency.The CRoS-NDN Tunnel Extension exhibits higher data-delivery efficiency than the base scheme.
Deployment plans involve evaluating CRoS-NDN using Afanasyev et al.'s proposal in the Future Internet Testbed with Security (FITS).The use of established software distributions ensures a reliable and standardized deployment process.
While sources lack specific information on the weaknesses of CRoS-NDN, potential challenges include dependency on the controller as a single point of failure, increased communication overhead with network size, and security and privacy concerns in the centralized control plane.
CRoS-NDN's strengths lie in preserving NDN features, supporting content mobility without FIB size explosion, and offering CDN-like features.It demonstrates robust behavior, superior efficiency, and scalability in various scenarios.

SRSC
The SRSC 16 routing method designed for Content-Centric Networks leverages an SDN controller to maintain a global name-to-location mapping directory for efficient content retrieval.As depicted in Figure 2, when source nodes issue Interests for named data objects, the network nodes first check their local caches.
In case of misses, the requests get forwarded to the SDN controller through a control channel.The controller looks up its authoritative content routing table to identify nodes currently storing or hosting the data object.It computes optimal paths based on network topology information.The controller configures forwarding entries proactively at individual network nodes to route future Interests along the shortest path to the target data source or cache.
Evaluations on an emulated ISP topology indicate that SRSC can reduce network-wide message complexity due to minimized forwarding floods.It also improves cache utilization by routing requests directly to caches assured of having the content.However, the requirement for every node to interact with the controller for the first Interest of a named data object severely limits scalability.
With increasing namespace sizes, the control channel overhead can overwhelm the controller's capacity.This also introduces latency overheads as nodes must await controller responses before forwarding virgin Interests.Failures of the controller can paralyze request resolutions network-wide.The controller communication delays and potential downtime make SRSC unsuitable for applications with tight timing bounds.Geographically distributed deployments may also suffer from propagation latencies between distant nodes and the controller.
In summary, SRSC highlights the potential of SDN-based centralized coordination to minimize NDN control traffic and improve cache performance.However, scalability and reliability concerns due to controller overload and latency warrant further investigation.Solutions to partition namespaces or deploy distributed controllers could help overcome these limitations in large-scale deployments.

SDAR
SDAR 18 (Software Defined Intra-Domain Routing) uses the centralized control plane of SDN to enhance the routing performance of Named Data Networking (NDN) and implement adaptive forwarding through intra-domain multi-path routing algorithms.It benefits from the controller's comprehensive view of network topology and conditions.The SDAR controller uses real-time network data to compute optimal single-or multi-path routes for each node, thereby reducing control overhead compared to distributed NDN protocols.Its dynamic multi-path routing adapts quickly to network changes or breakdowns, facilitated by continuous data collection.It introduces an efficient communication model between routers and the controller within a single administrative domain.Thus, SDAR supports the porting of multiple existing single-path and multi-path routing algorithms to the platform, enabling robust and adaptive intra-domain routing.
In addition, concentrating route computation in the controller minimizes the processing load of nodes.This comprehensive central optimization provides performance gains over distributed routing.SDAR efficacy was tested through NS3/ndnSIM simulations under realistic conditions.The results demonstrate its ability to handle various network scenarios with robustness and adaptability.
However, SDAR's total dependence on the controller raises concerns regarding scalability, latency, and reliability.As the network size increases, the controller can become a bottleneck.Latency also increases because of the communication between nodes and the controller.Finally, the controller represents a single point of failure.
In summary, SDAR showcases the powerful capabilities of centralized control based on SDN technology to enhance NDN routing agility, efficiency, and failure resiliency.However, relying on a single controller can limit its feasibility in large-scale networks.Further research is necessary to evaluate the scalability of SDAR in larger networks and to assess its performance under various network scenarios.A hybrid approach that involves distributed forwarding decisions could balance the advantages of centralization with scalability.

POF
Source Routing Over Protocol-Oblivious Forwarding (POF) for Named Data Networking (NDN) 19 is a proposed technique designed to tackle the challenges associated with deployment and routing in NDN.POF allows NDN packets to bypass IP reliance by offering an adaptable and programmable forwarding plane.The source routing schema over POF provides centralized management, facilitating quick responses to network changes and efficient forwarding of interest packets to the optimal cache node.Experiments demonstrate its potential to significantly reduce networking traffic flow and increase the cache hit rate compared with baseline schemas, outperforming alternatives like NLSR and CATT across various metrics.
The technique combines source routing and POF to address deployment and routing challenges in NDN.POF enables NDN packets to be directly carried without relying on IP (Figure 1), defining a generic flow instruction set for an adaptable and programmable forwarding plane.Source routing, a flexible and scalable schema, allows the sender to specify the forwarding path.When applied over POF, this schema demonstrates substantial performance improvements in terms of reduced networking traffic flow, increased cache hit rate, and outperforming alternatives in link traffic, cache rate, access distance, and content retrieval delay.
A source routing schema specifies the NDN packet format, handles failure and recovery, manages network flow, and adjusts traffic flow.
Source routing over POF for NDN exhibits advantages such as reduced networking traffic flow and increased cache hit rate compared to baseline schemas.The schema boasts low control overhead and ease of deployment, leveraging path diversity for improved performance.Experiments confirm its effectiveness, making it a viable deployment and routing plan for NDN.The ability to carry NDN packets without relying on IP and the centralized management of source routing enhance flexibility and efficiency for future networks.
F I G U R E 1 POF architecture. 19verall, the source routing schema over POF for NDN demonstrated feasibility regarding deployment, control overhead, performance improvement, and adaptability to future network requirements.
The source routing schema over POF offers substantial performance improvements, including a more than 50% reduction in networking traffic flow compared with baseline schemas, increased cache hit rate, and better overall network performance than alternatives like NLSR and CATT.The ability to optimize network efficiency and resource utilization is achieved through traffic flow adjustment and path diversity maximization.
Although the source routing schema over POF has several advantages, SDAR's reliance on the controller raises concerns about scalability, latency, and reliability.As the network expands, the controller may become a bottleneck.Furthermore, communication between nodes and the controller contributes to increased latency.Finally, the controller poses a single point of failure risk.

FCR-NS
FCR-NS 20 is a new routing protocol that combines the features of Software Defined Networking (SDN) and Named Data Networking (NDN) to enhance the performance of data forwarding and delivery.
The FCR-NS architecture consists of multiple network zones, each containing a data plane with switches and NDN users, and a control plane managed by an SDN controller.FCR-NS uses two types of packets: data packets for sending data between network elements and interest packets for searching and publishing data in the FCR-NS network.
To manage the routing and forwarding processes, the FCR-NS controller includes four tables.In addition, the proposed caching strategy in FCR-NS calculates the popularity of local data in switches and employs a bloom filter structure to enable fast forwarding.By offering a complete separation between the control and data planes and leveraging the SDN network paradigm, FCR-NS ensures efficient routing and forwarding of data packets and interest packets in the network.
FCR-NS, a proposed solution for SDN controllers, uses bloom filters and a popularity-based cache replacement strategy to enable fast route lookups and packet forwarding while reducing redundancy and improving the cache hit ratio.Moreover, FCR-NS employs a zone-based routing scheme, allowing the controller to dynamically adjust routing paths and policies based on network conditions and user preferences.This centralized routing, coupled with bloom filter forwarding, reduces overhead and latency compared with traditional NDN routing protocols, which rely on distributed algorithms and message exchanges.The performance of FCR-NS was evaluated using the ndnSIM simulator, with experiments demonstrating the efficiency of the proposed solution based on various performance indicators such as cache hit ratio and cache replacement ratio.
The cache replacement policy in FCR-NS has been compared with well-known alternative strategies like LFU, LRU, and Random, demonstrating the effectiveness of the proposed policy.The forwarding procedure in FCR-NS has been compared with strategies like Flooding Strategy, Best-route Strategy, and Random Strategy, showcasing its efficiency.The routing procedure in FCR-NS has been compared with OSPFN and NLSR, highlighting its high speed.
The proposed FCR-NS architecture aims to improve the performance of the current NDN network and accelerate the deployment of the NDN architecture in a real Internet network.However, FCR-NS also introduces some challenges, such as increased memory requirements due to bloom filters, scalability issues due to controller centralization, and computational complexity due to bloom filter operations.
While initial simulations demonstrate the potential of FCR-NS to improve NDN performance, further evaluation is needed on large-scale topologies and real user traffic.Testing can reveal limitations such as scalability, security, controller overhead, and computational complexity.Moreover, testing can also identify possible enhancements to address these limitations, such as distributed controllers, adaptive bloom filters, or hardware acceleration.
In summary, FCR-NS proposes interesting concepts around blooming filtering and centralized zone-based control to improve NDN performance.However, its advantages and applicability in practice require a more thorough investigation.Enhancements to address the controller bottleneck may also be beneficial.Overall, FCR-NS provides a useful foundation for blending SDN with NDN.

RISC
Information-Centric Networking (ICN) is a new approach to content delivery that prioritizes content delivery over traditional host-to-host communication.ICN aims to enhance content distribution efficiency, scalability, and security.
However, ICNs face several challenges, including content discovery, routing, and management.To overcome these challenges, we propose RISC (Routing mechanism for ICN incorporating Software-Defined Networking and Community division), a routing scheme that leverages the concepts of SDN and community division to improve content retrieval and outperform existing methods.RISC 21 includes four key components: community division, centralized management, intra-community routing, and inter-community routing.First, RISC divides the ICN topology into multiple communities based on a maximal tree scheme.This division helps to retrieve content conveniently and effectively, as each community has an information center that stores and manages the content and forwards information between communities.Second, RISC places all information centers under the control of a central controller, which oversees the entire network and coordinates the inter-community communication.This approach decouples the control plane from the data plane and simplifies network management.Third, RISC employs intra-community routing based on the same community information, enabling efficient communication within a community.Each node in a community can query the information center for the content location and forwarding path, and then retrieve the content from the nearest source.Fourth, RISC uses inter-community routing based on relationships between communities, which facilitates communication between different communities within the ICN topology.The central controller maintains a social graph that captures the affinity and trust among communities and uses it to select the best inter-community path for content delivery.
RISC demonstrates higher performance in terms of content retrieval and outperforms existing methods.The proposed community division scheme based on a maximal tree facilitates convenient and effective content retrieval, as it reduces the content search space and avoids unnecessary content duplication.The centralized management of information about content and forwarding enhances the overall performance of RISC, as it simplifies the network configuration and optimization, and avoids inconsistent or outdated information.The routing mechanism of RISC includes both intra-community routing, which is based on the same community information, and inter-community routing, which considers social relationships among communities.These routing mechanisms contribute to efficient communication within and between communities, further improving the performance of RISC.Experimental results validate the effectiveness of RISC, showing its ability to speed up content retrieval and surpass existing methods.
RISC has certain strengths and weaknesses from different aspects.On the positive side, RISC improves content retrieval, centralized management, and efficient routing, as discussed above.On the negative side, RISC has limited security considerations and scalability challenges, as the sources do not provide specific information about these aspects.It is unclear how RISC handles the security threats and attacks in ICN, or how RISC performs in large-scale ICN deployments.Therefore, further research or analysis is required to understand the weaknesses or limitations of RISC in terms of security and scalability.
To determine if it is possible to utilize Reduced Instruction Set Computing (RISC) in real-world scenarios, Further analysis and research are essential.It is crucial to consider factors like compatibility with existing network infrastructure, scalability, and potential challenges.When examining the feasibility of deploying RISC, it is essential to assess its security, performance, and scalability aspects in real-world situations.Furthermore, the practical implementation difficulties, such as integration with existing network protocols and devices, need to be addressed.More work is necessary to assess the feasibility and potential benefits of implementing RISC in real-world environments.

P4 NLSR
The paper aims to design an efficient NDN routing mechanism in a P4 environment 17 to facilitate network integration between NDN and TCP/IP.It addresses the lack of a centralized NDN routing architecture under centralized control, which is essential for the successful implementation of NDN concepts in the network.The proposed mechanism combines the NLSR protocol and the P4 environment to offer extensible NDN routing services.It allows programmable switches to transmit NLSR packets to the control plane with the extended data plane, enabling the control plane application to provide NDN routing services based on resource-location mapping.The main goal of the control plane is to establish resource-location mapping by processing NLSR routing packets (Figure 2).These services are then converted into data plane configuration and installed in the P4 switch.
The proposed mechanism has been evaluated through simulations, comparing it with traditional NLSR.The experimental results show that the proposed mechanism can significantly reduce the number of routing packets, with a slight overhead in the controller compared to NLSR simulation.The specific performance metrics, such as the exact reduction in routing packets or the amount of overhead in the controller, are not explicitly mentioned in the provided sources. 17wever, the paper highlights the effectiveness of the proposed mechanism in reducing routing packets and the minimal impact on the controller.Further details on the performance aspects of the contribution may require additional information or a more detailed analysis of the experimental results.
The paper does not explicitly discuss whether the proposed routing mechanisms would be feasible for use in the real world.However, the authors focus on designing and evaluating an efficient NDN routing mechanism within a P4 environment, which suggests that they believe it is possible to implement such mechanisms.The integration of the NLSR protocol with the P4 environment offers more flexibility and programmability for network research, due to the use of programmable switches.The experimental results demonstrate promising performance improvements, including a significant reduction in the number of routing packets.Even though the paper does not provide specific details on real-world deployment, the proposed mechanism's compatibility with traditional NDN routing protocols and its ability to reduce routing overhead indicate its potential feasibility in practical network scenarios.
The paper discusses a mechanism with strengths and weaknesses.On the strengths side, the mechanism offers extensible NDN routing services and reduces the number of routing packets significantly.The integration of the NLSR protocol and the P4 environment provides flexibility and programmability for network research.Moreover, the experimental results show promising performance improvements with only a slight overhead in the controller compared to the NLSR simulation.However, on the weaknesses side, the paper does not explicitly discuss the security aspects of the proposed mechanism, leaving potential vulnerabilities or security considerations unaddressed.The scalability of the proposed mechanism is not extensively discussed, and specific scalability challenges or limitations are not mentioned in the provided sources.

Empowering Command and Control through a Combination of Information-Centric Networking and Software-Defined Networking
The paper, titled "Empowering Command and Control through a Combination of Information-Centric Networking and Software Defined Networking," 22 explores the challenges faced in current operational scenarios for the deployment of armed forces and the emerging command and control (C2) approaches to address these challenges.It aims to confront the hurdles in C2 operations by amalgamating information-centric networking (ICN) and software-defined networking (SDN), focusing on achieving "C2 agility" within deployed networks.The proposed solution architecture integrates ICN and SDN in a specific application setting, enabling efficient data distribution within "ICN islands" in the military IP network while controlling interaction patterns among heterogeneous nodes in the Internet of Battle Things (IoBT).Initial experiments conducted on the modeled architecture exhibit promising results, although specific performance metrics used in the experiments or evaluation of the proposed architecture are not explicitly mentioned.While the paper does not outline any drawbacks, it is essential to recognize that integrating ICN and SDN may introduce additional complexity, requiring meticulous configuration and management.Furthermore, the performance and scalability of the architecture may depend on specific implementation and deployment scenarios, presenting challenges in real-world operational environments.In conclusion, the proposed architecture holds the potential to meet high-level operational requirements for "C2 agility" within deployed networks, but further research and experimentation are necessary to assess its strengths and weaknesses.

4.9
Secure command and control for the internet of battle things using novel network paradigms In article, 23 a secure network strategy is proposed to achieve Command and Control (C2) agility within the Internet of Battle Things (IoBT) heterogeneous environment.The approach integrates application and network layers with a multi-layer, defense-in-depth cybersecurity mechanism spanning hardware, network, and application levels.
The research investigates the intricacies of establishing secure Command and Control (C2) capabilities in the IoBT environment.It advocates for a multi-layered cybersecurity framework that encompasses hardware, network, and application layers to ensure the secure and timely dissemination of information.
The proposed framework leverages the Software-Defined Networking (SDN) paradigm to orchestrate network services, incorporating semantics-oriented data from technologies like Information Centric Networks (ICN) and Delay-Tolerant Networks (DTN).
Preliminary simulated scenarios showcased a notable enhancement in network metrics and resilience against cyber-attacks, illustrating the potential efficacy of the proposed approach.However, the specific performance metric utilized for evaluating the framework remains unspecified.
Although the paper highlights significant improvements in network resilience and performance, it lacks explicit details regarding the routing techniques employed and the specific performance metrics used for evaluation.This gap hinders a comprehensive understanding of the proposed solution's quantitative impact.
Despite these limitations, the paper presents a promising avenue for achieving secure and agile Command and Control within the IoBT environment.Further exploration into the routing techniques and performance metrics is warranted to ascertain the practical feasibility and effectiveness of the proposed framework.

OPEN CHALLENGES IN CENTRALIZED ROUTING
Centralized SDN-based routing in NDN has its own set of open issues (see Table 1).One of the main issues is the scalability 24 of NDN routing in SDN.The exponential growth of routing tables due to a rise in content objects can result in significant overhead and performance degradation.Furthermore, optimally leveraging the in-network caching capabilities of NDN with SDN's global network view is an open research question.Another issue is NDN uses hierarchical content names for routing, whereas SDN relies on flow-based routing.Integrating the two requires rethinking routing mechanisms.Hence, optimizing an integrated NDN-SDN network is a challenging task.Optimizing data dissemination strategies, caching policies, and routing protocols is necessary to ensure efficient and reliable data delivery.Furthermore, centralizing control can raise concerns about privacy and security, as all traffic and routing decisions can be monitored and controlled from a single location.Therefore, there is a need for further research and development to address these issues and ensure the scalability, reliability, and security of centralized SDN-based routing in NDN.

FUTURE RECOMMENDATIONS
Researchers are constantly looking for ways to improve computer network routing in NDN and to develop new methodologies that allow networks to run better and more efficiently as technology progresses.They employ a variety of • Outperforms other routing protocols in terms of convergence latency and efficiency.
• Scalable to large networks.
• Relies on a centralized controller.
• Can be slow to forward packets.
• Improves cache hit percentage.
• Slow to forward packets.
• Relies on a centralized controller.
SDAR 18 • Centralized controller can quickly adapt to changes in the network.
• Multi-path routing reduces traffic overhead.
• Relies on a centralized controller.
• May not be scalable to large networks.
• Improves cache hit rates.
• Requires POF switches, which may not be practical in all networks.
• Does not compare to other routing protocols.

FCR-NS 20
• Uses bloom filters to improve switch forwarding speed.
• Preserves a clear separation between the control and data planes.
• Integrates NDN with the Internet seamlessly.
• Requires more memory than other systems.
• Needs more testing on larger-scale networks.
RISC 21 • Improves content retrieval efficiency by incorporating SDN and community division.Uses a community partition scheme for easy and effective content retrieval.
• Effectiveness has only been proved in experimental findings.
• May not be scalable to large networks.
• Facilitates NDN integration with TCP/IP-based networks.
• Provides routing and addressing flexibility.
• Lacks a comprehensive comparison to other NDN routing algorithms.

Note:
The table provides a high-level overview of key centralized routing protocols proposed for Named Data Networking (NDN), summarizing their related works, advantages, and disadvantages.This aids in understanding the current state and evolution of routing in NDN.The protocols differ in their architectures and trade-offs but generally aim to improve efficiency, scalability, and integration with the broader Internet.Further research is needed to determine optimal routing strategies as NDN matures.
techniques, including metaheuristic algorithms, artificial intelligence (AI), 25,26 Software-Defined Networking (SDN), and Information-Centric Networking (ICN).Researchers are incorporating AI to construct intelligent algorithms that learn from network data, similar to learning from experience, resulting in networks that gradually become smarter, faster, and more trustworthy through real-time feedback.An exciting concept under exploration is Federated Learning, 27 which involves training machine learning models on multiple devices or networks while preserving privacy, allowing networks to learn from each other without revealing personal information.This collaborative approach can enhance performance and efficiency.The combination of metaheuristic algorithms, AI, SDN, and ICN has enormous potential to change network functions, providing multiple opportunities for quicker, more innovative, and trustworthy networks.Determining the best optimization technique for certain conditions in ICN will necessitate concentrated effort.combining machine learning (ML) with software-defined networks to optimize NDN routing is expected to be a very active research area in the future.
Furthermore, Reinforcement learning (RL), 28 a machine learning (ML) type, allows networks to learn from previous interactions and make informed judgments.RL has been used in many areas, resulting in the creation of clever autonomous systems.Researchers should investigate these concepts to build networks that can constantly adapt, learn, and improve to make the digital experience as smooth and efficient as possible.
Researchers are actively exploring innovative techniques, such as metaheuristic algorithms, artificial intelligence, Software Defined Networking (SDN), and Information-Centric Networking (ICN), to optimize NDN routing.These techniques improve the performance, scalability, and adaptability of NDN routing in different scenarios and environments.Among these techniques, machine learning (ML) methods have shown great promise for NDN routing optimization.ML is a branch of artificial intelligence that enables systems to learn from data and experience without explicit programming.Two relevant ML methods for NDN routing are federated learning and reinforcement learning.Federated learning is a secure and scalable distributed ML technique that enables networks to learn optimal routing strategies from each other's experiences without exposing private or sensitive data.It can also reduce communication overhead and latency compared with centralized ML approaches.Reinforcement learning is an ML technique that allows systems to autonomously learn from dynamic interactions with their environment, thus improving their decisions over time.Reinforcement learning can enable networks to adapt to changing conditions and demands, such as traffic patterns, congestion, failures, and user preferences.Reinforcement learning can also allow networks to discover new and better routing paths that may not be obvious or known beforehand.Implementing federated and reinforcement learning for NDN routing requires a careful system design tailored to the caching and multiparty communication architecture of NDN.Potential approaches include distributed learning across NDN nodes with coordination from controllers and intelligent caching agents that learn to route and caching policies.These approaches must consider various factors such as network topology, data popularity, user feedback, security threats, and resource constraints.These ML techniques can bring significant advantages for NDN routing optimization, such as responsive and intelligent routing, privacy-preserving collaboration, and self-optimization.As learning improves routing policies, NDN performance can be enhanced while reducing human intervention and overhead.However, challenges remain in terms of security, interpretability, and sample efficiency.Before NDN routing can widely adopt federated and reinforcement learning, there are a few challenges to overcome.In addition to ML, another technique that can optimize NDN routing is the combination of optimization algorithms, such as ant colony optimization and particle swarm optimization, with SDN's centralized control.Metaheuristic techniques can optimize NDN caching, load balancing, and forwarding by adjusting network parameters with SDN controllers.These metaheuristic techniques can improve NDN routing performance by finding near-optimal solutions in a reasonable time.They can also adapt to network changes by iteratively updating their solutions.However, these techniques have drawbacks such as high computational complexity, difficulty in parameter tuning, and convergence uncertainty.These need to be overcome before integration with SDN for NDN routing.The convergence of modern AI, SDN programmability, and ICN data-centricity can transform NDN routing.Intelligent and autonomous NDN networks can optimize performance, efficiency, and resilience based on data insights.However, careful research is imperative to use these technologies effectively.

CONCLUSION
This survey provides an overview of the recent research applying software-defined networking (SDN) to improve routing in Named Data Networking (NDN).NDN is a promising Internet design that focuses on content and not welcomers.Efficient routing and forwarding remain open research problems in NDN.The separation of the control and data planes in SDN provides opportunities to address these challenges.
Several studies have proposed SDN-based NDN routing frameworks to enable flexible, programmable routing control.Among the key benefits are support for new routing strategies, optimization of cache use, and reduction of overhead through a global view of the network and centralized routing decisions.However, SDN-based NDN routing introduces new challenges, including scalability, controller placement, and consistency across controllers.
Future research should address these challenges.Approaches such as distributed controller architectures, hierarchical control, and machine learning techniques require further exploration to scale SDN-NDN networks.Studies should also focus on efficient ways to collect and share global network states while maintaining cache consistency.As research matures, integrating SDN with NDN can enable more efficient, flexible, and scalable content-centric networks.