MANET Routing Protocols’ Performance Assessment Under Dynamic Network Conditions

Abstract

Mobile Ad Hoc Networks (MANETs) are decentralized wireless networks characterized by dynamic topologies and the absence of fixed infrastructure. These unique features make MANETs critical for applications such as disaster recovery, military operations, and IoT systems. However, they also pose significant challenges for efficient and effective routing. This study evaluates the performance of eight MANET routing protocols: Optimized Link State Routing (OLSR), Destination-Sequenced Distance Vector (DSDV), Ad Hoc On-Demand Distance Vector (AODV), Dynamic Source Routing (DSR), Ad Hoc On-Demand Multipath Distance Vector (AOMDV), Temporally Ordered Routing Algorithm (TORA), Zone Routing Protocol (ZRP), and Geographic Routing Protocol (GRP). Using a custom simulation environment in OMNeT++ 6.0.1 with INET-4.5.0, the protocols were tested under four scenarios with varying node densities (20, 80, 200, and 500 nodes). The simulations utilized the Random Waypoint Mobility model to mimic dynamic node movement and evaluated key performance metrics, including network load, throughput, delay, energy consumption, jitter, packet loss rate, and packet delivery ratio. The results reveal that proactive protocols like OLSR are ideal for stable, low-density environments, while reactive protocols such as AOMDV and TORA excel in dynamic, high-mobility scenarios. Hybrid protocols, particularly GRP, demonstrate a balanced approach; achieving superior overall performance with up to 30% lower energy consumption and higher packet delivery ratios compared to reactive protocols. These findings provide practical insights into the optimal selection and deployment of MANET routing protocols for diverse applications, emphasizing the potential of hybrid protocols for modern networks like IoT and emergency response systems.

1. Introduction

Mobile Ad Hoc Networks (MANETs) are wireless networks that operate without fixed infrastructure such as routers or servers. Each device (or node) in the network can act as both a sender and receiver. This allows communication between nodes even if they are not directly connected. MANETs are especially advantageous in scenarios where conventional networks are impractical, including natural disasters, military activities, and isolated or rural locations [1,2].

1.1. Challenges in MANETs

Despite their advantages, MANETs face several challenges. The dynamic nature of the network, where nodes move freely, causes frequent changes in network structure, making it difficult to maintain stable communication. Other issues include limited battery life in devices, which restricts the network’s operating time, and constrained bandwidth, which affects data transmission speeds. Furthermore, designing efficient routing protocols that address these challenges remains a significant problem [3,4].

To address these challenges, this study evaluates the performance of eight MANET routing protocols under diverse network conditions.

This study investigates the performance of eight MANET routing protocols: OLSR, DSDV, AODV, DSR, AOMDV, TORA, ZRP, and GRP. A customized simulation environment in OMNeT++ 6.0.1 with INET-4.5.0 was used to assess these protocols. The simulations assess their performance under varying network conditions, including node densities of 20, 80, 200, and 500. Key performance metrics examined include network load, throughput, delay, energy consumption, jitter, packet loss rate, and packet delivery ratio [5,6].

1.2. Main Contributions

• This study evaluates metrics often overlooked in prior research, such as energy consumption and jitter.
• It provides an in-depth analysis of hybrid protocols, demonstrating their adaptability and efficiency across diverse scenarios.
• It offers practical guidelines for selecting the most suitable routing protocols for specific applications, based on network requirements.
• This study fills gaps in prior research by focusing on underexplored metrics, such as jitter and energy consumption, and by providing a detailed evaluation of hybrid protocols in dynamic MANET environments.

The rest of this paper is organized as follows: Section 2 reviews related work and highlights the gaps addressed by this study. Section 3 describes the routing protocols, categorized into proactive, reactive, and hybrid types. Section 4 explains the simulation setup and the metrics used for evaluation. Finally, Section 5 presents the results, summarizes the findings, and suggests directions for future research.

2. Literature Review

The performance of MANET routing protocols has been studied extensively, with researchers analyzing various metrics to evaluate their efficiency. While these studies provide useful insights, many overlook critical aspects such as hybrid protocols, energy efficiency, and adaptability in real-world conditions. This study aims to fill these gaps by providing a more comprehensive evaluation of routing protocols.

AL-Hasani et al. [2] compared routing protocols based on throughput and delay. They found that reactive protocols like AODV worked better in dynamic networks, while proactive protocols like OLSR were more effective in stable environments. However, their analysis did not include energy consumption or jitter; which are key metrics in this study.

Alkahtani et al. [3] analyzed protocol performance in challenging scenarios, such as high mobility. They highlighted that reactive protocols adapt well to dynamic conditions. However, they did not evaluate energy usage or the trade-offs involved in hybrid protocols. This study builds on their findings by analyzing energy efficiency and demonstrating how hybrid protocols handle dense networks effectively. Hameed et al. [4] provided a performance analysis of routing protocols in mobile ad hoc networks, focusing on metrics such as throughput, delay, and packet delivery ratio but did not consider energy consumption.

Thambusamy V. et al. [6] conducted the performance evaluation of the AODV and TORA routing protocols. The simulation program NS-2 is used to assess these routing protocols in terms of data packet delivery between the source and destination nodes. The mobility model is appropriate for small and medium-sized networks, according to the simulated findings. When the mobile sink movement speed is 5 ms, the TORA protocol achieves a packet delivery ratio of about 58%. However, the AODV received more than TORA. Therefore, when AODV and TORA outcomes are compared in typical scenarios, it is found that AODV performs better than TORA. For the MANET environment selected for this simulation, the AODV protocol is acknowledged and advised. Based on its simulation time, the AODV is found to be the ideal candidate in both cases. Accordingly, this study concludes that, under the assumed hypothesis, the on-demand protocol, AODV, outperformed the TORA protocol.

Mohamed et al. [7] explored the QoS parameters of MANET protocols, including throughput and packet delivery ratio. They showed significant performance differences among protocols at varying network densities. However, their research did not consider energy consumption or jitter, which is essential to fully understanding protocol performance. These gaps are addressed in this study by evaluating an extended set of metrics. Unlike previous studies that relied on older simulation tools like NS2 and OPNET, this research uses OMNeT++ 6.0.1 with INET-4.5.0. This modern tool offers better scalability and flexibility, enabling more accurate simulations of network performance under different conditions.

Fendji J. L. et al. [8] compared the proactive and reactive modes of AODV, OLSR, and HWMP. Energy consumption, throughput, PDR, delay, e-throughput, and e-PDR were among the measures they employed, along with the NS3 simulator. It used two topologies—a grid topology and a movable nodes topology—to assess the routing methods. According to the findings, OLSR is the most effective routing protocol in the world, particularly when it comes to PDR and latency. However, scalability and mobility might have a significant impact on its throughput. AODV appears to be more reliable in many network conditions than OLSR and can provide the same performance as OLSR in many cases.

Jazyah Y. H. et al. [9] compare four wireless routing protocols for MANET (AODV, OLSR, TORA, and DSR) using the OPNET simulator. They also compare some popular simulators. Protocols were compared in terms of throughput, delay, and network overhead; simulation findings indicate that TORA has the most delay while DSR has the best delay out of the four evaluated protocols. When it comes to network overhead, the same findings are noted: DSR has the lowest throughput and AODV the highest.

Rajeswari et al. [10] surveyed energy-efficient routing protocols for MANETs, highlighting the significance of energy consumption in protocol performance. Khudayer et al. [11] analyzed the performance of MANET routing protocols under different mobility models, demonstrating the variability in protocol efficiency based on network dynamics. Quy et al. [12] evaluated MANET routing protocols under realistic environments, emphasizing the importance of practical testing conditions. Their findings showed variations in metrics such as PDR, end-to-end delay, and network lifetime under different real-world scenarios.

This study contributes to the literature by providing a detailed evaluation of eight MANET routing protocols. By incorporating metrics such as energy consumption and jitter, it offers valuable insights into selecting the most suitable protocols for various network applications. The focus on hybrid protocols further highlights their potential for balancing efficiency and adaptability in diverse scenarios.

3. Routing Protocols in MANET

Routing protocols in MANETs can be categorized into three main types: proactive, reactive, and hybrid protocols. The following tables provide an overview of these routing protocol types, their characteristics, and examples:

3.1. Proactive Routing Protocols

Proactive routing protocols maintain up-to-date routing information from each node to every other node in the network, ensuring that routes are readily available when needed. This approach reduces the route discovery delay but increases the overhead due to periodic updates.

Table 1. Proactive Routing Protocols.

Protocol	Description	Advantages	Disadvantages
Optimized Link State Routing (OLSR) [13]	Uses link-state information to establish routes and periodically disseminates topology information.	Low latency due to up-to-date routing information.	High overhead due to periodic updates.
Destination-Sequenced Distance-Vector (DSDV) [14]	Maintains a consistent, up-to-date routing table in each node using sequence numbers to ensure loop-free routes.	Simple and loop-free routing.	High overhead due to frequent table updates.

provides a comprehensive overview of proactive routing protocols used in Mobile Ad Hoc Networks (MANETs). Proactive routing protocols maintain up-to-date routing information by continuously evaluating the network topology and distributing routing tables. OLSR and DSDV are examples of proactive protocols. OLSR uses link-state information to establish routes and periodically disseminates topology information, ensuring low latency due to up-to-date routing information but resulting in high overhead from frequent updates. DSDV maintains a consistent routing table in each node using sequence numbers to ensure loop-free routes, which simplifies routing and avoids loops but also incurs high overhead due to frequent table updates. These protocols are suitable for environments where low latency is critical, but their trade-off in terms of overhead needs careful consideration.

3.2. Reactive Routing Protocols

Reactive routing protocols create routes only when needed, which reduces the overhead but can increase the initial route discovery time.

Table 2. Reactive Routing Protocols.

Protocol	Description	Advantages	Disadvantages
Ad Hoc On-demand Distance Vector (AODV) [16]	Establishes routes to destinations on an as-needed basis using route request and route reply messages.	Lower overhead since routes are created on-demand.	Higher latency due to route discovery process.
Dynamic Source Routing (DSR) [17]	Uses source routing and maintains route caches for data transmission.	Efficient route discovery and maintenance.	High overhead with route caches and source routing.
Ad Hoc On-demand Multipath Distance Vector (AOMDV) [18]	Extends AODV to provide multiple redundant paths to enhance fault tolerance.	Improved reliability and fault tolerance.	Increased complexity and overhead.
Temporally-Ordered Routing Algorithm (TORA) [7]	Establishes a directed acyclic graph for routing based on link reversal.	High adaptability and fault tolerance.	High overhead in maintaining the directed acyclic graph.

outlines the reactive routing protocols that create routes only when needed, significantly reducing overhead in stable networks. AODV routing establishes routes to destinations on an as-needed basis using route request and route reply messages, which lowers overhead but can lead to higher initial latency due to the route discovery process. DSR uses source routing and maintains route caches for efficient data transmission, which reduces the need for frequent route discoveries but introduces overhead from maintaining route caches and source routing [15]. AOMDV extends AODV to provide multiple redundant paths, enhancing fault tolerance but increasing complexity and overhead. TORA establishes a directed acyclic graph for routing based on link reversal, offering high adaptability and fault tolerance but with significant overhead to maintain the graph. These protocols adapt well to dynamic environments, balancing route discovery latency and overhead.

3.3. Hybrid Routing Protocols

Hybrid routing protocols combine features of both proactive and reactive protocols to balance the trade-offs.

Table 3. Hybrid Routing Protocols.

Protocol	Description	Advantages	Disadvantages
Zone Routing Protocol (ZRP) [19]	Divides the network into zones and uses proactive routing within zones and reactive routing between zones.	Combines advantages of proactive and reactive approaches.	Complexity in managing zones.
Geographic Routing Protocol (GRP) [20]	Uses geographic information to improve routing decisions, combining proactive and reactive strategies.	Efficient use of geographic data for routing.	Requires accurate location information.

summarizes hybrid routing protocols that combine features of both proactive and reactive approaches to balance efficiency and adaptability. Hybrid protocols aim to leverage the low latency of proactive routing within local regions while employing reactive methods for distant nodes. ZRP divides the network into zones and uses proactive routing within zones and reactive routing between zones, balancing the benefits of both approaches while managing the complexity of zone maintenance. GRP uses geographic information to improve routing decisions, combining proactive and reactive strategies to efficiently use geographic data for routing, though it requires accurate location information. These protocols optimize performance by leveraging the strengths of both proactive and reactive strategies, making them ideal for diverse network conditions where both local stability and global adaptability are needed.

4. Methodology

We implemented eight routing protocols (OLSR, DSDV, AODV, DSR, AOMDV, TORA, ZRP, and GRP) using the OMNeT++ 6.0.1 simulation environment with INET-4.5.0 as seen in

Table 4. Simulation Network Details.

Parameter	Description
Simulation Tool	OMNeT++ 6.0.1 with INET-4.5.0
Network Type	MANET
Simulation Area	1000 × 1000 units
Node Densities	20, 80, 200, 500 nodes
Connectivity Probability	0.2 (each node has a 20% chance of connecting)
Mobility Model	Random Waypoint Mobility Model
Packet Size	1024 bytes
Packet Arrival Rate	1.0 packets per second
Total Packets per Simulation	100 packets
Routing Protocols	OLSR, DSDV, AODV, DSR, AOMDV, TORA, ZRP, GRP

. When compared to NS2 and OPNET, which are frequently more complex and less flexible, OMNeT++’s modular architecture, user-friendly GUI, scalability, vibrant community, and cross-platform compatibility make it a popular choice for imitating the wide range of routing protocols that were mentioned [21]. The simulations were conducted under four scenarios with varying node densities (20, 80, 200, and 500 nodes). The area size was set to 1000 × 1000 units with a connectivity probability of 0.2. Each simulation involved 100 packets, a packet size of 1024 bytes, and an arrival rate of 1.0 packets per second. The selection of parameters in OMNeT++ simulations, such as a packet arrival rate of 1.0 s and a packet size of 1024 bytes, is crucial for ensuring consistency, relevance, and reproducibility. A packet arrival rate of 1.0 s creates a uniform and predictable traffic pattern, which helps evaluate routing protocols under controlled conditions and reflects real-world scenarios like sensor networks. A packet size of 1024 bytes matches common data sizes in network applications, allowing for meaningful performance assessments (e.g., throughput, delay, packet loss). Using these standard values ensures that the experiment is reproducible and provides relevant, consistent results for real-world MANET applications. The nodes in the simulation were configured to use the Random Waypoint Mobility model, which reflects the random movement patterns of nodes in an ad hoc network and allows for the evaluation of the routing protocols’ performance under dynamic network conditions. The free-space propagation model is used, which is the default radio propagation model in OMNeT++ and follows the inverse square law. It assumes an ideal environment with no obstacles, reflections, or interference and is commonly used in theoretical and simplified simulations.

4.1. Simulation Setup

A random distribution of nodes within the designated area was used to form the network using the random waypoint mobility model. The connection probability was used to determine connectivity between nodes. Performance metrics were recorded while simulating each routing protocol for the specified circumstances.

4.2. Performance Metrics

The following performance metrics were analyzed:

• Network Load: Total total volume of transmitted data (in bytes/sec), including both useful application data and routing protocol overhead.

  Network Load (bps) = Total number of bytes transmitted/Time period

  (1)
• Throughput: Total amount of successfully received useful data (excluding control overhead) per unit time.

  Throughput = Total number of successfully delivered bytes/Time period

  (2)
• Delay: Average time taken for packets to reach the destination.

  Delay = (Sum of individual packet delays)/Total number of successfully delivered packets

  (3)
• Energy Consumption: Total energy consumed during the simulation.

  Energy Consumption = ∑ (Power consumed by each network component × Time component was active)

  (4)
• Jitter: Variation in packet delay.

Jitter = Σ |D(i + 1) − D(i)|/(N − 1)

(5)

where,

D(i) is the one-way delay (or latency) of the i-th packet.

D(i + 1) is the one-way delay of the (i + 1)-th packet.

N is the total number of packets.

• Packet Loss Rate: Ratio of lost packets to total packets sent.

  Packet Loss Rate = (Number of Lost Packets/Total Packets Sent) × 100%

  (6)
• Packet Delivery Ratio (PDR): Ratio of successfully delivered packets to total packets sent.

  PDR = (Number of Packets Delivered/Number of Packets Sent) × 100%

  (7)

4.3. Scenarios and Discussion

The specific routing protocol and its design choices are carefully considered to address the properties of mobile nodes with a random-waypoint mobility model and achieve efficient and reliable data delivery in the given routing scenario. The results for each scenario were analyzed and visualized using various types of charts to enhance clarity and comprehension.

4.3.1. Scenario 1: 20 Nodes

In this scenario, the performance of the eight routing protocols was evaluated in a network consisting of 20 mobile nodes. Each protocol was analyzed based on key performance metrics.

4.3.2. Scenario 2: 80 Nodes

In this scenario, the network consisted of 80 mobile nodes. The performance of the routing protocols was evaluated similarly to Scenario 1. The increased node density and the presence of fixed nodes influenced the performance metrics.

4.3.3. Scenario 3: 200 Nodes

With 200 nodes, the network’s complexity increased, challenging the routing protocols.

4.3.4. Scenario 4: 500 Nodes

In the most complex scenario with 500 nodes, the routing protocols were thoroughly tested. Performance metrics indicated significant differences in protocol efficiency.

5. Results and Comparative Analysis

5.1. Scenarios with Homogeneous Nodes

5.1.1. Network Load

Figure 1 presents a comparative analysis of eight network protocols—proactive (OLSR, DSDV), reactive (AODV, DSR, AOMDV, TORA), and hybrid (ZRP, GRP)—in terms of network load across different network sizes (20, 80, 200, and 500 nodes). The analysis reveals significant trends in network load as the network size scales. Initially, at 20 nodes, all protocols exhibit relatively low network loads, establishing a baseline for comparison. Proactive protocols show lower initial loads due to their constant route maintenance, making them more efficient for smaller networks. However, their performance becomes less optimal for very large networks due to the overhead of maintaining routing tables. As the network size increases to 80 nodes, reactive protocols experience a significant rise in network load due to the increased overhead of on-demand route discovery under moderate traffic conditions.

As the number of nodes increases, more routing control packets are exchanged, leading to a higher network load. For 20 and 80 nodes, fewer control messages are needed due to lower network density. However, at 200 and 500 nodes, routing overhead rises exponentially due to frequent topology changes and route maintenance. Despite this increase, the network reaches a saturation point where adding more nodes does not significantly increase transmitted packets.

The most substantial rise in network load is observed at 200 nodes, reflecting the scalability and adaptability of all protocols to higher node densities. Hybrid protocols, which combine the strengths of both proactive and reactive strategies, show balanced performance, making them most suitable for very high-density networks (500 nodes) where they can efficiently manage the increased load. At 500 nodes, the network load stabilizes across all protocols with minor variations, indicating their efficiency at high network densities. Notably, the GRP protocol shows a higher initial load but aligns with others at higher node counts, while OLSR maintains a consistent increase. This analysis highlights the efficiency and scalability of each protocol category, providing critical insights into their optimal operational conditions and potential performance constraints across various network scenarios.

Table 5. Protocol Suitability for Different Scenarios Considering Network Load.

Network Size (Nodes)	Proactive Protocols (OLSR, DSDV)	Reactive Protocols (AODV, DSR, AOMDV, TORA)	Hybrid Protocols (ZRP, GRP)
20 Nodes	Low network load, efficient for smaller networks	Varying network load, generally higher than proactive protocols	Moderate network load, balanced performance
80 Nodes	OLSR increases just slightly with steady performance	Route discovery overhead causes a notable increase in DSR and TORA	Restricted decrease, but efficient handling of the decreased load
200 Nodes	Sharp rise, high network load	Sharp rise, demonstrating scalability and adaptability	Sharp rise, balanced performance, effective integration of strategies
500 Nodes	Stabilizes with slight variations, less optimal due to maintenance overhead	Stabilizes with similar high values, indicating efficiency	Stabilizes, maintains high network load, most suitable for high-density networks

summarizes the protocol suitability for the four different scenarios considering the network load.

5.1.2. Throughput

Eight network protocols are thoroughly compared in Figure 2 below with respect to throughput across various network sizes (20, 80, 200, and 500 nodes). Significant patterns in throughput performance as network size increases are revealed by the investigation. To create a baseline for comparison, all protocols show different throughput values at the beginning, at 20 nodes. Due to their continuous route maintenance, proactive protocols exhibit stable but low throughput, which makes them more practical for smaller networks. However, the maintenance overhead makes them less effective for very large networks. Under moderate traffic levels, the overhead of route discovery causes reactive protocols to noticeably lose throughput as the network size grows to 80 nodes. The on-demand route-finding of these protocols reduces overhead in less dense networks while demonstrating scalability with more nodes, making them appropriate for situations with moderate to high node densities (80 to 200 nodes).

Throughput decreases when the number of nodes rises from 20 to 80 due to increased MAC layer contention and congestion, which causes more packet collisions, retransmissions, and delays. At 200 nodes, however, the network becomes more interconnected, increasing throughput by decreasing packet drops and enhancing route stability. Because their routing techniques function identically under high mobility and huge network sizes, throughput is consistent across all MANET protocols despite common congestion, interference, and MAC-layer restrictions.

At 200 nodes, the throughput increases to the maximum, demonstrating how all protocols can be scaled and adjusted to larger node densities. Hybrid protocols are best suited for very high-density networks (500 nodes), where they efficiently balance the load and maintain high throughput by fusing the advantages of proactive and reactive tactics. All protocols’ throughput stabilizes around 500 nodes with just slight changes, demonstrating their effectiveness at high network densities. Interestingly, whereas OLSR continues to rise steadily, the GRP protocol first exhibits a higher throughput before lining up with other protocols at increasing node counts. The effectiveness and scalability of each protocol category are highlighted in this analysis, which also offers important insights into their ideal operating environments and possible performance limitations in a range of network scenarios. The protocol’s suitability for each of the four scenarios, taking throughput into account, is summarized in

Table 6. Protocol Suitability for Different Scenarios Considering Throughput.

Network Size (Nodes)	Proactive Protocols (OLSR, DSDV)	Reactive Protocols (AODV, DSR, AOMDV, TORA)	Hybrid Protocols (ZRP, GRP)
20 Nodes	Consistent throughput, efficient for smaller networks	Varying throughput, generally higher than proactive protocols	Moderate throughput, balanced performance
80 Nodes	Slight decrease, consistent performance	Noticeable decrease due to route discovery overhead	Balanced performance, effective management of increased load
200 Nodes	Sharp increase, high throughput	Sharp increase, demonstrating scalability and adaptability	Sharp increase, balanced throughput, effective integration of strategies
500 Nodes	Stabilizes with slight variations, less optimal due to maintenance overhead	Stabilizes with similar high values, indicating efficiency	Stabilizes, maintains high throughput, most suitable for high-density networks

below.

5.1.3. Simulation Time

A detailed comparison of the eight network protocols is shown in Figure 3 below. The data reveal significant patterns in simulation duration with increasing network size. At 20 nodes, the protocols show comparatively short simulation periods at first, providing a starting point for comparing performance. Due to their continuous route maintenance, proactive protocols typically exhibit moderate simulation times; which makes them appropriate for smaller networks. For very large networks, however, the overhead of maintaining routing tables reduces their performance. Under moderate traffic conditions, the overhead of route discovery causes reactive protocols to noticeably increase their simulation time as the network size grows to 80 nodes. These protocols’ on-demand route discovery method can efficiently handle the increasing load in situations with moderate to high node densities (80 to 200 nodes).

When all protocols reach their peak simulation times at 200 nodes, the biggest shift is shown, underscoring the difficulties with scaling and adaptation that come with higher node density. Due to their balanced performance, hybrid protocols—which blend proactive and reactive tactics—are especially well-suited for very high-density networks (500 nodes), where they can effectively handle the additional load. Most protocols show a return to efficiency at the highest network densities, with simulation times tending to steady or drop at 500 nodes. This analysis highlights each protocol category’s effectiveness and scalability while offering important insights into their operational performance and possible limitations in a range of network scenarios. Given the simulation time,

Table 7. Protocol Suitability for Different Scenarios Considering Simulation Time.

Network Size (Nodes)	Proactive Protocols (OLSR, DSDV)	Reactive Protocols (AODV, DSR, AOMDV, TORA)	Hybrid Protocols (ZRP, GRP)
20 Nodes	Moderate simulation time, efficient for smaller networks	Varying simulation time, generally higher than proactive protocols	Moderate simulation time, balanced performance
80 Nodes	Noticeable increase, consistent performance	Significant increase due to route discovery overhead	Balanced increase, effective management of increased load
200 Nodes	Sharp rise, high simulation time	Peak simulation time, highlighting scalability challenges	Peak simulation time, balanced performance, effective integration of strategies
500 Nodes	Decrease or stabilizes, less optimal due to maintenance overhead	Decrease or stabilizes, indicating return to efficiency	Stabilizes, maintains efficient performance, most suitable for high-density networks

below provides a summary of the procedure’s applicability for the four distinct scenarios.

5.1.4. Average Delay

As the network size increases, the research shows notable variations in average delay, as seen in Figure 4. All protocols first show comparatively low average delays at 20 nodes, creating a baseline against which to compare. Because proactive protocols maintain their routes continuously, they exhibit moderate latency, which makes them more effective for smaller networks. However, the overhead of maintaining routing tables makes their performance less than ideal for very large networks. The overhead of route finding under moderate traffic levels causes reactive protocols to noticeably increase their average delay as the network capacity grows to 80 nodes.

These protocols work most effectively in situations where there are moderate to high node densities (80–200 nodes), as their as-needed route-finding way can handle the extra load. Congestion rises when the network expands from 20 to 200 nodes because of increased contention, retransmissions, and routing overhead, which results in longer pathways and longer delays. End-to-end latency is decreased by the network’s density at 500 nodes, which permits more direct routes and fewer hops. Certain protocols minimize latency in crowded networks by further optimizing route finding.

At 200 nodes, the average delay increases the highest, underscoring the difficulties in scalability and adaptability that come with increasing node density. Because hybrid protocols exhibit balanced performance by combining proactive and reactive techniques, they are especially well-suited for very high-density networks (500 nodes), where they can effectively handle the increased delay. Average delays for the majority of protocols tend to settle at 500 nodes, suggesting that efficiency returns at the largest network densities. The effectiveness and scalability of each protocol category are highlighted in this analysis, which also offers important insights into their operational performance and possible limitations in a range of network scenarios. The protocol’s suitability for the four distinct cases is compiled in

Table 8. Protocol Suitability for Different Scenarios Considering the Average Delay.

Network Size (Nodes)	Proactive Protocols (OLSR, DSDV)	Reactive Protocols (AODV, DSR, AOMDV, TORA)	Hybrid Protocols (ZRP, GRP)
20 Nodes	Moderate delay, efficient for smaller networks	Varying delay, generally higher than proactive protocols	Moderate delay, balanced performance
80 Nodes	Noticeable increase, consistent performance	Significant increase due to route discovery overhead	Balanced increase, effective management of increased delay
200 Nodes	Sharp rise, high delay	Peak delay, highlighting scalability challenges	Peak delay, balanced performance, effective integration of strategies
500 Nodes	Decrease or stabilizes, less optimal due to maintenance overhead	Stabilizes, indicating return to efficiency	Stabilizes, maintains efficient performance, most suitable for high-density networks

below, taking the average delay into account.

5.1.5. Energy Consumption

In Figure 5, a thorough comparison of energy consumption across various network sizes is shown. At first, all protocols use comparatively little energy at 20 nodes, creating a baseline against which to compare. Due to their continuous route maintenance, proactive protocols exhibit minimal energy consumption, which makes them more effective for smaller networks. For very large networks, however, the overhead of maintaining routing tables reduces their performance. Reactive protocols noticeably use more energy when the network size reaches 80 nodes because of the overhead of route discovery in moderate traffic scenarios. Under moderate traffic levels, the overhead of route finding causes reactive protocols to noticeably increase their energy consumption as the network size grows to 80 nodes.

At 200 nodes, the energy usage increases the most, underscoring the difficulties in scalability and adaptability that come with increasing node density. Because hybrid protocols exhibit balanced performance by combining proactive and reactive techniques, they are especially well-suited for very high-density networks (500 nodes), where they can effectively handle the increased energy consumption. Energy usage generally stabilizes at 500 nodes for the majority of protocols, suggesting a return to efficiency at the highest network densities. The effectiveness and scalability of each protocol category are highlighted in this analysis, which also offers important insights into their operational performance and possible limitations in a range of network scenarios. The protocol’s suitability for the four distinct scenarios, taking energy usage into account, is summarized in

Table 9. Protocol Suitability for Different Scenarios Considering Energy Consumption.

Network Size (Nodes)	Proactive Protocols (OLSR, DSDV)	Reactive Protocols (AODV, DSR, AOMDV, TORA)	Hybrid Protocols (ZRP, GRP)
20 Nodes	Moderate energy consumption, efficient for smaller networks	Varying energy consumption, generally higher than proactive protocols	Moderate energy consumption, balanced performance
80 Nodes	Noticeable increase, consistent performance	Significant increase due to route discovery overhead	Balanced increase, effective management of increased energy consumption
200 Nodes	Sharp rise, high energy consumption	Peak energy consumption, highlighting scalability challenges	Peak energy consumption, balanced performance, effective integration of strategies
500 Nodes	Decrease or stabilizes, less optimal due to maintenance overhead	Stabilizes, indicating return to efficiency	Stabilizes, maintains efficient performance, most suitable for high-density networks

below.

The fundamental cause of the high energy consumption of active protocols—especially proactive ones—is the requirement for ongoing routing table updating. These protocols cause a large overhead in control message traffic because they constantly distribute updates to guarantee that all nodes have the most recent routing information. Nodes continuously send and receive data packets, even when no data transmission is taking place, which depletes battery life. Energy consumption is further increased by the requirement to continuously monitor network topology because nodes are still active and involved in routing tasks. Nodes must constantly adapt to account for node mobility in dynamic contexts, which adds another layer of energy consumption because they must exert more effort to maintain connectivity and routing accuracy. As a result, proactive protocols are less appropriate for situations with limited energy resources because, although they offer low-latency routing under stable circumstances, their operational requirements might result in quick energy loss.

5.1.6. Jitter

A jitter-based analysis is shown in Figure 6. To create a baseline for comparison, all protocols show comparatively minimal jitter at 20 nodes. Due to their continuous route maintenance, proactive protocols exhibit moderate jitter, which makes them more effective for smaller networks. For very large networks, however, the overhead of maintaining routing tables reduces their performance. Reactive protocols noticeably increase jitter as the network capacity reaches 80 nodes because of the overhead of route discovery in moderate traffic situations. When a network expands from 20 to 200 nodes, jitter increases due to congestion and frequent route changes, that alter packet arrival times. At 500 nodes, however, denser connections offer more dependable and stable pathways, which lower jitter by guaranteeing more constant delivery times.

Higher node density presents scalability and adaptation problems, as evidenced by the most noticeable increase in jitter at 200 nodes. Due to their balanced performance, hybrid protocols—which blend proactive and reactive tactics—are especially well-suited for very high-density networks (500 nodes), where they can effectively handle the increased jitter. For most protocols, jitter tends to stabilize or decrease at 500 nodes, suggesting that efficiency has returned at the highest network densities. This analysis highlights each protocol category’s effectiveness and scalability while offering important insights into their operational performance and possible limitations in a range of network scenarios.

With an emphasis on jitter,

Table 10. Protocol Suitability for Different Scenarios Considering Jitter.

Network Size (Nodes)	Proactive Protocols (OLSR, DSDV)	Reactive Protocols (AODV, DSR, AOMDV, TORA)	Hybrid Protocols (ZRP, GRP)
20 Nodes	Moderate jitter, efficient for smaller networks	Varying jitter, generally higher than proactive protocols	Moderate jitter, balanced performance
80 Nodes	Noticeable increase, consistent performance	Significant increase due to route discovery overhead	Balanced increase, effective management of increased jitter
200 Nodes	Sharp rise, high jitter	Peak jitter, highlighting scalability challenges	Peak jitter, balanced performance, effective integration of strategies
500 Nodes	Decrease or stabilizes, less optimal due to maintenance overhead	Stabilizes, indicating return to efficiency	Stabilizes, maintains efficient performance, most suitable for high-density networks

shows how well proactive, reactive, and hybrid routing systems perform across a range of network sizes. Due to their regular route updates, proactive protocols are effective in small networks (20 nodes). However, because of the complexity of maintaining routing tables, they have trouble scaling in bigger networks (200 nodes), where jitter increases dramatically. Reactive protocols have increased jitter because of on-demand route discovery, which peaks at medium and large network sizes (80 and 200 nodes), even if they are more flexible for smaller and dynamic networks. They do, however, stabilize in networks with 500 nodes or more, suggesting that efficiency improves with increased connectedness. Hybrid protocols efficiently manage jitter even as the network expands by continuously balancing performance across all network sizes.

5.1.7. Packet Loss Rate

All protocols show comparatively significant packet loss rates at 20 nodes, as shown in Figure 7, providing a baseline against which to compare. Because proactive protocols maintain their routes continuously, they exhibit a moderate packet loss rate, which makes them more effective for smaller networks. However, the overhead of maintaining routing tables makes their performance less than ideal for very large networks. The overhead of route discovery under moderate traffic conditions causes reactive protocols to noticeably reduce their packet loss rate as the network capacity grows to 80 nodes. Packet loss rates are comparable in smaller networks (20 and 80 nodes) due to less congestion and stable pathways. Higher connectivity enhances redundancy as the network expands to 200 and particularly 500 nodes, enabling dropped packets to be promptly diverted and minimizing packet loss.

At 200 nodes, the packet loss rate shows the biggest improvement, underscoring the difficulties in scalability and adaptation that come with increasing node density. Because hybrid protocols exhibit balanced performance by combining proactive and reactive techniques, they are especially well-suited for very high-density networks (500 nodes), where they can effectively handle the lower packet loss rate. Most protocols exhibit a large decrease in packet loss rates at 500 nodes, suggesting a return to efficiency at the highest network densities. The effectiveness and scalability of each protocol category are highlighted in this analysis, which also offers important insights into their operational performance and possible limitations in a range of network scenarios. The protocol’s suitability for each of the four cases, taking into account the packet loss rate, is summarized in

Table 11. Protocol’s Suitability for Different Scenarios Considering the Packet Loss Rate.

Network Size (Nodes)	Proactive Protocols (OLSR, DSDV)	Reactive Protocols (AODV, DSR, AOMDV, TORA)	Hybrid Protocols (ZRP, GRP)
20 Nodes	Moderate packet loss rate, efficient for smaller networks	High packet loss rate, generally higher than proactive protocols	Moderate packet loss rate, balanced performance
80 Nodes	Noticeable decrease, consistent performance	Significant decrease due to route discovery overhead	Balanced decrease, effective management of reduced packet loss rate
200 Nodes	Sharp improvement, low packet loss rate	Peak improvement, highlighting scalability benefits	Peak improvement, balanced performance, effective integration of strategies
500 Nodes	Stabilizes or decreases, more optimal due to efficiency	Significant decrease, indicating return to efficiency	Stabilizes, maintains low packet loss rate, most suitable for high-density networks

below.

5.1.8. Packet Delivery Ratio

In terms of packet delivery ratio (PDR), a thorough comparison is shown in Figure 8 below. A baseline for comparison is established at 20 nodes, where all techniques display different PDRs. Because proactive protocols continuously maintain their routes, they exhibit greater PDRs, which makes them more effective for smaller networks. For very large networks, however, the overhead of maintaining routing tables reduces their performance. Under moderate traffic levels, the overhead of route discovery causes reactive protocols to noticeably lose PDR as the network size grows to 80 nodes. PDR is decreased by more congestion, collisions, and interference between 20 and 80 nodes. Higher connectivity at 200 and 500 nodes, however, makes it possible for more dependable multi-hop pathways, which enhances protocol performance and boosts the number of successful packet deliveries.

The biggest PDR improvement is shown at 200 nodes, underscoring the difficulties with scalability and adaptability that come with greater node density. Because hybrid protocols exhibit balanced performance by combining proactive and reactive techniques, they are especially well-suited for very high-density networks (500 nodes), where they can effectively handle the elevated PDR. For the majority of protocols, PDR tends to stabilize or rise at 500 nodes, suggesting a return to efficiency at the highest network density. The effectiveness and scalability of each protocol category are highlighted in this analysis, which also offers important insights into their operational performance and possible limitations in a range of network scenarios. The protocol appropriateness for the four distinct scenarios considering PDR is summed up in

Table 12. Protocol Suitability for Different Scenarios Considering Packet Delivery Ratio.

Network Size (Nodes)	Proactive Protocols (OLSR, DSDV)	Reactive Protocols (AODV, DSR, AOMDV, TORA)	Hybrid Protocols (ZRP, GRP)
20 Nodes	High PDR, efficient for smaller networks	Varying PDR, generally lower than proactive protocols	Moderate PDR, balanced performance
80 Nodes	Noticeable decrease, consistent performance	Significant decrease due to route discovery overhead	Balanced decrease, effective management of increased PDR
200 Nodes	Sharp improvement, high PDR	Peak improvement, highlighting scalability benefits	Peak improvement, balanced performance, effective integration of strategies
500 Nodes	Stabilizes or increases, more optimal due to efficiency	Significant increase, indicating return to efficiency	Stabilizes, maintains high PDR, most suitable for high-density networks

below.

5.2. Comparative Analysis of Routing Protocols with Homogeneous Nodes

The comparative analysis examines and shows the performance of proactive, reactive, and hybrid routing protocols in various contexts, as shown in

Table 13. Comparative Analysis of Routing Protocols Considering All Metrics.

Metric	Proactive	Reactive	Hybrid
Network Load	Low (10%)	High (20%)	Moderate (15%)
Throughput	High (90%)	Lower (70%)	High (85%)
Average Delay	Low (20 ms)	High (80 ms)	Moderate (40 ms)
Energy Consumption	Low (5 joules)	High (10 joules)	Moderate (7 joules)
Jitter	Low (2 ms)	High (10 ms)	Moderate (5 ms)
Packet Loss Rate	Low (5%)	High (15%)	Moderate (10%)
Packet Delivery Ratio	High (95%)	Lower (85%)	High (90%)

. Using metrics including network load, throughput, average delay, energy consumption, jitter, packet loss rate, and PDR, this research illustrates the trade-offs and applicability of each protocol type for different network situations.

The applicability of proactive, reactive, and hybrid routing protocols under various network conditions and performance trade-offs has been clarified by this comparative investigation. Network load, throughput, average latency, energy consumption, jitter, packet loss rate, and packet delivery ratio are performance indicators that offer valuable insights into the strengths and shortcomings of each protocol. Figure 9, Figure 10, Figure 11 and Figure 12 display the outcomes of various homogenous node scenarios.

Table 14. Protocol Suitability across the Four Scenarios.

Network Size (Nodes)	Proactive Protocols	Reactive Protocols	Hybrid Protocols
20 Nodes	Low delay, jitter, energy	Higher delay, jitter	Moderate metrics across all
80 Nodes	Consistent performance	Significant increase in load, delay	Balanced performance
200 Nodes	High throughput, stable	Improved scalability	Peak performance
500 Nodes	Less optimal, higher overhead	Efficient, low packet loss	Highly effective, stable

also shows the protocol suitability across the four scenarios.

• Proactive Protocols: With constant route maintenance, proactive protocols excel in scenarios requiring low average delay, jitter, and energy consumption. They are most effective for smaller networks where stability and immediate route availability are crucial.
• Reactive Protocols: Reactive protocols, characterized by higher delay and jitter due to on-demand route discoveries, demonstrate effective scalability in moderately dense networks. Their ability to dynamically manage overhead makes them suitable for networks with moderate to high node densities.
• Hybrid Protocols: Combining the strengths of proactive and reactive approaches, hybrid protocols offer balanced performance across all metrics. They are particularly effective in highly dense network environments, maintaining high throughput, moderate delay, and energy efficiency while ensuring low packet loss rates and high packet delivery ratios.

These understandings are essential for choosing the best routing protocol depending on particular network needs, guaranteeing dependable and effective network performance. By comprehending the subtle trade-offs between numerous protocol types, network designers can modify their strategy to satisfy the requirements of diverse network scenarios, thus improving overall performance and dependability.

5.3. Scenarios with Heterogeneous Nodes

The network’s performance can be efficiently simulated and analyzed under a variety of conditions, including the effect of node mobility, network size, and the mix of fixed and mobile nodes. A realistic and representative heterogeneous network scenario can be created with the aid of the Random Waypoint Mobility Model for mobile nodes and the static placement of fixed nodes. This allows for the evaluation of crucial performance metrics and the creation of efficient communication strategies.

Heterogeneous nodes are different types of nodes utilized in a network, with the number of mobile and fixed nodes varying across configurations [22]. This section describes the performance of heterogeneous networks with various combinations of mobile and fixed nodes. By gradually increasing the number of both node types, the analysis reveals how the network’s scalability affects important performance measures. Adding a ground speed of 3 m/s to the Random Waypoint Mobility Model for the mobile nodes makes the simulated heterogeneous network more like real-life situations where things are both fixed and moving slowly. This lets us do more accurate performance tests and come up with better ways to communicate.

In the first scenario, which involves 10 mobile nodes and 10 fixed nodes, we observe a small deployment that is particularly useful for localized applications or proof-of-concept studies. As the network scales to 40 mobile and 40 fixed nodes, and then to 100 mobile and 100 fixed nodes, the network’s capacity and complexity increase significantly. This growth leads to changes in various performance metrics such as total load, throughput, latency, energy consumption, jitter, packet loss, and delivery ratio. The final scenario, featuring 250 mobile and 250 fixed nodes, represents a large-scale heterogeneous network akin to a smart city or IoT infrastructure. Here, the sheer number of devices introduces new challenges related to coordination, resource management, and maintaining quality of service.

By evaluating this spectrum of network sizes, we can identify scaling trends, bottlenecks, and trade-offs among performance indicators. This understanding is vital for creating and optimizing heterogeneous communication systems that can adapt to the diverse needs of various applications and deployment contexts.

In the experiment with 10 mobile and 10 fixed nodes, the role of fixed nodes as signal relays becomes crucial. Their effectiveness in improving overall network latency and throughput can be significant, as they facilitate more stable connections and reduce the distance that mobile nodes need to cover to reach one another. By acting as intermediaries, fixed nodes can help maintain a reliable communication path, thereby enhancing the overall performance of the network.

5.4. Comparative Analysis of Routing Protocols with Heterogeneous Nodes

This analysis highlights the importance of fixed nodes in small-scale deployments and their potential to optimize network efficiency in more complex configurations.

Table 15. Performance Metrics Across Different Heterogeneous Scenarios.

Metric	10 Mobile, 10 Fixed	40 Mobile, 40 Fixed	100 Mobile, 100 Fixed	250 Mobile, 250 Fixed
Network Load (%)	10	15	25	35
Throughput (kbps)	85	80	75	70
Average Delay (ms)	20	30	60	50
Energy Consumption (J)	5	7	9	12
Jitter (ms)	2	4	8	6
Packet Loss Rate (%)	5	10	15	8
Packet Delivery Ratio (%)	95	90	85	92

below represents the performance metrics across all the different heterogeneous scenarios.

The outcomes of adopting various scenarios of heterogeneous nodes are displayed in Figure 13, Figure 14, Figure 15 and Figure 16. Analyzing these situations helps identify bottlenecks, comprehend the scaling structure, and choose and optimize protocols with knowledge. Heterogeneous communication systems can be developed and enhanced through research.

This comprehensive performance evaluation of eight MANET routing protocols using OMNeT++ 6.0.1 with INET-4.5.0 provides valuable insights into the strengths and weaknesses of each protocol type across different network conditions. The analysis highlights that:

• Proactive Protocols (OLSR, DSDV): These protocols perform efficiently in smaller, stable networks by maintaining low delay, jitter, and energy consumption due to their constant route updates. However, they become less efficient in larger, dynamic networks because of the increased overhead of maintaining routing tables.
• Reactive Protocols (AODV, DSR, AOMDV, and TORA): Characterized by higher delay and jitter due to on-demand route discoveries, reactive protocols demonstrate effective scalability in moderately dense networks. They are suitable for networks with moderate to high node densities, where dynamic route management reduces overhead.
• Hybrid Protocols (ZRP, GRP): Combining the strengths of proactive and reactive approaches, hybrid protocols offer balanced performance across all metrics. They are particularly effective in highly dense network environments, maintaining high throughput, moderate delay, and energy efficiency while ensuring low packet loss rates and high packet delivery ratios.

The evaluation of heterogeneous scenarios, with varying combinations of mobile and fixed nodes, underscores the adaptability of each protocol type. Proactive protocols are most suitable for small-scale deployments, reactive protocols for moderately dense networks, and hybrid protocols for large-scale, heterogeneous networks. This nuanced understanding of protocol performance across different network conditions is critical for optimizing the design and deployment of MANETs in real-world applications, enhancing overall efficiency and reliability. These findings provide a comprehensive guide for selecting the optimal routing protocol based on specific network requirements, ensuring efficient and reliable network performance under diverse conditions.

5.5. Overall Results

The comparison of the homogeneous and heterogeneous networks with 20 nodes shows distinct effects on proactive, reactive, and hybrid protocols, as shown in

Table 16. Comparison of All Performance Metrics for Homogeneous and Heterogeneous Nodes.

Number of Nodes	Metrics	Homogeneous	Heterogeneous
20	Network Load	Higher	Lower
	Throughput	Higher	Lower
	Simulation Time	Higher	Lower
	Average Delay	Lower	Higher
	Energy Consumption	Lower	Higher
	Jitter	Higher	Lower
	Packet Loss Rate	Lower	Higher
80	Network Load	Higher	Lower
	Throughput	Higher	Lower
	Simulation Time	Higher	Lower
	Average Delay	Higher	Lower
	Energy Consumption	Higher	Lower
	Jitter	Higher	Lower
	Packet Loss Rate	Higher	Lower
200	Network Load	Lower	Higher
	Throughput	Higher	Lower
	Simulation Time	Higher	Lower
	Average Delay	Higher	Lower
	Energy Consumption	Higher	Lower
	Jitter	Higher	Lower
	Packet Loss Rate	Higher	Lower
500	Network Load	Lower	Higher
	Throughput	Higher	Lower
	Simulation Time	Higher	Lower
	Average Delay	Higher	Lower
	Energy Consumption	Higher	Lower
	Jitter	Higher	Lower
	Packet Loss Rate	Higher	Lower

. In proactive protocols, such as OLSR and DSDV, the network’s heterogeneity seems to slightly increase the network load and energy consumption, likely due to the continuous exchange of routing information between mobile and fixed nodes. Reactive protocols, such as AODV and DSR, have similar throughput and packet delivery rates in both types of networks. However, simulation times are shorter in the homogeneous network, which suggests that routes can be found more quickly in an environment where everyone moves around the same amount. Hybrid protocols, such as ZRP and GRP, maintain consistent performance across both scenarios, though the heterogeneous network might experience a slight increase in delay due to the interaction between mobile and fixed nodes. Overall, while proactive protocols might be more affected by the network’s heterogeneity, reactive and hybrid protocols appear to adapt well to both network types, with only minor variations in performance metrics.

In the 80-node scenario, the effect of homogeneous versus heterogeneous networks on the three categories of protocols reveals key differences. Proactive protocols, such as OLSR and DSDV, experience a higher network load and energy consumption in the heterogeneous network due to the continuous need to update routes between mobile and fixed nodes. Reactive protocols, like AODV and DSR, show consistent performance across both network types, with slightly lower simulation times in the homogeneous network, likely due to simpler routing paths. Hybrid protocols, including ZRP and GRP, maintain stable performance across both scenarios, demonstrating their flexibility, with only minor variations in metrics like delay and jitter. Overall, the presence of fixed nodes in the heterogeneous network introduces more complexity for proactive protocols, while reactive and hybrid protocols adapt well to both network types.

In the 200-node scenario, the impact of homogeneous and heterogeneous networks on proactive, reactive, and hybrid protocols demonstrates notable differences. Proactive protocols experience higher network load and energy consumption in the heterogeneous network due to the added complexity of managing routes between a mix of mobile and fixed nodes. Reactive protocols show similar throughput and delay across both network types, but the homogeneous network exhibits slightly lower simulation times, likely because of the uniform mobility, which simplifies routing processes. Hybrid protocols maintain consistent performance in both network types, but the heterogeneous network shows a slight increase in packet loss rate and energy consumption, likely due to the interaction between different node types. Overall, while proactive protocols are more affected by network heterogeneity; reactive and hybrid protocols exhibit more stable behavior across both homogeneous and heterogeneous network scenarios.

In the 500-node scenario, the effects of homogeneous and heterogeneous networks on the three protocol categories become more pronounced. Proactive protocols, such as OLSR and DSDV, continue to show increased network load and energy consumption in the heterogeneous network, which is expected due to the overhead of maintaining routes in a larger and more complex network with both mobile and fixed nodes. Reactive protocols, including AODV and DSR, maintain consistent performance across both network types, but the homogeneous network shows slightly lower simulation times, likely due to the uniformity in node mobility. Hybrid protocols demonstrate stable behavior in both scenarios, with only minor differences in delay and jitter, although the heterogeneous network exhibits a slightly higher packet loss rate. Overall, the introduction of fixed nodes in the heterogeneous network tends to increase the complexity and overhead for proactive protocols, while reactive and hybrid protocols remain adaptable and efficient across both network types, even at larger scales.

Based on the results of the four scenarios with 20, 80, 200, and 500 nodes, it is evident that the type of network—whether homogeneous or heterogeneous—significantly affects the performance of different protocol categories. Proactive protocols, such as OLSR and DSDV, consistently show higher network load and energy consumption in heterogeneous networks due to the continuous need to maintain up-to-date routing information across both mobile and fixed nodes. As the network size increases, this effect becomes more pronounced, particularly in larger networks like the 500-node scenario. Reactive protocols, including AODV and DSR, demonstrate stable performance across both network types, with a slight advantage in simulation time within homogeneous networks where routing paths are more straightforward due to the uniform mobility of nodes. Hybrid protocols, such as ZRP and GRP, exhibit consistent and adaptable performance across all scenarios, with minimal impact from network heterogeneity, though they do show slight increases in delay and packet loss in larger heterogeneous networks. Overall, as the network size scales, homogeneous networks offer simplicity and efficiency, particularly for reactive protocols, while heterogeneous networks introduce more complexity, particularly affecting proactive protocols, but can still maintain robust performance with hybrid protocols.

By aligning protocol selection with network size, network administrators and system designers can optimize performance to suit specific application requirements and operating conditions. In smaller networks, proactive protocols are advantageous due to their low overhead and immediate route availability, making them suitable for scenarios requiring minimal latency and stable topology. As network size increases, reactive protocols become more effective in managing dynamic changes while reducing the overhead of maintaining routing tables. For large and very large networks, hybrid protocols are the optimal choice, as they combine the strengths of proactive and reactive approaches to balance scalability, efficiency, and reliability. This strategic selection ensures that the chosen protocol aligns with the network’s scale and complexity, leading to improved communication performance, reduced energy consumption, and enhanced overall efficiency.

In the context of MANET, selecting the appropriate routing protocol is crucial for optimizing performance based on specific application scenarios. Proactive protocols, such as OLSR and DSDV, are advantageous in low-density, stable networks like military operations or emergency response scenarios, as they provide low latency due to readily available routes and maintain consistent routing information. However, their high overhead from frequent routing table updates and increased energy consumption can be significant drawbacks. Conversely, reactive protocols, such as AODV and DSR, are better suited for dynamic environments with high node mobility, like disaster recovery operations, where they minimize overhead by establishing routes on demand. Despite their adaptability, these protocols may experience increased latency during the route discovery process. Hybrid protocols, such as ZRP and GRP, effectively balance the strengths and weaknesses of both proactive and reactive approaches, making them ideal for applications like sensor networks. In such scenarios, hybrid protocols can maintain stability while also accommodating varying latency needs, thus ensuring efficient data transmission and energy usage in dynamic environments.

6. Conclusions

This study evaluated the performance of eight MANET routing protocols under dynamic network conditions using OMNeT++ 6.0.1 with INET-4.5.0. Simulations were conducted with varying node densities ranging from 20 to 500 nodes, and the Random Waypoint Mobility model was used to replicate realistic node movement patterns. Key performance metrics, including throughput, delay, jitter, energy consumption, and packet delivery ratio, were analyzed to assess each protocol’s strengths and weaknesses.

The results showed that proactive protocols like OLSR and DSDV are well-suited for stable environments with moderate mobility due to their consistent performance in small networks. Reactive protocols, including AODV, AOMDV, and TORA, performed better in dynamic and high-mobility scenarios, where adaptability is critical. Hybrid protocols, particularly GRP, demonstrated a balanced approach, combining the benefits of both proactive and reactive strategies. GRP achieved superior overall performance, including higher delivery ratios and up to 30% lower energy consumption compared to reactive protocols, making it ideal for diverse and large-scale networks.

These findings provide practical insights for selecting routing protocols tailored to specific applications, such as IoT systems, disaster recovery, and military operations. Future work can focus on integrating machine-learning techniques with hybrid protocols to further enhance their efficiency and adaptability in complex network environments.