Performance analysis and enhancement of position-based routing protocols in MANETS

Abstract

A Mobile Ad-hoc Network (MANET) is an assortment of wireless nodes that creates a temporal network between mobile nodes with no centralized administrator. The performance of MANET can be assessed by virtue of implementation of routing protocols. Continuous movement of mobile nodes in MANET results in frequent topology changes causing a major performance challenge. Several MANET routing protocols are available within different categories, mainly topology-based and position-based routing protocols. This work presents a new performance evaluation framework, aiming to find optimal position-based routing protocols features achieving best performance within different environmental conditions considering; terrain areas, density of nodes and mobility speeds.

The performance analysis carried out in this work focused on a set of metrics including; delay, throughput, packet delivery ratio and power consumption. Results have indicated that LAR protocol achieved best performance considering all metrics, while DREAM protocol proven to attain best energy consumption. Results were investigated in terms of scalability, mobility and energy efficiency, in order to select optimal protocol features and provide recommendations for future design and implementation of position-based routing protocols.

Keywords

MANET routing protocols throughput delay PDR

1. Introduction

MANET represents a form of new technology for wireless networks in which mobile devices communicate among each other without stationary infrastructure by using multi-hop links. According to Johnson et al. [4], MANET routing protocols help different wireless mobile nodes to establish a wireless network that functions independent of any centralized administration. Here, point-to-point routing is used across a multi-hop connection between two nodes [6]. This communication involves the mechanism of finding the shortest path from a source node to the desired destination node along which data packets can be transferred. Therefore, efficiency of the routing protocol implemented determines the success, reliability and proficiency of communications in MANET. MANETs routing protocols are divided into two major categories; topology-based and position-based routing protocols. Topology-based protocols face various challenges and becomes unsuitable with a high-density mobile environment and with increased network size, leading to high bandwidth consumption and increased end to end delay [17].

Therefore, position-based routing protocols can be utilized to enhance the scalability and efficiency of MANETS. Position-based strategy eliminates and resolves topology-based limitations after utilizing physical location and distance information of the source, neighbors and destination nodes, tracked by sensors or positioning techniques, in order to enhance packet forwarding design in the networks [20, 12, 5].

This research work focuses on investigating the limitations and capabilities of position-based routing protocols in MANETs. The objective is to provide a new framework of selecting appropriate MANET routing protocols achieving optimal performance, considering network size, network density and node mobility speed. This work is conducted on the following steps:

•
State of art analysis of literature with regard to position-based routing protocols performance taking into consideration different MANET environmental conditions and protocol features.
•
The design and implementation of a new hierarchical evaluation framework for performance analysis of position-based protocols taking into consideration best and worst case scenarios of MANET environmental conditions.
•
The customization and implementation of several position-based and topology-based routing protocols within hierarchical evaluation structure:

–
Highly Dynamic Destination-Sequence Distance-Vector Routing (DSDV).
–
Location Aided Routing Protocol (LAR).
–
Distance Routing Effect Algorithm for Mobility (DREAM).
–
Ad Hoc On-Demand Distance Vector (AODV).
–
Zone Routing Protocol (ZRP).

•
A detailed investigation of performance metrics achieved after assessing selected routing protocols within best and worst case scenarios. The results of this investigation allowed summarizing best position-based protocol features achieving optimal performance within MANET networks.

Section 2 describes related work regarding to MANET different routing protocols; operations, performance and the theoretical concepts. Section 3 presents the evaluation methodology. Section 4 describes the performance metrics, and simulation scenarios. Section 5 provides the analysis of results. Section 6 concludes this work and provide future recommendations.
2. Related work

A performance comparison of four topology-based routing protocols including DSDV, AODV, Temporally Ordered Routing Algorithm (TORA), and Dynamic Source Routing (DSR), was conducted in [10]. It was observed that AODV demonstrates significantly better performance as compared to DSR and TORA protocols. An intensive performance comparison between a set of MANET routing protocols been carried out in Bhatia et al., (2015) using different simulation scenarios with varying network sizes and mobile node speeds. Evaluated protocols represents three major topology-based routing protocols including proactive, reactive and hybrid protocols. It was observed that, reactive routing protocols significantly outperform proactive routing protocols taking into consideration performance parameters such as throughput and packet delivery ratio. However, in context of E2E and NRL, proactive routing protocols perform better than the reactive routing protocols. Here, AODV was observed to be the best routing protocol in the context of overall performance followed by protocols DSR, OLSR and DSDV that exhibited average performance and, ZRP showed the worst performance.

Husain et al. [11] tested routing protocols LAR, DSR and AODV for different metrics with respect to node density. According to claimed results, position-based routing protocol LAR has outperformed topology-based routing protocols DSR and AODV in different scenarios. Furthermore, Ahvar et al., (2007) analyzed the performance of three MANET routing protocols; DSR, AODV and LAR-1. This study has focused on energy consumption with respect to varying node mobility, network size, and network load. In low density networks, it was observed that DSR consumes the lowest energy consumption, however in high density networks LAR-1 achieved best power consumption efficiency. Similar conclusion regarding LAR and DSR energy consumption was reached by Neelesh and Roopam [14]. In addition, it was recorded that LAR demonstrates optimal performance to DSR in terms of routing overhead, packet loss ratio, end to end delay, and PDR.

Routing protocols scalability is considered an important performance factor to be measured. In this context, Jörg [9] investigates the performance scalability of a four different MANET routing protocols in terms of network size, these protocols were AODV, DSR, LAR-1 and ZRP). It was noticed that in large scale networks (network size up-to 200 nodes) LAR-1 outperformed other protocols in terms of PDR and routing overhead. In addition, Broustis et al. [8] measured the performance of; DSR, AODV, TORA and LAR, it was found that DSR does not scale well in terms of end-to-end delay in large-scale MANETs as compared to its performance observed in small scale MANETS.

Performance analysis of position-based and topology-based routing protocols has been conducted by several studies [3, 7, 12, 22, 23]. A set of common protocols were investigated in these studies including AODV and DSR for Topology-based protocols (AODV and DSR), and GSR (Geographic Source Routing), GPSR (Geographic Position Source Routing) for Position-based routing protocols. Results have shown different performance levels achieved by selected protocols, in which GPSR and GSR achieved lowest PDR using CBR rate. However, GSR outperformed AODV and DSR in with reference to latency and PDR. A similar study reported in Alnabhan et al. [12], in which a group of protocols from both categories (DSDV, DSR, ZRP, and PDGR) were selected considering packet- delivery ratio, End-to End delay, and routing overhead. Results confirmed that position-based protocol (PDGR) achieves better performance than topology-based protocols (DSDV, DSR and ZRP) in high dynamic and dense environments.

Accordingly, the performance of MANET routing protocols has been widely investigated in previous research. However, these studies were conducted with single view and do not represent all environmental settings, hence reported results are not matured enough to be accurately used in reaching recommendations of future protocol design and developments. Hence, in this work a constructive and hierarchal evaluation framework is presented and utilized to integrate all possible MANET environmental settings and scenarios ranging from best to worst, allowing for the extraction of optimal protocol features achieving best performance in all conditions.

3. Selected routing protocols

In this work, four different MANET protocols; LAR, DREAM, ZRP, and AODV have been selected to conduct an intensive performance evaluation study. LAR and DREAM are considered position-based protocols, ZRP and AODV are topology-based protocols. These protocols were selected based different routing strategies, routing philosophy and routing metrics:

1.
Routing strategy:

–
Geographic position assisted (LAR, DREAM).
–
Flat routing or distance based (AODV).
–
Hierarchal routing (ZRP).

2.
Routing philosophy:

–
On-demand or reactive (LAR, AODV).
–
Proactive (DREAM).
–
Hybrid: reactive and proactive (ZRP).

3.
Routing metrics (path): All of the selected protocols use the shortest path metrics instead of the ZRP use the local shortest path.

3.1 Location Aided Routing (LAR)

LAR utilizes Global Positioning System (GPS) to estimate destination node coordinates. LAR route discovery is based on flooding, and reduces control messages overhead [14]. Below are major advantages and disadvantages of LAR protocol:

Advantages:

•
Restricted directional flooding to expected region instead of entire flooding.
•
Low control and memory overhead.
•
Resource consumption as well as overhead appear to be low due to exchange of minimum required control information.

Disadvantages:

•
Designed for large networks for nodes greater than 1000.
•
Requires GPS.

3.2 DREAM

In DREAM, update and control messages are distributed in a frequency relative to node mobility and distance effect. Thus, the stationary nodes do not have to send any update messages. Hence, routing overhead is further reduced in DREAM [15]. Below are some major advantages and disadvantages of DREAM protocol:

Advantages:

•
Restricted directional flooding to expected region instead of entire flooding.
•
Low control overhead and Memory Overhead.
•
Energy consumption medium.

Disadvantages:

•
Requires GPS.
•
Suitable for medium network size.

3.3 AODV

AODV is a proactive routing protocol. It consists of two mechanisms; route discovery and route maintenance. Route discovery incorporates two sub-phases called route request and route reply [2]. If an established route is broken, the sender is notified, hence the sender initiates another route discovery process to establish a route. AODV incurs less memory overhead and establishes unicast routes between source and destination nodes and thus, the network utilization tends to be minimal. In AODV routes are built on demand, hence routing traffic is minimal and overall overhead is reduced.

AODV is capable of establishing optimal routes between communicating nodes in the network on the basis of sequence numbers [3]. For route discovery, AODV utilizes three different control messages such as Route Requests (RREQ), Route Errors (RERRs) and Route Replies (RREPs) [13]. AODV messages are incorporated in UDP packets to find out shortest path between the source and destination nodes and sustain route subsequently. Below are some major advantages and disadvantages of AODV protocol.

Advantages:

•
Routes are created on demand. Destination sequence numbers can be used to find the latest route to the destination.
•
Delay for connection setup is minimized.
•
GPS is not required.

Disadvantage:

•
Route maintenance is carried out between each pair of hosts on a regular basis during the entire lifetime of the application.
•
Multiple RREP packets may be received by the source in response to a single RREQ packet that may consume higher control overhead.
•
Periodic beaconing leads to unnecessary bandwidth consumption.

3.4 ZRP

Zone Routing Protocol (ZRP) is a hybrid MANET routing protocol combining advantages of both proactive and reactive routing protocols [21]. In ZRP, every node proactively create routes to neighborhood nodes, within a hop distance that lies within the range of the radius of the chosen zone. It is based on partitioning the network into various overlapping zones of different sizes. Nodes located within the zone are considered interior nodes, nodes residing on the boundary of the zone are referred as peripheral nodes and all other nodes considered as exterior nodes.

A source node requiring to send data to a specific destination node has to initiate the route discovery process. If the destination node exists within the routing zone of the source node, the route must have been already available and routing process can be accomplished by asset of intra zone phase and otherwise the packet needs to be sent to peripheral nodes via border casting. Below are some major advantages and disadvantages of ZRP protocol:

Advantages:

•
Designed especially for large networks (i.e. nodes greater than 1000)
•
It represents an integration between reactive and proactive routing protocols.

Disadvantages:

•
The increased separation between source and destination nodes, results in zone increase consequently, hence more number of border casts are required to find border zone bandwidth.
•
Selection of zone radius is considered a performance challenge.

4. Performance evaluation metrics

MANET routing protocols are evaluated taking into consideration different network settings and based on a set of performance metrics. This research work, considers four different performance metrics namely; throughput, packet delivery ratio (PDR), end-to-end delay as well as energy consumption.

4.1 Throughput

Throughput is the number of packets received at the destination during a specific simulation time [19]. It is expressed in terms of packets per second or bits per second as well. MANET throughput is significantly affected by the factors like limited bandwidth, limited battery power, and repeatedly changing topology [11]. Throughput of a MANET can be calculated as per the equation that follows.

$\displaystyle\!\!\!\!\!\!\!\!\textit{Throughput}=\left(\frac{\textit{Number of% Delivered Packets}}{\textit{Simulation Time}}\right)$ (1)

Table 1

Simulation scenarios parameters

Parameters	Scenario 1	Scenario 2	Scenario 3
Network size (m $\times$ m)	500 $\times$ 500–1750 $\times$ 1750	Best case, worst case	Best case, worst case
Network density ( $n$ )	30	30–150	Best case, worst case
Mobility (m/s)	20	20	5–40

4.2 Packet Delivery Ratio (PDR)

PDR is the ratio between packets successfully received by the destination to the total number of packets sent by source node Tarunpreet and Anil [19]. PDR can be expressed as:

$\displaystyle\!\!\!\!\!\!\!\!\!\!\!\!\textit{PDR}\!=\!\left(\!\frac{\textit{% Packets Received at Destination}}{\textit{Packets Sent By Source}}\!\right)*100\%$ (2)

4.3 End to end delay

End-to-End delay (E2E Delay) refers to the average amount of time taken by each packet to travel from the source to destination nodes [19]. It is expressed in seconds. It includes several components such as; delay at the MAC layer while preparing packet for transmission, transmission delay, waiting time at intermediate nodes/routers. Hence, E2E Delay depends on network dynamics, weak signal strength on the communication link and link breakages. E2E delay can be determined using the following mathematical expression:

$\displaystyle\textit{E2E Delay}=N(\textit{Transmission Delay}+$

(3) $\displaystyle\quad\textit{Propagation Delay}+\textit{Processing Delay})$

Where $N$ is the total number of transmissions.
4.4 Energy consumption

In order to enhance the network life time, it is important to reduce energy consumption at each node.

MANET non-fixed infrastructure incurs more overhead on each node, which results in more consumption of transmitted power and reduced network life time. Node activities such as receiving, sending, forwarding and mobility, consume most of nodes’ battery power (Ahvar, 2007). Average energy consumption can be determined as follows:

$\displaystyle\!\!\!\!\!\!\!\!\!\textit{\@setsize{\small}{11pt}{\ixpt}{\@ixpt}% AvgEngConsumption}=\frac{(\textit{\@setsize{\small}{11pt}{\ixpt}{\@ixpt}% InitialEnergy}-\textit{\@setsize{\small}{11pt}{\ixpt}{\@ixpt}EndEnergy})}{N}$ (4)

Where $N$ is the total number of nodes in the network.

5. Simulation and result analysis

A set of performance and environmental parameters and settings were defined for the simulation process. Constant Bit Rate (CBR) traffic model was implemented at the data rate of 5 packets per second. Transmission range was set to 250 m and two source-destination pairs of nodes were chosen. In addition, MAC layer protocol was IEEE 802.11. Size of beacon packet size was set to 64 bytes. The queue size, network bandwidth and data packet size were initialized to be 50 packets, 2 Mbps and 512 bytes respectively.

Three different scenarios were conducted implementing a hierarchal evaluation model allowing to assess selected MANET protocols in several network conditions representing best to worst situations:

1.
The first scenario utilizes six different network spaces ranging from 500 m $\times$ 500 m to 1750 m $\times$ 1750 m.
2.
The second scenario used seven different groups of nodes density, with their numbers ranging from 30 to 150 nodes. This scenario considered results achieved from worst case network size scenario.
3.
The third scenario focused on varying node mobility speeds, were speed range from 5 to 40 meters per second. This scenario considered results achieved from first and second scenario.

Table 1 summarizes all three scenarios Parameters with a focus on network size, network density and mobility.

Table 2 shows the default network parameters and summarizes the parameters used in these scenarios.

Table 2
Overall network parameters

Network simulation Value

parameters

Map size 500 $\times$ 500, 750 $\times$ 750, 1000 $\times$ 1000,

1250 $\times$ 1250, 1500 $\times$ 1500,

1750 $\times$ 1750 m

Number of nodes 30, 50, 70, 90, 110, 130, 150 nodes

Mobility speed 5, 10, 20, 30, 40 m/s

Simulation time (sec) 150 s/effective (2–148)

Mobility model Random way-point

Traffic type Constant Bit Rate (CBR)

Radio range 250 meter

MAC layer type IEEE 802.11

Bandwidth of links 2 Mbps

Number of connections First node $\rightarrow$ final node

Packet size 512 Byte

Pause time 2 second

Transmission range 250 m

Table 3
Performance metrics vs scenario parameters

Parameters Throughput PDR E2E Delay Energy consumption

Network size (increase) Decrease Decrease Increase Decrease

Density (increase) Relative Relative Increase Increase

Mobility speed (increase) Decrease Decrease Increase Increase

Figure 1.
Average packet delivery ratio during scenario 1.

Figure 2.
Average throughput during scenario 1.

Figure 3.
Average E2E delay during scenario 1.

5.1 Performance metrics relationship vs scenarios

Parameters	Throughput	PDR	E2E Delay	Energy consumption
Network size (increase)	Decrease	Decrease	Increase	Decrease
Density (increase)	Relative	Relative	Increase	Increase
Mobility speed (increase)	Decrease	Decrease	Increase	Increase

Increasing network size, result in decreasing throughput, for the reason that when network size grows, number of nodes in the network increases leading to higher mobility and more frequent topology changes. At the same time, high node mobility and more frequent topology change causes link breakages and leads to generate more control messages, which in turn result in increased end-to-end delay, reduced PDR and high congestion rate. In the same concern, increased node’s mobility and density consumes more energy reducing network lifetime. However, if network size increase reducing the connectivity of nodes, this will reduce nodes power consumption. Nodes consume their energy during connectivity sensing and packets transmission. Table 3 summarizes the relationship between suggested scenario and performance metrics.

Figure 4.

Average energy consumption during scenario 1.

5.2 Result analysis discussion

This section presents results of executed 10 trials of each simulation scenario. Figure 1 shows throughput results for AODV, DREAM, LAR, and ZRP in scenario 1. As it is shown, throughput of all four routing protocols tend to decrease with increasing number of nodes and network grid size. As described before, increasing network grid size, provides more freedom for nodes to move which results in frequent topology changes. In this situation, it is more difficult for finding a route to destination regardless of selected routing protocol. Also, frequent topology changes cause several link breakdown and nodes to be unreachable, hence PDR is reduced.

The reason that position-based protocols LAR and DREAM demonstrate better throughput performance as compared to AODV and ZRP protocols during smaller network is due to route discovery method being used. During routing LAR and DREAM, packets are sent to the neighboring node that appears to be nearest to the desired destination node.

Figure 5.

Average packet delivery ratio during scenario 2.

Figure 6.

Average throughput for during scenario 2.

As depicted in Fig. 1, it is clear that PDR for all protocols LAR, DREAM, AODV and ZRP tends to decrease with increase network size and in frequency of topology changes. However, among these four protocols, AODV appears to have achieved the best PDR performance and ZRP the lowest. Looking at the worst case grid size of 1750 $\times$ 1750 m, all four tested protocols achieved very limited PDR.

Figure 3 demonstrates average E2E delay for all four protocols under investigation AODV, LAR, DREAM, and ZRP. It can be noticed that E2E delay grows uniformly along with the grid size. However, E2E delay of LAR protocol remains significantly less affected by network size compared to AODV and DREAM protocols, especially at largest grid size 1750 $\times$ 1750 m. This is because LAR incurred low overhead leading the network to be less congested as compared to AODV and ZRP.

Table 4

Routing protocols’ performance metrics comparison at worst network size

	LAR	DREAM	AODV	ZRP
Throughput	1	3	1	4
PDR	1	3	1	4
E2E	1	3	2	4
Energy consumption	1	2	3	4

Figure 7.

Average E2E delay during scenario 2.

Figure 8.

Average energy consumption during scenario 2.

Figure 4 demonstrates the average energy consumption for LAR, DREAM, AODV and ZRP protocols. Energy consumed by each sensor node during the entire network life time is due to both sensing and communication. For all protocols, the energy consumption slightly decreases as the network size grow because the sensing and communication between nodes becomes limited. It was measured that ZRP and AODV consume more energy compared to other protocols LAR and DREAM. The reduced energy consumption at position based routing protocols is due to less connectivity sensing and control packets overhead.

The summary of scenario 1 evaluation shows that all routing protocols demonstrate best performance as network size is small, by increasing the network size performance of all protocols start decreasing, where the worst results are achieved in large-scale network size. Table 4 shows that LAR protocol achieves best accumulative performance considering worst-case network size (1750 $\times$ 1750 m). A scale between (1 best to 4 worst) is used to describe the level of achieved performance result.

5.3 Evaluating network density effect under worst case network size (1750

\times

1750 m)

In the second simulation scenario different groups of node density were used ranging from 30 to 150 nodes and largest network grid size (1750 $\times$ 1750 m) was used. Figures 5 and 6 show the average throughput and PDR for selected protocols through different network density ranges. It was noted that the average throughput linearly increased from 70 Kbps to 250 Kbps and PDR linearly increased from 5% to 20% as network density increases. Because topology became quite dense and the network connectivity is high, the distance became close which leads to decrease of packet dropping.

Table 5
Routing protocols’ performance metrics comparison during scenario 2

	LAR	DREAM	AODV	ZRP
Throughput	1	1	3	4
PDR	1	1	3	4
E2E	1	3	2	4
Energy consumption	1	2	3	4

Figure 9.

Average packet delivery ratio during scenario 3.

Figure 10.

Average throughput during scenario 3.

For the same reason, average E2E delay was decreased as network density increased, this was demonstrated in Fig. 7. However, it’s clear from Fig. 8 average energy consumption linearly increased from 10% to 40% as the node density increased. This is because sensing and connectivity overhead between nodes became higher. ZRP had the worst energy consumption performance.

Table 5 summarizes accumulative performance results for the second scenario considering largest network scale and different node densities. During scenario 2, it was obvious that LAR and DREAM routing protocols performed the best comparing to remaining routing protocols. This could be attributed to the low overhead of LAR and DREAM which causes the network to be less congested as compared to AODV and ZRP. Table 5 provides a simple performance comparison between selected protocols in scenario 2, in which worst case network size and different network densities were considered.

Figure 11.

The average E2E Delay during scenario 3.

Figure 12.

The average energy consumption during scenario 3.

5.4 Evaluating node mobility effect under worst case network size and node density

In the third simulation scenario, nodes mobility effect is measured under largest network size (1750 $\times$ 1750 m) and using highest node density (150 nodes), representing worst cases from scenario 1 and scenario 2 respectively. Range of mobility speeds between 5 to 40 m/s. Figures 9–12 present average result of PDR, throughput, E2E delay and energy consumption according to the mobility speed changes. DREAM and AODV have best PDR and throughput performance during moderate mobility speed, however, when speed reaches 40 m/s, LAR and ZRP perform the best performance. ZRP had the worst average E2E delay during this scenario. While all protocols have shown a moderate consumption of energy in Fig. 12, DREAM shows the best power consumption at highest mobility speeds.

Table 6 summarizes performance metrics results for all routing protocols considering various mobility speeds as described in scenario 3. LAR routing protocol demonstrate best performance in large-scale and high dense networks and during high mobility speed. LAR PDR increased from 17% to 26% also Throughput increased from 155 Kbps to 255 Kbps. Protocols such as DREAM and AODV demonstrated good performance, however at low mobility speed.

Table 6
Routing protocols’ performance metrics comparison during scenario 3

	LAR	DREAM	AODV	ZRP
Throughput	1	4	2	2
PDR	1	4	3	2
E2E	1	1	1	2
Energy consumption	2	3	1	4

After evaluating selected routing protocols in all three scenarios representing different MANET network conditions, LAR routing protocol demonstrated best performance results in terms of PDR, throughput, and E2E delay.

6. Conclusion and future recommendations

This work investigates a set of MANETs routing protocols utilizing a hierarchal evaluation framework taking into consideration methods of operation, drawbacks and performance measures. Four different MANET routing protocols LAR, DREAM (position-based) and AODV and ZRP (topology-based) were selected for investigation. The analysis was based on performance parameters including PDR, Throughput, E2E Delay, and average energy consumption. Three main simulation scenarios were used, the first one focused on the network scale size, the second scenario considered the worst case setting of scenario 1 with varying node density. The last scenario utilized worst case results of scenarios 1 and 2, using varying node mobility speeds. The evaluation method focused on assessing the performance of these routing protocols within different integrated and rigorous network environments. Also, it was accurately used to identify the relationship between selected protocols features and achieved performance. The simulation scenarios demonstrated harsh and unpredictable MANET environment for more valid results and analysis. Results justify efficiency of position-based protocols such as LAR and DREAM in more dynamic and dense MANET environments. LAR showed best average performance in all matrices within high node speed and increased number of nodes and network size. In terms of energy consumption, DREAM demonstrated best average energy consumption in all three scenarios.

In conclusion, to achieve optimal QoS measures, future protocols should ensure scalability, reliability and adaptation to environmental changes and network contexts. Based on position-based routing protocols achieving best performance, the following features should be considered in the implementation of future position-based routing protocols:

•

Adaptive and scalable to MANET environmental conditions and contextual parameters. This includes network size, topology changes, density, mobility and node energy level. An integration between LAR and DREAM protocol features, facilitates the interchange between protocol advantages and allows the successful adaptation.

•

Reliable and optimal route construction considering a set of quality metrics such as; lower energy consumption, highest connectivity, least congestion and packet loss.

References

Husain

Shringar

Kumar

and Amit

, Performance comparison of topology and position based routing protocols in vehicular network environments, International Journal of Wireless and Mobile Networks (IJWMN) (2011), 20–32.

Perkins

Belding-Royer

and Das

, Ad hoc on-demand distance vector (AODV) routing, RFC 3561, IETF Network Working Group Internet-Draft, (2003), 1–37.

Johnson

D.B.

and Maltz

D.A.

, The dynamic source routing protocol for multi-hop wireless ad hoc networks, Computer Science Department Carnegie Mellon University, 2013, pp. 20–32.

Johnson

D.B.

Maltz

D.A.

and Broch

, Dynamic source routing in Ad-hoc wireless network, Computer Science Department, Carnegie Mellon University, 2011, pp. 153–181.

Wadhwa

Deepika Kochher

and Kumar

, A review of comparison of geographic routing protocols in mobile adhoc network, Rajasthan, India: Research India Publications, 2014, pp. 51–58.

Scott

DaDilva

and Davis

, Mobile ad hoc network routing protocols methodologies and applications, Department of Computer Engineering, 2014, pp. 20–25.

Ehsan

and Zartash

A.U.

, Performance comparison of ad hoc wireless network routing protocols, in: Multitopic Conference, Proceedings of INMIC, 8th International, IEEE, 2004.

Broustis

Jakllari

Repantis

and Molle

, A comprehensive comparison of routing protocols for large-scale wireless MANETs, in: 3rd Annual IEEE Communications Society on Sensor and Ad Hoc Communications and Networks (SECON ’06), 2006, pp. 951–956.

Jörg

D.O.

, Performance comparison of MANET routing protocols in different network sizes, in: Computer Science Project, Institute of Computer Science and Applied Mathematics Computer Networks and Distributed Systems (RVS), 2013, pp. 53–71.

10.

Koorush

and Talooki

V.N.

, Performance comparison of routing protocols for mobile ad-hoc networks, in: Asia-Pacific Conference on Communications, APCC, 2006, pp. 1–5.

11.

Girons

and Quesada

Y.J.

, A threshold-based hybrid routing protocol for MANET, Department of Telematics, Norwegian University of Science and Technology, 2009, pp. 80–96.

12.

Alnabhan

Alshuqran

Hammad

and Alnawayseh

, Performance evaluation of unicast routing protocols in MANETs – current state and future prospects, iJIM, http://www.i-jim.org, 2017, pp. 27–35.

13.

Mandal

and Mirsa

, Performance comparison of AODV/DSR on demand routing protocols for ad hoc networks in constrained situation, in: ICPWC International Conference, IEEE, 2005, pp. 86–89.

14.

Neelesh

and Roopam

, Optimization of performance metrics of lar in ad-hoc network, International Journal of Wireless and Mobile Networks (IJWMN) (2012), 71–85.

15.

Basagni

and Chlaa

, Distance routing effect algorithm for mobility (DREAM), in: Proceedings of the ACM/IEEE International Conference on Mobile Computing and Networking (Mobicom), 1998, pp. 76–84.

16.

www.isi.edu/nsnam. Retrieved March 23, 2017, from Network simulator-ns-2.

17.

Sonam

and Sahu

, Topology vs position based routing protocols in mobile ad hoc networks: A survey, International Journal of Engineering Research and Technology (IJERT) (2012), 301–315.

18.

Camp

Boleng

Williams

Wilcox

, and Navidi

, Performance comparison of two location based routing protocols for ad hoc networks, in: INFOCOM, 2002, pp. 1678–1687.

19.

Tarunpreet

and Anil

, QoS comparison of MANET routing protocols, I J Computer Network and Information Security (2015), 64–73.

20.

Young

and Ko

N.H.

, Wireless Networks, 2010, pp. 307–322.

21.

Haas

Pearlman

and Samar

, The zone routing protocol (ZRP) for ad-hoc networks, in: Internet Draft, draft-ietf-manet-zonezrp-04.txt, 2002.

22.

Bujari

Palazzi

C.E.

and Ronzani

, A comparison of stateless position-based packet routing algorithms for FANETs, IEEE Transactions on Mobile Computing (1) (2018), 1–1.

23.

Smiri

Boushaba

Abbou

R.B.

and Zahi

, Geographic and topology based routing protocols in vehicular ad-hoc networks: Performance evaluation and QoS analysis, in: Intelligent Systems and Computer Vision (ISCV), 2018 International Conference on, IEEE, 2018 April, pp. 1–8.

Network simulation	Value
parameters
Map size	500 $\times$ 500, 750 $\times$ 750, 1000 $\times$ 1000,
	1250 $\times$ 1250, 1500 $\times$ 1500,
	1750 $\times$ 1750 m
Number of nodes	30, 50, 70, 90, 110, 130, 150 nodes
Mobility speed	5, 10, 20, 30, 40 m/s
Simulation time (sec)	150 s/effective (2–148)
Mobility model	Random way-point
Traffic type	Constant Bit Rate (CBR)
Radio range	250 meter
MAC layer type	IEEE 802.11
Bandwidth of links	2 Mbps
Number of connections	First node $\rightarrow$ final node
Packet size	512 Byte
Pause time	2 second
Transmission range	250 m

Performance analysis and enhancement of position-based routing protocols in MANETS

Abstract

Keywords

1. Introduction

3. Selected routing protocols

• Restricted directional flooding to expected region instead of entire flooding. • Low control overhead and Memory Overhead. • Energy consumption medium. Disadvantages: • Requires GPS. • Suitable for medium network size. 3.3 AODV

4.1 Throughput

(3) Propagation Delay + Processing Delay ) Where N is the total number of transmissions. 4.4 Energy consumption

Table 5 Routing protocols’ performance metrics comparison during scenario 2

Table 6 Routing protocols’ performance metrics comparison during scenario 3

References

•
Restricted directional flooding to expected region instead of entire flooding.
•
Low control overhead and Memory Overhead.
•
Energy consumption medium.

Disadvantages:

•
Requires GPS.
•
Suitable for medium network size.

3.3 AODV

(3) $\displaystyle\quad\textit{Propagation Delay}+\textit{Processing Delay})$

Where $N$ is the total number of transmissions.
4.4 Energy consumption

Table 5
Routing protocols’ performance metrics comparison during scenario 2

Table 6
Routing protocols’ performance metrics comparison during scenario 3