DI-AODV Model with Sustainable Data Management Scheme to Improve Performance and Data Accessibility in VANET

doi:10.21203/rs.3.rs-5327559/v1

Cooperation and coordination between mobile nodes (vehicles) and static nodes (roadside infrastructure) are crucial in vehicular ad hoc networks (VANETs). Rapid topological modifications and the perception and transmission of physical environment parameters are the main problems of VANETs. Effective routing is a difficult problem in VANETs because to the vehicles' selfishness and the absence of a central coordinator. Effective routing requires optimal resource allocation, and this work suggested the DI-AODV model to boost VANET performance. In DI-AODV model, AODV protocol is modified by using Dijkstra Shortest Path Algorithm. Then, Data Replication method which is Sustainable Data Management Scheme is provided to ensure data delivery. The experimental result shows the proposed DI-AODV model provides better performance compared with other existing protocols in VANET.

Systems and Networking

AODV

Data Replication

DI-AODV

DSDV

Dijkstra's Shortest Path algorithm

GPSR

VANET

VANET is a type of wireless multi-hop network that is constrained by the need for rapid topology changes due to high node mobility. Inter-vehicle communication is becoming a viable topic of research, standardisation, and growth as the number of vehicles equipped with computing technology and wireless communication devices grows. Routing protocols play an important role in network communication. Any routing protocol's principal purpose is to identify the best way to communicate between nodes (vehicles).

Because of the high mobility of the vehicles, the topology of the VANET is continually changing. The majority of VANET vehicles these days are capable of merging their own system with other existing technologies such as GPS (J. Costa et al., 2017).

In today's computing world VANETs communication provides plentiful applications which are primarily divided into different categories such as Safety-related applications, value-added applications, and comfort-related applications (Mahajan & Patil, 2022).

Getting focus on Safety-related applications there are times when emergency warnings, stopped vehicle warnings, lane shifting warnings, road conditions warnings, and other alerts must be sent to vehicles on the same route. But, there are chances of failure of networks due to highly dynamic nature of the VANET. Here, the purpose of this work is to check the availability of data, efficiency of network and test the performance of mobile nodes across this dynamic vehicular environment (M. R. Ghori et al., 2018).

The VANET model's device and communication link characteristics vary greatly, resulting in a high degree of heterogeneity. Many issues face applications intended for such a heterogeneous sort of network, including frequent network partition, communication link failure, data node crash, and fluctuations in network connection performance, among others. Data replication is a technique for increasing data availability in the event of problems such as those stated above. In networks like mobile ad hoc networks and vehicular ad hoc networks, network division occurs owing to node mobility. The data maintained by the other node may not be accessed during a network partition. Data accessibility can be improved by replicating that data on several other nodes (P. J. Kumar & Ilango, 2017).

Sudesh Kumar et. al. (S. Kumar et al., 2022) mentioned that AODV protocol gives better performance in terms of throughput and packet delivery ratio metrics than DSDV. But, in sparse region, DSDV performs better than AODV.

Amit Choksi et. al. (Choksi & Shah, 2024) introduced machine learning-assisted distance and residual energy-based clustering algorithm. This research shows that SOMNNDP performs exceptionally well in terms of QoS and consumes less energy when compared to other multi-hop cluster routing algorithms such as KMDP and FCMDP.

G. P. Kour Marwah et. al. (Marwah et al., 2022) introduced H-WDFOA-VANET approach which hybridizes the notion of rider integrated cuckoo search, DCM and SBACO. The suggested approach minimized the drop value to 0.15504, delay time to 15.64318s.

Mahmood et. al. (Sohail et al., 2023) proposed BBSF technique which uses blockchain-based Hyperledger Sawtooth framework to secure data. When packets are delivered to the correct destination node, BBSF offers a low network latency, low hop counts, a minimal network overhead time, and good packet delivery ratio.

Seliem et. al. (Seliem et al., 2019) proposed an analytical solution for the end-to-end delay probability distribution based on the closed-form formulation of the re-healing delay in one-way VANET traffic and the PMF of the number of gaps from source to destination. The service provider can quickly establish a bound on the worst-case end-to-end delay for a Vehicular Ad hoc Network by using a closed-form expression for the lower bound of the end-to-end delay probability density distribution. A closed-form expression for the upper bound of the probability distribution of the end-to-end delay is established when the distance between the source and destination in a VANET is restricted.

Paranjothi et. al. (A. Paranjothi et al., 2018) highlighted the consideration of obstacle shadowing region during message dissemination distinguishes Hybrid-Vehcloud from prior systems such as Cloud-assisted Message Downlink Dissemination Scheme (CMDS), Cross-Layer Broadcast Protocol (CLBP), and Cloud-VANET. In obstacle shadowing regions, the Hybrid-Vehcloud uses mobile gateways (such as buses) in the vehicular cloud to disseminate messages, whereas in non-obstacle shadowing regions, it uses a multi-hop approach. This network simulation ensures that messages are delivered on time.

Costa et. al. (J. Costa et al., 2017) proposed the DDBC protocol selects the best cars to serve as relay nodes by identifying each relay vehicle as a cutting vertex in each edge-induced subgraph G[E'u] using Tarjan's technique to represent the most central node. The DDBC protocol was presented as a way to distribute data with minimal overhead and delay while yet providing adequate coverage for non-critical services. In terms of diverse data distribution requirements, DDBC provides better delay and number of transmissions, but it does not provide improved coverage.

Kumar et. al. (P. J. Kumar & Ilango, 2017) depict the different replication methods used in MANET, VANET and Cloud Computing environment. Priority Competition split replication protocol improves the file access and its availability among peers in MANET. Community partition aware Replication for ASNET improves effective data sharing and information propagation among social network peers. Data accessibility among peers in social ad-hoc network can be increased further in future. Replication Aware Data dissemination using bloom filter to select candidate vehicle for replication increases availability of data among vehicles in VANET either in the sparse region or in the dense region. There is scope to decrease energy consumption. Load balancing using replication method in cloud computing decreases the queuing delay faced by the applications while executing it on cloud servers. The queuing delay can be reduced further by increasing the performance the system.

As claimed by authors- Janech et. al. (Janech et al., 2012) data replication algorithms like SPA (Simple Pulling Algorithm), DRA (Dependent Replication Algorithm) and IRA (Independent Replication Algorithm) give satisfactory results in DDBS VANET environment. In comparison to other algorithms, the IRA algorithm achieves the highest data actuality. In terms of network burden, the IRA algorithm produces poor results.

Because it does not make copies of OBU devices, the data actuality of the SPA method is the lowest, ranging from 3 to 10%. Because the OBU transmits information about its local database and also generates replicas, the DRA process requires more data to be delivered than the SPA technique. The veracity of the statistics varies between 7 and 14 percent. In comparison to other algorithms, the IRA algorithm takes the most data. The IRA algorithm generates the most replicas and distributes the most data. The veracity of the statistics varies between 10% and 21%.

According to Lieskovsky et. al. (A. Lieskovsky et al., 2011) RSU(Road-Side Unit) initiates the data replication process in DDBS in VANET. But, most data is distributed by OBU (On-Board Unit). OBU nodes actively replicate the data which causes improvement of data availability. For data distribution two strategies: the PUSH method and the PULL method were used.

Zhang et. al. (Y. Zhang & G. Cao, 2011) proposed a vehicle platooning methodology that identifies the platoon fast and predicts the split process. According to the authors, the two proposed cooperative replication data replication algorithms can decrease intra-platoon data redundancy while simultaneously lowering data access costs. V-PADA (best-location) and V-PADA (neighbouring) provide substantially higher data availability than the non-cooperative solution, the connection-based solution, and the distance-based solution, according to their simulation results.

The research proposed a DI-AODV model with Data Replication Approach to reduce average end to end delay time observed in the traditional AODV protocol used for VANET. In this approach, the real-world map for the city is given as an input in the form of network provided with the scenario templates along with setting for various parameters like area, speed of vehicles, number of accidents, etc. As per proposed model, the AODV nodes find their eight possible intermediate nodes from eight directions (N, E, W, S, NE, NW, SE, and SW).

Dijkstra Shortest Path method is applied to determine the shortest path for packet forwarding across all nodes. Dijkstra Shortest Path method determines the shortest path from a given source node to every other node. It may also be used to calculate the costs of shortest pathways from a single node to a single destination node by stopping the procedure once the shortest path to the target node has been identified. After finding the shortest path apply the efficient traffic pattern among the CBR, Pareto and Exponential traffic pattern.

Additionally, we have added the data replication method which will prevent the loss of the packets especially observed during an accident events. For this method the special data packet has to be generated referred as urgent data packet whose priority count determines the importance of the packet. As the data is replicated priority count increases also the number of message replicas grows.

DI-AODV with Data Replication Framework output is feed to the network simulator along with the simulation parameters in the form of configurations such as config-1, config-2, etc. to extract the results.

Algorithm 1. DI-AODV with Data Replication Method

Step-1. AODV Protocol Initialization for N number of nodes

Find 8 possible intermediate nodes (N, E, W, S, NE, NW, SE, SW) for transmission of message;

Set HopCount = no. of intermediate nodes;

Step-2. Apply CBRTraffic Pattern

Set packetInterval;

Set value for packetSize;

Step-3. Apply Dijkstra Shortest Path Algorithm

Find the closest node;

Find the potential distances which are compared to find the shortest path;

Step-4. Apply Data Replication Method

Set number of replicas to 0;

When packet is received from lower layer

If Packet Type is the Urgent Data

Then

Replicate the data;

Increase the priority counter;

If new route is created

Then

Replicate the data;

Increase the number of replicas;

Increase the priority counter;

Send replicated packet to Network Layer;

If route is deleted

Then

Delete replicated data;

When packet is received from upper layer

If Packet Type is the Urgent Data

Then

If there is replicated the data

Then

Decrease the number of replicas;

Decrease the priority counter;

Dijkstra's algorithm is a graph search algorithm that solves the single-source shortest path problem for a graph with nonnegative edge path costs, resulting in a shortest path tree. Routing and subroutines in other graph algorithms frequently employ this approach.

The algorithm finds the shortest path, or the path with the lowest cost, between a given source node and all other nodes. By halting the process once the shortest path to the target node has been found, it may also be used to determine the costs of shortest pathways from a single node to a single destination node. For example, if the graph's vertices represent nodes (vehicles) and edge path costs reflect the driving distances between pairs of nodes connected by a direct road, then Dijkstra's algorithm can be used to find the shortest path between one node and every other node.

After identifying the intermediate nodes, the shortest path is found using Dijkstra's Shortest Path algorithm, as shown in Algorithm 2.

Algorithm 2. Dijkstra Shortest Path Algorithm

While I <= number of nodes

closestNode = getClosestNode()

visited[closestNode] = True

While J <= number of nodes

If (!visited[j])

then potentialDistance = Distances[closestNode] + 1

If (potentialDistance < Distances[J])

then Distances[J] = potentialDistance

predecessors[J] = closestNode

//Here Distances[J] represents the current known distance from the source node to //the current node J being considered.

//And potentialDistance is the sum of the distance from the source node to the //current node and the distance from the current node to its neighboring node, //which is being considered as a potential candidate for the shortest path from the //source node to the neighboring node.

Procedure getClosestNode()

Set minDistance =

Set closestNode = -1

While i < numberOfNodes

If (!visited[i] && distances[i] < minDistance)

Then minDistance = distances[i]

closestNode = i

Data Replication method is used as a sustainable data management scheme. Data replication is a method of storing the same data in multiple locations to improve access and reduce the impact of intermittent node connectivity in vehicular ad hoc networks (VANETs). The topology of VANETs changes dynamically due to high vehicle mobility, which can lead to frequent disconnections.

Algorithm 3 gets invoked when there is vulnerable situation, it creates replicas to ensure data accessibility. The packets are forwarded according to their priority counter managed during data replication stage.

Algorithm 3. Data Replication Algorithm

Set numReplicas = 0

While packet is received from lower layer

If PACKET_TYPE = URGENT_DATA

Then replicate the data

numReplicas = 1

If new route is created

Then replicate the data

numReplicas++

send replicated packet to Network Layer

If route is deleted

Then delete replicated data

While packet is received from upper layer

If PACKET_TYPE = URGENT_DATA

Then If there is replicated the data

Then numReplicas--

This section presents the experimental outcomes and study of DI-AODV model against the existing popular routing protocols (DSDV, AODV and GPSR) in VANET. For experimental analyses, Random Waypoint Mobility based Map (Mumbai) used. This simulation run for 10, 20,30,40,50,100,150,200,250, and 300 vehicles in a 10000 m × 10000 m unit region for 2500s.

Table 1 shows the packet loss ratio (PLR) obtained in well-known VANET Protocols such as DSDV, AODV and GPSR (Mahajan & Patil, 2024) along with proposed DI-AODV model.

Table 1. Comparison of Various VANET Protocols with the proposed DI-AODV protocol model using Packet Loss Ratio (in percentage)

No. of Nodes	DSDV	GPSR	AODV	DI-AODV
10	3	0.125	0.3	0
20	5.286	0.375	0.509524	0.051706
30	6.8	1	0.68	0.059559
40	6.861	1.214286	0.891111	0.062305
50	10.676	2.25	1.089286	0.065531
100	10.893	3.2	1.143	0.095057
150	12.744	5.363636	1.275207	0.13089
200	16.72	6.210891	1.667633	0.164474
250	17.669	6.60367	1.766864	0.320513
300	22.338	6.698885	2.234572	0.416667

According to Table 1 and Graph 1, the DI-AODV provides the lowest Packet Loss Ratio (PLR) value when compared to other options, and its value never goes above 0.5% for any number of cars.

Table 2 demonstrates the packet delivery ratio (PDR) obtained in well-known VANET Protocols along with proposed DI-AODV model.

Table 2. Comparison of Various VANET Protocols with the proposed DI-AODV protocol model using Packet Delivery Ratio (in percentage)

No. of Nodes	DSDV	GPSR	AODV	DI-AODV
10	97	99.875	99.7	100
20	94.714	99.625	99.49048	99.94829
30	93.2	99	99.32	99.94044
40	93.139	98.78571	99.10889	99.9377
50	89.324	97.75	98.91071	99.93447
100	89.107	96.8	98.857	99.90494
150	87.256	94.63636	98.72479	99.86911
200	83.28	93.78911	98.33237	99.83553
250	82.331	93.39633	98.23314	99.67949
300	77.662	93.30112	97.76543	99.58333

As referred to Table 2 and Graph 2, we can say that the DI-AODV gives highest value of packet delivery ratio (PDR) as compared to others and its value also approximately 100% for all number of vehicles. DSDV gives the lowest PDR than others.

Table 3 gives the details of average end-to-end delay observed in well-known VANET Protocols along with proposed DI-AODV model.

Table 3. Comparison of Various VANET Protocols with the proposed DI-AODV protocol model using Average End-to-End Delay (in ms)

No. of Nodes	DSDV	GPSR	AODV	DI-AODV
10	81.56	36.125	32.625	4.166667
20	41.81	32.625	27.000	3.205128
30	15.17	11.84	26.500	2.941176
40	11.13	11.51	11.857	1.644737
50	9.24	10.375	11.143	1.308901
100	7.46	9.29368	10.778	0.95057
150	4.81	8.25	8.895	0.655308
200	3.49	7.70297	5.457	0.623053
250	3.41	7.449541	3.967	0.595593
300	3.28	6.61157	3.814	0.517063

As per Table 3 the minimum end-to-end delay observed in DI-AODV than others potentially less than 4.2ms in all node scenarios which is more than most of the cases of other protocols.

DSDV protocol gives largest value of end-to-end delay for number of vehicles less than 50.

Table 4 gives the average throughput given by well-known VANET Protocols along with proposed DI-AODV model.

Table 4. Comparison of Various VANET Protocols with the proposed DI-AODV protocol model using Average Throughput (in Mbps)

No. of Nodes	DSDV	GPSR	AODV	DI-AODV
10	8.302	3.062201	16.920	133.84
20	8.000	2.452107	33.000	174.16
30	7.704	2.162162	90.912	190.4
40	7.518	2.077922	149.201	339.92
50	3.692	1.810843	123.981	427.28
100	3.267	1.62	184.817	588.56
150	3.216	0.9	286.868	854
200	2.622	0.622622	394.733	898.24
250	2.391	0.394089	403.523	939.68
300	2.337	0.256	419.721	1082.48

Table 4 and Graph 4 depicts that DI-AODV produces high value of throughput as compared to other protocols, DSDV, AODV, GPSR.

This research proposed a DI-AODV model to improve performance and data accessibility in VANET using Sustainable Data Management Scheme such as Data Replication. In this paper, DI-AODV model, using Dijkstra’s Shortest Path Algorithm, Data Replication method implemented.

The experimental outcomes verified that DI-AODV model provides assured performance improvement giving low packet loss ratio, high packet delivery ratio, low average end-to-end delay and high average throughput when data transfer in VANET as compared to other well-known VANET protocols, AODV, GPSR and DSDV.

Data Availability Statement

Authors agree to make data and materials supporting the results or analyses presented in their paper available upon reasonable request. It is up to the author to determine whether a request is reasonable.

A. Lieskovsky, J. Janech, & T. Baca. (2011). Data replication in distributed database systems in VANET environment. 2011 IEEE 2nd International Conference on Software Engineering and Service Science, 304–307. https://doi.org/10.1109/ICSESS.2011.5982314
A. Paranjothi, M. S. Khan, & M. Atiquzzaman. (2018). Hybrid-Vehcloud: An Obstacle Shadowing Approach for VANETs in Urban Environment. 2018 IEEE 88th Vehicular Technology Conference (VTC-Fall), 1–5. https://doi.org/10.1109/VTCFall.2018.8690729
Choksi, A., & Shah, M. (2024). Machine Learning-assisted Distance Based Residual Energy Aware Clustering Algorithm for Energy Efficient Information Dissemination in Urban VANETs. INTERNATIONAL JOURNAL OF NEXT-GENERATION COMPUTING, 15, 1–20. https://doi.org/10.47164/ijngc.v15i1.1472
J. Costa, W. Lobato, A. M. de Souza, D. Rosário, L. A. Villas, & E. Cerqueira. (2017). Centrality-based data dissemination protocol for vehicular ad hoc networks. 2017 IEEE 16th International Symposium on Network Computing and Applications (NCA), 1–4. https://doi.org/10.1109/NCA.2017.8171333
Janech, J., Lieskovsky, A., & Krsak, E. (2012). Comparation of Strategies for Data Replication in VANET Environment. https://doi.org/10.1109/WAINA.2012.179
Kumar, P. J., & Ilango, P. (2017). Data Replication in Current Generation Computing Environment. Seventh Sense Research Group, 45(10), 488–492. https://doi.org/10.14445/22315381/IJETT-V45P292
Kumar, S., Sharma, S., Khan, S., Singh, P., & seema. (2022). SIMULATION-BASED PERFORMANCE EVALUATION OF VANET ROUTING PROTOCOLS UNDER INDIAN TRAFFIC SCENARIOS. ICIC Express Letters, 67–74. https://doi.org/10.24507/icicel.16.01.67
M. R. Ghori, K. Z. Zamli, N. Quosthoni, M. Hisyam, & M. Montaser. (2018). Vehicular ad-hoc network (VANET): Review. 2018 IEEE International Conference on Innovative Research and Development (ICIRD), 1–6. https://doi.org/10.1109/ICIRD.2018.8376311
Mahajan, P., & Patil, G. (2022). A BRIEF GLANCE TOWARDS THE SIMULATION IN THEVEHICULAR ENVIRONMENT. 2320–2882.
Mahajan, P., & Patil, G. (2024). COMPARATIVE PERFORMANCE ANALYSIS OF AODV, DSDV, AND GPSR ROUTING PROTOCOLS IN VEHICULAR AD-HOC NETWORKS. Journal of Nonlinear Analysis and Optimization, 15, 70–80.
Marwah, G., Jain, A., Malik, P., Goraya, D.-M., Tanwar, S., Ovidiu Calin, S., Mihaltan, T., Sharma, R., & Alkhayyat, A. (2022). An Improved Machine Learning Model with Hybrid Technique in VANET for Robust Communication. Mathematics, 10, 4030. https://doi.org/10.3390/math10214030
Seliem, H., Shahidi, R., Ahmed, M. H., & Shehata, M. (2019). On the End-to-End Delay in a One-Way VANET. IEEE Transactions on Vehicular Technology, PP, 1–1. https://doi.org/10.1109/TVT.2019.2916936
Sohail, H., Mscs, M., Elmagzoub, M., Rajab, A., Rajab, K., Ahmed, A., Shaikh, A., Ali, A., Jamil, H., Akhtar, N., & Khan. (2023). BBSF: Blockchain-Based Secure Weather Forecasting Information through Routing Protocol in Vanet. 23, 5259.
Y. Zhang & G. Cao. (2011). V-PADA: Vehicle-Platoon-Aware Data Access in VANETs. IEEE Transactions on Vehicular Technology, 60(5), 2326–2339. https://doi.org/10.1109/TVT.2011.2148202

Graphs 1 to 4 are available in the Supplementary Files section

The authors declare no competing interests.

Graphs.docx

DI-AODV Model with Sustainable Data Management Scheme to Improve Performance and Data Accessibility in VANET

Status:

Version 1

Abstract

Figures

1. Introduction

2. Related Work

3. DI-AODV Model with Sustainable Data Management Scheme

4. Experimental Results

5. Conclusion

Declarations

References

Graphs

Additional Declarations

Supplementary Files

Status:

Version 1