Performance Analysis of Stationary and Deterministic AODV Model

kothaig@srmist.edu.in Abstract —Vehicular Adhoc Network (VANET) is an emerging technology that provides a digital communication among vehicles, persons and Road-Side Units (RSU). VANETs are highly vulnerable to cyber-attacks. These cyber-attacks make a wrong illusion on traffic jam, can inject false information regarding traffics and injects large amount of spam messages that disrupts the normal functionalities. The main objective of the research work is to implement and analyze the different models that help in improving the traffic management. The scenarios are simulated, and the performance is analyzed using the OMNET++


AODV and Mobility Models
The protocols, agents and other models are provided by the open source model library called INET framework for OMNET++ simulation environment [R11]. The Adhoc On-Demand Distance Vector (AODV) routing protocol is one of the reactive protocols where routes are created on demand [R9]. Instead of maintaining up-to-date route information, the AODV routing protocol broadcasts the messages over the network for route discovery process [R3]. This model undergoes Route Discovery and Route Maintenance. Route Discovery is done through Route REQUEST (RREQ) and Route REPLY (RREP) and Route Maintenance is done through Route ERROR (RERR).
When a source node wants to send a message to the destination node at first it checks in the Route When the route request message is received multiple times, the duplicate copies will be discarded, by comparing the Broadcast ID and Source ID pairs.
The previous nodes broadcast ID is stored when RREQ is forwarded. If the Route Reply is not received before time expires, the entry will be deleted [R4]. All the nodes in the network will monitor the neighbourhood nodes and if any active route gets lost then it will send the Route Error Message (RERR) to all other nodes in the network

Analysis of Results
The simulation output of the AODV Stationary model and AODV Mobility model for fixed number of nodes as 20. The experiment run for 12 seconds and the calculated values based on the node shut down and start up time is tabulated are tabulated.

Stationary model
The  network. If five nodes shutdown then total number of events is 1713. If the ten and fifteen nodes shutdown then total number of events are 1678 and 1538. In a network when there is increase in number of nodes shutdown then total number of events gets decreases in this model.

Deterministic models
Deterministic mobility models are models that help in describing the motion using non-random mathematical models. The various deterministic mobility models are Linear Mobility helps in moving the nodes with constant speed and acceleration in linear manner. If the nodes move around circular and rectangular that is parallel to XY plane with constant speed, then it is Circle Mobility and rectangle mobility model. If the node orients towards the position of any other mobility model, then it is facing mobility model [R8].
Linear mobility model: The above Table II shows that when the nodes in a network are in linear mobility model there are 8523 events occurs by 12 seconds even after one node shutdown in the network. If five nodes shutdown then total number of events is 6780. If the ten and fifteen nodes shutdown then total number of events are 6354 and 5589. In a network when there is increase in number of nodes shutdown then total number of events gradually gets decreases in this model.

Fig. 3. Linear Deterministic Mod
Circle mobility mode: From table III it is known that when the all the nodes in a network are in Circle Mobility model the total number of events occurred at one node, five nodes, ten and fifteen nodes shutdown then total number of events are lesser when compared to Linear Deterministic Mobility model.  Time in (sec)

Combining deterministic models
No. of Events

Fig. 5. Linear and Circle Deterministic Mobility Model
In table IV, the linear and Circle Deterministic Models are combined, there are 12114 events occurred by 12 seconds even after one node shutdown from the network. If five nodes, ten and fifteen nodes shutdown then total number of events are 8452, 7154 and 5273. By combining two Deterministic Models there is dramatic increase in number of events than the linear and circle models. In table V, the Stationary and Deterministic Models are combined, there are 9559 events occurred by 12 seconds even after one node shutdown from the network. The source and destination node are in stationary model and all other eighteen nodes are in Circular Deterministic Models. If five nodes, ten and fifteen nodes shutdown then total number of events are 15721, 14594 and 6704. By combining two Different AODV Models there is dramatic increase in number of events during five and ten nodes shutdown than one node shutdown. But there is decrease in number of events for fifteen nodes shutdown when compared to one node shut down.

No. of Events
Time in (sec)

Conclusion
In this paper the analysis of performance of Stationary model and Deterministic Mobility models are done using AODV reactive routing protocol. The simulation is carried out in OMNET++ Simulator and the results are presented. After the analysis, we observed that combining Stationary and Deterministic model gives 20% higher performance when compared to other models in terms of number of events and time for more than one node shutdown. By combining two Deterministic Models gives 10% higher performance when compared to other models in terms of number of events and time for one node shutdown. The work can be extended for detecting the black hole attack by finding the average end to end delay, packet delivery ratio and throughput.