Academia.eduAcademia.edu

Figure 6 – uploaded by Subodh Shah

See full PDFdownloadDownload figure
Figure 7 -CAD-HOC generated highway scenario Figure 6 -CAD-HOC generated airport terminal

Figure 7 -CAD-HOC generated highway scenario Figure 6 -CAD-HOC generated airport terminal

Related Figures (15)

Figure 1 shows the normal process of conducting simulation in ns. The user has to imagine of a scenario, the number of nodes to be placed in the scenario, and then write the TCL scripts (tcl file) specifying the node configurations parameters and some other ns commands required to start and stop ns. The user has also to create the movement and connection files that together represent the scenario. These files can be generated using the cbrgen or setdest utilities, or the user can write TCL scripts to generate the movement and connection files (which is very tedious). Once these files are produced they are fed as input to ns, using which ns creates the scenario during simulation. The output of the simulation is a trace file (.tr), which is logged with each and every event that took place during the simulation. This file can than be used for obtaining measures such as mobility, throughput, end-to-end delay, and packet loss measurement. An optional output is the NAM supported file (.nam) that logs the necessary events to help visualize the scenario using the NAM. The NAM is a post simulation process that shows how the nodes moved and how they were connected during the simulation.
Figure 1 shows the normal process of conducting simulation in ns. The user has to imagine of a scenario, the number of nodes to be placed in the scenario, and then write the TCL scripts (tcl file) specifying the node configurations parameters and some other ns commands required to start and stop ns. The user has also to create the movement and connection files that together represent the scenario. These files can be generated using the cbrgen or setdest utilities, or the user can write TCL scripts to generate the movement and connection files (which is very tedious). Once these files are produced they are fed as input to ns, using which ns creates the scenario during simulation. The output of the simulation is a trace file (.tr), which is logged with each and every event that took place during the simulation. This file can than be used for obtaining measures such as mobility, throughput, end-to-end delay, and packet loss measurement. An optional output is the NAM supported file (.nam) that logs the necessary events to help visualize the scenario using the NAM. The NAM is a post simulation process that shows how the nodes moved and how they were connected during the simulation.
Figure 2 — The main screen of CAD-HOC
Figure 2 — The main screen of CAD-HOC
Movement (mobility) can be specified using either mobility models such as Brownian, Column, Pursue an Nomadic or as user-defined movements (curvature motion).
Movement (mobility) can be specified using either mobility models such as Brownian, Column, Pursue an Nomadic or as user-defined movements (curvature motion).
Table 1: Parameters used during realistic scenario simulation
Table 1: Parameters used during realistic scenario simulation
Table 2: Simulation results for airport terminal scenario
Table 2: Simulation results for airport terminal scenario
Table 3: Simulation results for the highway scenario 2.5.3. Conference Room Scenario
Table 3: Simulation results for the highway scenario 2.5.3. Conference Room Scenario
Table 4: Simulation results for conference room scenario The results show that there was huge loss of about 28% of the packets in the scenario. AODV and DSR had almost the same loss. DSDV showed the highest throughput and minimum delay, while AODV showed the lowest throughput and the maximum delay. This result depicts that DSR, a source initiated routing protocol, works better in this scenario. The results obtained in this simulation match the results obtained by Johannson [12], which showed that there was a 25.4% loss for DSDV, and losses of about 2% and 6% for DSR and AODV, respectively. This very much matches the patterns of results obtained in our simulation, thus proving the usefulness of using realistic scenarios, which can now be designed and generated effortlessly using CAD-HOC.
Table 4: Simulation results for conference room scenario The results show that there was huge loss of about 28% of the packets in the scenario. AODV and DSR had almost the same loss. DSDV showed the highest throughput and minimum delay, while AODV showed the lowest throughput and the maximum delay. This result depicts that DSR, a source initiated routing protocol, works better in this scenario. The results obtained in this simulation match the results obtained by Johannson [12], which showed that there was a 25.4% loss for DSDV, and losses of about 2% and 6% for DSR and AODV, respectively. This very much matches the patterns of results obtained in our simulation, thus proving the usefulness of using realistic scenarios, which can now be designed and generated effortlessly using CAD-HOC.
Graphs depicting throughput vs. number of nodes, for all three scenarios, are shown below. Throughput here represents the received throughput, a combination of receiving data and control packets. The AODV protocol shows higher throughput but it should be due to the more number of control packets being received. This is logical as the graphs for delay and percentage loss show that AODV performed worse than the other two protocols.
Graphs depicting throughput vs. number of nodes, for all three scenarios, are shown below. Throughput here represents the received throughput, a combination of receiving data and control packets. The AODV protocol shows higher throughput but it should be due to the more number of control packets being received. This is logical as the graphs for delay and percentage loss show that AODV performed worse than the other two protocols.
The graphs for delay vs. number of nodes are shown for all of the three scenarios. The graphs for delay show that DSR has the least delay for all the scenarios while AODV had the highest delay. This shows that for low or high mobility scenarios DSR has the lowest delay, while DSDV and AODV have higher delay. This is due to the fact that source initiated routing protocols incur less communication overhead and traffic.
The graphs for delay vs. number of nodes are shown for all of the three scenarios. The graphs for delay show that DSR has the least delay for all the scenarios while AODV had the highest delay. This shows that for low or high mobility scenarios DSR has the lowest delay, while DSDV and AODV have higher delay. This is due to the fact that source initiated routing protocols incur less communication overhead and traffic.
2.6.3. Percentage Loss
2.6.3. Percentage Loss
The graphs for percentage loss vs. number of nodes are shown for all of the three scenarios. The graphs fo percentage loss show that DSR has the lowest loss for all the three scenarios and AODV has the highest loss a compared to other two protocols.
The graphs for percentage loss vs. number of nodes are shown for all of the three scenarios. The graphs fo percentage loss show that DSR has the lowest loss for all the three scenarios and AODV has the highest loss a compared to other two protocols.
If we consider that DSR had less loss and less delay and assume that the throughtput is shown to be high for DSDV and AODV (due to control packets being received), we can conclude that DSR is a better protocol for the airport and highway scenarios, which represent moderate to high mobility. In the conference scenario, however, the throughput for DSR and DSDV was nearly equal. Even though the emphasis of our work was placed on creating the CAD-HOC tool and not in conducting comprehensive simulation studies, the little realistic simulations we have done raise a question on the generally accepted fact that the choice of ad-hoc routing protocol is mainly application dependent [6].
If we consider that DSR had less loss and less delay and assume that the throughtput is shown to be high for DSDV and AODV (due to control packets being received), we can conclude that DSR is a better protocol for the airport and highway scenarios, which represent moderate to high mobility. In the conference scenario, however, the throughput for DSR and DSDV was nearly equal. Even though the emphasis of our work was placed on creating the CAD-HOC tool and not in conducting comprehensive simulation studies, the little realistic simulations we have done raise a question on the generally accepted fact that the choice of ad-hoc routing protocol is mainly application dependent [6].
Related topics:
Computer ScienceComputer NetworksRouting protocolsMobile Ad Hoc NetworkAd hoc networkVehicular ad hoc networkNetwork SimulatorWireless Ad Hoc NetworkIntelligent NetworksScenario Generation
580 California St., Suite 400
San Francisco, CA, 94104
© 2025 Academia. All rights reserved