1	On-demand Routing in MANETs: The Impact of a Realistic Physical Layer Model Liang Qin, Thomas Kunz Department of Systems and Computer Engineering Carleton University, Canada
2	Motivation Mobile Ad Hoc Network Routing Protocols: Limited radio transmission range and node mobility Problems: Most existing performance comparisons based on ideal propagation model (free space model, two-ray ground model) Our simulation results with shadowing model show severe performance degradation
3	Propagation Models Ideal Models Free space model and two-ray ground reflection model Shadowing Model b: loss exponent, corresponding to mean transmission range s:shadowing deviation
4	Impact on Routing Protocols Signal strength fluctuates in shadowing model Ideal model (left) and shadowing model (right) over the same distance between two nodes Fluctuation at least 2 orders of magnitude
5	Impact on Routing Protocols Signal strength fluctuation causes active links to “break” Simulations of AODV & DSR show that Packet Delivery Ratio (PDR) decreases significantly Broken links introduce more Route Discovery processes Cause more control messages and longer packet delay
6	One Example in DSR Transmitter did not receive an ACK from the receiver and salvaged the packet Result in duplicated packets and increased traffic
7	Improving Performance Based on AODV and DSR Apply new signal strength threshold during Route Discovery Process Higher than the threshold used in NS2 In order to select more stable links to resist to signal strength fluctuation DSR: Route Request & Route Reply, AODV: Route Request only Different from other proposals (e. g, SSA): Power strength only applying on Route Discovery
8	Improving Performance in DSR Reduce unnecessary control messages in DSR Turn the “shortening” option off to prevent “weak link” joining the route again No Route Request for undeliverable Route Error and Route Reply messages No salvage for Route Error and Route Reply messages
9	Simulation Environment The Network Simulator (NS2) Simulation metrics: Packet Delivery Ratio Total number of control messages Average hop count Average packet latency Simulation parameters 1500x300 m area, 4 data packets/s, 20 sources, 64 Bytes/packet, 20m/s max node speed Equal node coverage (Equals to 50 mobile nodes in ideal model)
10	Simulation Results Performance Comparisons with Different b values
11	Simulation Results Performance Comparisons with Different Mobility Patterns
12	Conclusion Selecting reliable links by using higher signal strength threshold can significantly improve performance PDR increases Total number of control messages and average packet delay decrease Future work Route maintenance Appropriate signal strengths threshold for different node density and b values