Simulation of MAC layer EEL 6507 Fall 2003 Scott Owen & Eric Donnelly
Covered in this presentation: Background Approach Simulation Results Conclusions
Background Distinguishing features of compared to other protocols: – CSMA/CA, instead of the more common CSMA/CD – Exponential backoff algorithm
Background Modes: Infrastructure and Ad-hoc – Focus on Ad-hoc mode infrastructure network ad-hoc network AP wired network AP: Access Point
Background RTS/CTS – Used to solve “hidden node” problem
Background Timing diagram:
Background Timing diagram (Our Simulation):
Approach Goal: – To create a simulation for the MAC layer that would allow various conditions to be tested – As a test case we will investigate the optimum Maximum Transmission Unit (MTU) size, for different device population densities – Coding was done in Matlab
Matlab Code State Machine Nodes optionally mobile (random linear path) All nodes aware of each other – Can transmit to any node – Node might not be in range – Node might fall out of range mid-transmission – Hidden nodes possible All nodes have data to transmit at all times (not Poisson)
State Machine Ready To TX Tx Rx ACK Tx ACK Wait Tx ACK Rx Rx ACK Wait Backoff
Simulation Results
Conclusions The goal of a suitable simulation of the MAC layer was accomplished Our test experiment relating the optimum MTU to the number of active nodes showed: – Sparsely populated areas are more efficient when all nodes have larger MTU’s – Densely populated areas operate better when all nodes have small MTU’s