Dear all, I am doing some simulations with ns-2.31 to calculate the saturation throughput of the IEEE 802.11 DCF, just as Bianchi did in his work "Performance Analysis of the IEEE 802.11 Distributed Coordination Function". I have used NOAH to get rid of the routing effect on the performance of mac layer. One weird phenomenon has confused me for months. When CWmax and CWmin are both small,say, CWmin=CWmax=32, or CWmin=2,CWmax=128, the saturated throughput of the network drops down initially, then it bounces back and fluctuates. I think in this high contention network the throughput should drops down to some low value. However the throughput in my simulation will never drops to that value. Anyone has experienced some similar thing?
Thanks a lot! Daniel -- Daniel City University of Hong Kong Dept.of EE