Verifying VSAT simulation settings and results
2 posts · Closed
Hi,
I am currently testing network speeds and want to verify that I am using the correct settings and getting expected outcomes. I am testing for a VSAT internet connection with speeds from 1Mbps to 4Mbps and latencies 50ms-750ms.
I used the following settings on the emulator:
- I used a Wi-Fi connection and then set the speed from 1Mbps to 4Mbps.
- I then set the latency for 50ms and 750ms for each speed.
- I also had packet loss enabled during the test with 0.1% packets.
My results are showing that 1Mbps and 750ms latency, compared to 4Mbps and 750ms latency, has almost no difference in time. Would this be a reasonable expectation? Would using an Ethernet connection make a difference in time?
Thanks
Yes, that's a reasonable expectation, as there is a relationship between Bandwidth-delay Product and TCP buffer size. You can check this article on the topic: How to Calculate TCP throughput for long distance WAN links.
For a standard 64K TCP window it would be: (64K * 8) bits / 0.75 seconds = 699040 bits/sec = 0.7 Mbps.
So with this latency the maximum achievable throughput is less than either limit (1 Mbps or 4 Mbps) and that's why there is no difference in time.