0

I would like to evaluate the error-detection capabilities of TCP and compare them to CRC. Specifically, I want to generate random errors or model-based errors (I have the error model) on TCP packets and observe how the TCP checksum behaves. While I've heard that TCP checksum is weaker than CRC, I'm interested in quantifying the extent of this weakness. Can you please guide me on how to test or simulate this?

Thank you so much

Arash
  • 225
  • 1
  • 11
  • 2
    You can use e. g. [this code](https://stackoverflow.com/a/3954192/987358) to calculate the checksum and run tests with it. – Michael Butscher Jul 02 '23 at 06:17
  • 2
    *"I'm interested in quantifying the extent of this weakness"* - Why not study existing research about this topic, like [Performance of checksums and CRCs over real data](https://www.researchgate.net/publication/3334567_Performance_of_checksums_and_CRCs_over_real_data)? – Steffen Ullrich Jul 02 '23 at 07:32
  • Thank you all (@MichaelButscher and @SteffenUllrich) I basically want to introduce some errors that TCP cannot detect, and I wonder how I should proceed. – Arash Jul 02 '23 at 18:13

0 Answers0