On the base of study of packet marking scheme’s convergence

a new method of choosing marking probability had been put forward to get the optimization of convergence. Meanwhile

a new safe verified packet marking scheme had been brought forward to countermeasure the information forged by routes which had been controlled by an attacker

and this forged information should disturb the arithmetic of rebuilding the attacking path. Lastly

this scheme’s performances were emulated and tested and the result shows that the performances are better than other packet marking schemes.