Non-repeatedly marking traceback model for wireless sensor networks

When networks are under attack, traceback schemes can be used to locate the compromised node and remove the compromised node from the network. To resolve the problem that marked information has been easy to be covered by the information of down-stream node in the traceback schemes based on probabilistic packet marking (PPM), `a non-repeatedly marking traceback scheme for wireless sensor networks (NMtrace) is proposed to locate the compromised nodes in this study´. In the NMtrace, when a packet is forwarding to the base station, if the edge information of the packet has been marked by forwarding nodes, the packet will not be marked again by the other nodes. The security shows that the NMtrace can resilience against various attacks which are launched by the compromised forwarding nodes collude with compromised source node. The performance evaluation shows that with the length of the attack path increasing, `the NMtrace needs fewer´ packets to reconstruct the attack path than the PPM-based schemes, the complete convergence rate less affected by the marking probability p and the energy `consumption of the NMtrace is also´ less than PPM-based schemes.