IP Spoofing Attack Mitigation using Extreme Learning Machine to Promote Secure Data Transmission

The most important and existing problem in wireless network security is authentication of nodes. The traditional scheme for each node authentication is done by their IP and MAC addresses. But these addresses are highly vulnerable because they are easily spoofed by the attackers. This spoofing attack causes serious damage to the existing network performance. The node is validated using cryptographic authentication. This mechanism exists but is not popular due to its high needs. Hence, this paper would provide the various existing methodology to detect and prevent this IP Spoofing attack in the network. The proposed system uses Received Signal Strength (RSS) of each node, which gives the physical location of nodes. In case multiple attackers having same node identity we make use of the following mechanisms to determine the number of attackers-the multiclass detection problem and cluster based mechanism.