Research On Blackhole Attack Detection And Defense In Wireless Sensor Networks

Wireless sensor networks are typically deployed in the communication environments which are relatively harsh or some unattended areas. These scenarios have quite high requirements for the security of the network. However, the above workspaces easily lead the sensor networks to cause various security problems,especially the attacks against routing protocol vulnerabilities. The blackhole attack as a typical routing attack can do serious damage to the network. Now most of the detection and defense technologies of blackhole attacks are flawed, such as high complexity algorithm, bad detection effect, and being unable to detect collaborative blackhole attacks. Therefore, it is necessary to make further research on blackhole attack detection and prevention methods.This paper analyzes the current detection methods of blackhole attack and summarizes the disadvantages of these methods. According to different types of blackhole attacks, this paper introduces the principle of blackhole attack in detail.Thus the behavior characteristics of blackhole attack are acquired. Based on the behavior characteristics and the defense strategy of blackhole attack, this paper establishes an integrated detection and defense model, including protocols security improvement and blackhole attack signature detection. The model combines the pre-shared symmetric key identity authentication security mechanisms and nodes attributes-and-dubiety based signature detection to improve the performance of detecting and defensing blackhole attack.This paper applies the above model to a wireless sensor network with DSR protocol and optimizes security protocol and the detection algorithm. The improved model introduces improved DSR security protocol and improved behavioral characteristics detection algorithm. The improvement signature detection is accomplished by adding Trap-DSR detection module based on the honeypot idea to the blackhole signature detection module.The improved effect of the optimization model is validated using NS2 simulation platform. The result shows that improved DSR protocol enhances security without reducing the packet delivery rate and end-to-end delay. Besides, the improved behavioral characteristics detection algorithm is superior to that of before improvement in terms of detection accuracy.