基于随机博弈机制的人工智能网络安防算法研究

On Artificial Intelligence Network Security Algorithm Based on Stochastic Game Mechanism

为解决当前人工智能网络安防算法中存在的恶意节点识别困难、捕捉效率低下等不足,提出了一种基于随机博弈机制的人工智能网络安防算法。首先,采取MD5加密与四轮32位秘钥加密方式,设计了随机博弈加密机制的用户ID秘钥生成与解密方案,以实现对网络恶意节点的精确识别,消除因节点相似度较高而出现的漏查现象;随后,通过用户ID鉴别与坐标一次定位捕获的方式,构建了一种基于博弈定位机制的恶意节点捕捉方法,可实现恶意节点在隐藏状态下的高效识别,提高网络的抗攻击能力。通过仿真实验可以得到,与常见的几种网络安防技术相比,该算法具有更低的捕捉误差和更强的抗攻击能力。

In order to solve the problems of malicious node identification and low capture efficiency in current artificial intelligence network security algorithms,an artificial intelligence network security algorithm based on stochastic game mechanism is proposed.Firstly,by using MD5 encryption and 4 rounds of 32-bit secret key encryption,a scheme of user ID secret key generation and decryption based on random game encryption mechanism is designed to realize the accurate identification of malicious nodes in the network and eliminate the phenomenon of missing checks due to the high similarity of nodes.Then,through the way of user ID identification and coordinate one-time location capture,a vicious game location mechanism is constructed.The method of Italian node capture can realize the efficient identification of malicious nodes in hiding state and improve the network's anti-attack ability.The simulation results show that compared with several common schemes,the proposed algorithm has the outstanding advantages of low capture error and high anti-attack intensity and has higher practical deployment value.