Underwater Acoustic Physical Layer Key Generation Technology Based On Channel State Information

In recent years,underwater acoustic communication technology has developed rapidly.As a new type of wireless communication,underwater acoustic communication has incomparable advantages over other communication methods,so it has been widely concerned.In order to improve the security of underwater communication,the key technology is usually used for information transmission at the physical layer.However,the underwater acoustic channel is extremely complex,with high delay,low bandwidth,high noise level,multipath effect and topological movement,which seriously damages the accuracy and efficiency of information transmission and hinders the reliability of underwater communication.At the same time,due to the impact of the marine environment,the transmission information is easily obtained by the eavesdropping node,or even interfered with.The above situation will bring great security risks to underwater communication.To solve this problem,this paper proposes a key generation technology for underwater acoustic physical layer based on channel state information.This technology refers to the communication node using pilot signals for information transmission,and then quantizing the received information into a bit sequence to complete the key extraction.Then,the inconsistent key sequence is negotiated for error correction,and a secure transmission key is obtained.The research contents and innovations of this article include:(1)To solve the problems of low randomness and high disagreement rate of keys,a key generation scheme for underwater acoustic channels is proposed.In this scheme,the initial key is extracted from the underwater acoustic channel using the dynamic signal cooperation double-layer compensation K-means clustering quantization(DLCCKQ)method.First,in the process of information transmission,the pilot signal is dynamically added to ensure the randomness of the key,and the channel estimation is performed in cooperation with the Linear Minimum Mean Square Error(LMMSE).The weight matrix is used to correct the influence of underwater noise and improve the uniformity of the key.Secondly,the sensed sample value is filtered and pre-processed to improve the accuracy of post-order research.Then the filtered sample value is characterized by quantization.Compared with the classical quantization scheme,on the basis of the K-means clustering quantization method,a doublelayer compensation calculation is carried out by transferring the clustering center and the difference between each other,and the quantization is carried out after the sample value moves to the vicinity of the quantization center.Finally,the experiment verifies that the key consistency and key generation rate are jointly optimized under the premise of ensuring the security of the key.(2)In order to solve the inconsistency of the initial key after quantization,a strong security key agreement scheme based on authentication mechanism is proposed.First,the hash function is used to authenticate the identity of the communication node,and the nodes that fail the authentication are discarded directly,and the nodes that pass the authentication are negotiated for error correction.Then,based on the traditional key error correction scheme,the error correction coding scheme of the session key control coding system(SKCCS)is proposed.The scheme uses interleaving technology to eliminate the redundancy of the measured value and realize the safe transmission of the key in the open channel.