Research On Wireless Channel Secret Key Extraction Under Non-ideal Channel Conditions

For security application scenarios in military wireless communication networks,based on the inherent physical layer properties of wireless channels,physical layer security solutions can provide signal-level security protection for wireless networks at the basement,provide an effective s upplement t o u pper-layer c ryptographic-based s ecurity methods.However,the non-ideal channel conditions caused by the complex military environment and other factors will lead to a sharp decline in security performance.Aiming at the physical layer key extraction under non-ideal wireless channel conditions,this paper studies the specific methods,scheme design and prototype implementation of the t echnology.And combined with the new transmission medium Intelligent Reflecting Surface,this paper try to solve the problems caused by the unsatisfactory channel conditions in the field of key extraction.First,aiming at the key extraction with poor channel reciprocity,under the typical air-ground communication scenario,this paper proposed a single-bit channel feature quantization method based on the change of the position of the maximum channel frequency amplitude response.Then,based on this quantification method,this paper designed a set of key extraction schemes.Furthermore,this paper used Xilinx’s ZYNQ platform to build a real-time channel key extraction system,carried out experimental tests under different UAV air-to-ground communication scenarios such as nodes are static or moving,verified the feasibility and reliability of the proposed key extraction technology over airto-ground channels.On the basis of security and reliability,the key extraction rate can achieve 28 bps.Secondly,combined with the Intelligent Reflecting Surface,this paper studied the design of key extraction schemes in low signal-to-noise ratio environments and quasistatic channel environments.At first,considering the single-user key extraction in a low signal-to-noise ratio scenario,aiming at the key inconsistency rate problem of the traditional scheme,this paper designed a new key extraction scheme based on the asymmetric transformation of the Intelligent Reflecting Surface,which can effectively improve the distribution of channel detection value,make it concentrate on the more robust quantization area,to reduce the key inconsistency rate.The simulation results show that compared with the traditional quantization scheme,the proposed scheme have an 5%-10% improvement in the key inconsistency rate.Finally,this article extends the scenario to multi-users.Aiming at the problem that the overall entropy of multi-user channels in a quasi-static environment is too low,resulting in low key extraction efficiency,this article designed a new type of key extraction scheme combined intelligent reflecting surface random transformation and user allocation.The random transformation of the intelligent reflector is used to increase the entropy of the quasi-static channel environment,which improves the key extraction efficiency.And combined with the designed user allocation algorithm,the channel capacity of each user is increased to ensure the robustness of the key extraction process.The final simulation results show that the proposed scheme can greatly improve the independence and the channel capacity of each user channel,which can improve the efficiency and robustness of key extraction.Compared with the traditional scheme,the proposed scheme also brings a certain degree of performance gain in terms of the key inconsistency rate.Aiming at the non-ideal channel conditions such as poor channel reciprocity and weak channel time variation,this paper proposed the design of the corresponding physical layer key extraction schemes.Based on testing and simulation,the performance of the proposed scheme is verified.The work in this paper has some reference value on how to design physical layer key extraction scheme under non-ideal channel conditions.