Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Log inSign up
Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Blog
Job Board
Q3AS

© 2026 Aqora Quantum S.A.S.

TermsPrivacyLegal Notice

International Electronics Symposium (IES)

Deep Learning based Channel Reciprocity Technique for Secret Key Generation

Dewi Cahyo Wulandari, Mike Yuliana, Amang Sudarsono

Published Aug 6, 2024• 1 citation
Publisher

Abstract

Wireless technology is a development of modern technology where data exchange and communication can be carried out without cables, which cannot be separated from daily activities. But behind the convenience of this wireless technology, there is a problem or threat in the form of tapping and data theft. So in this paper, a key generation system will be created which is expected to be used for a security system for wireless technology. In this paper, the technology used is LoRa. However, the key generation rate on LoRa is often low due to environmental factors. One way to overcome this problem is by conditioning the value of the high RSSI correlation coefficient. In this paper, we propose, a key generation system that was created to increase Channel Reciprocity based on the Deep Learning pre-process method and the Uniform Quantization method to obtain a high correlation coefficient value. From the experimental results, it can be seen that RSSI correlation values were successfully increased using the deep learning method that has been designed in all scenarios and the secret key is successfully generated. The most number of keys obtained is 18, with each key having a length of 128 bits.

DOI
10.1109/ies63037.2024.10665843

Related papers

Deep Learning-Based Channel Reciprocity Learning for Physical Layer Secret Key Generation

Haoyu He, Yanru Chen, Xinmao Huang +4 more

Using the physical layer channel information of wireless devices to establish the highly consistent secret keys is a promising technology for improving the security of wireless networks. Nevertheless, in the time division duplex system, the reciprocity of the wireless channel that is the basic principle of key generation is impaired by nonsimultaneous sampling and noise factors. Existing physical layer key generation approaches rely on hand-crafted feature extraction algorithms, which have high overhead or security issues and are impractical in real-world situations. This paper presents a novel physical layer key generation method to extract highly consistent keys from imperfect channel responses, which exploits channel reciprocity through deep learning. Specifically, we first design the Channel Reciprocity Learning Net (CRLNet), a neural network for efficiently learning channel reciprocity features from the wireless channel in TDD OFDM systems. Later, a new key generation scheme based on CRLNet is developed that can achieve a high key agreement rate. Experiments indicate that the CRLNet-based key generation scheme performs excellently in terms of key generation rate, key error rate, and randomness, confirming that our method has better performance and lower overhead than existing methods.

Security and Communication Networks•Mar 19, 2022•25 citations

LoRa-Key: Secure Key Generation System for LoRa-Based Network

Weitao Xu, Sanjay Kumar Jha, Wen Song Hu

Physical layer key generation that exploits reciprocity and randomness of wireless fading channels has attracted considerable attention in recent years. Despite much research efforts in this field, the problem of wireless key generation at long distance and low data rate remains unknown and has not been studied. In this paper, we conduct extensive experiments and analysis in real indoor and outdoor environments to explore the feasibility of wireless key generation for long range (LoRa)-based network. Our experimental results show that: 1) the low transmission rate will lead to low channel reciprocity which makes wireless key generation significantly challenging and 2) when the requirement of high reciprocity is fulfilled, two nodes can generate the same secret key even when they are far away from each other (a few kilometers). Building on the strengths of existing secret key extraction approaches, we present LoRa-Key, the first complete key establishment protocol for LoRa network by exploring the shared randomness extracted from measured received signal strength indicator. LoRa-Key employs a number of signal processing techniques to improve key generation rate significantly. Moreover, we propose a novel compressive sensing-based reconciliation framework to reduce mismatch rate. Experimental results show that LoRa-Key can achieve key establishment rates of 18 bit/s in stationary scenario and 31 bit/s in mobile scenario. To the best of our knowledge, this is the first work that studies key generation protocol for LoRa network.

IEEE Internet of Things Journal•Dec 18, 2018•120 citations

A Low-Complexity Deep Learning Approach to Enhance Secret Key Generation for IoT Networks

Anas Alashqar, Ehsan Olyaei Torshizi, Raed Mesleh +1 more

This article presents a novel low-complexity deep learning model for physical-layer secret key generation (PSKG), specifically designed to enhance wireless security in the Internet of Things (IoT). PSKG typically generates cryptographic keys by exploiting the reciprocal nature of wireless channels; however, this reciprocity is frequently compromised in time-division duplex (TDD) systems due to hardware imperfections and noise, which significantly complicate the key generation process. To effectively address these challenges, the proposed deep neural network (DNN) is developed to efficiently learn and enhance reciprocity features, even in the presence of imperfect channel state information (CSI). Furthermore, the study introduces a DNN-based PSKG method that strategically leverages the phase of channel frequency responses for key extraction. The results conclusively demonstrate that the proposed model substantially reduces the key disagreement ratio (KDR) and enhances randomness, thus providing a robust and practical solution for securing wireless communications in IoT environments.

International Wireless Communications and Mobile Computing Conference (IWCMC)•May 12, 2025•1 citation

Exploring the Feasibility of Physical Layer Key Generation for LoRaWAN

Weitao Xu, Sanjay Kumar Jha, Wen Song Hu

Physical layer key generation that exploits reciprocity and randomness of wireless fading channels has attracted considerable attention in recent years. Despite much research efforts in this field, the problem of wireless key generation at long distance and low data rate remains unknown and has not been studied. In this paper, we conduct extensive experiments and analysis in real indoor and outdoor environments to explore the feasibility of wireless key generation for Long Range Wide Area Network (LoRaWAN). Based on the results, we present LoRa-Key, the first complete key establishment protocol for LoRaWAN by exploring the shared randomness extracted from measured Received Signal Strength Indicator (RSSI). LoRa-Key employs a number of signal processing techniques to improve key generation rate significantly. Experimental results show that LoRa-Key can achieve key establishment rates of 13 bit/s in stationary scenario and 21 bit/s in mobile scenario. To the best of our knowledge, this is the first work that studies key generation protocol for LoRaWAN.

IEEE International Conference on Trust, Security and Privacy in Computing and Communications (TrustCom)•Aug 1, 2018•15 citations

Physical Layer Key Generation from Wireless Channels with Non-ideal Channel Reciprocity: A Deep Learning Based Approach

Cheng Feng, Li Sun

Physical layer encryption is a promising solution to provide transmission security for Internet-of-Things (IoT). To realize physical layer encryption, channel state information (CSI) as a random source to generate secret keys should be exploited. The reciprocity of wireless channel between legitimate communication nodes which is typically assumed for time division duplex (TDD) systems, is of critical importance to ensure the consistency in generated keys at two nodes. However, in practical TDD systems, due to the asynchronous measurement of CSI, the receiver noise and the the difference in hardware, etc., the reciprocity of channel measurements between legitimate parties is hard to be guaranteed, which makes physical layer key generation approach infeasible. To address this issue, we propose a wireless key generation scheme based on deep learning techniques. Firstly, a feature extraction network, which consists of two auto-encoders trained jointly, is developed to extract the common feature from the different yet correlated channel observations at two legitimate nodes, which will be used as the source to generate keys. Then a quantizer with an adaptive guard band width is devised to output a binary secret-key sequence from the feature vector. Numerical results show that the proposed scheme achieves remarkable improvement in terms of key disagreement rate (KDR) and key generation rate (KGR) compared existing counterparts in the literature.

IEEE Vehicular Technology Conference•Jun 1, 2022•13 citations

Enhancing Channel Reciprocity of Secret Key Generation Scheme by Using Modified Polynomial Regression Method

Mike Yuliana, Wirawan Wirawan, Suwadi Suwadi +2 more

Channel reciprocity is a very crucial component in establishing a secret key generation scheme. Non-simultaneous measurement conditions and noise cause low channel reciprocity, which is indicated by measurements of channel parameters that are not perfectly identical in both users. The stages of the secret key generation scheme that directly quantifies the channel parameters will have an impact on the high difference in the bits produced. In this paper, we propose the use of Modified Polynomial Regression methods to enhance channel reciprocity that would lower the rate of key disagreement between two users. We also propose a combination of the Modified Polynomial Regression method with multilevel quantization to overcome the trade-off between performance parameters, i.e. the key disagreement rate and key generation rate. The test results conducted in an indoor wireless environment show that our proposed scheme has better performance compared to the existing scheme in terms of increasing reciprocity, key generation rate and decreasing key disagreement rate. The resulting secret key has also fulfilled the randomness requirements set by the National Institutes of Standards and Technology statistical suite.

International Conference on Computer Engineering, Network, and Intelligent Multimedia (CENIM)•Nov 1, 2018•8 citations
Wireless technology is a development of modern technology where data exchange and communication can be carried out without cables, which cannot be separated from daily activities. But behind the convenience of this wireless technology, there is a problem or threat in the form of tapping and data theft. So in this paper, a key generation system will be created which is expected to be used for a security system for wireless technology. In this paper, the technology used is LoRa. However, the key generation rate on LoRa is often low due to environmental factors. One way to overcome this problem is by conditioning the value of the high RSSI correlation coefficient. In this paper, we propose, a key generation system that was created to increase Channel Reciprocity based on the Deep Learning pre-process method and the Uniform Quantization method to obtain a high correlation coefficient value. From the experimental results, it can be seen that RSSI correlation values were successfully increased using the deep learning method that has been designed in all scenarios and the secret key is successfully generated. The most number of keys obtained is 18, with each key having a length of 128 bits.
Using the physical layer channel information of wireless devices to establish the highly consistent secret keys is a promising technology for improving the security of wireless networks. Nevertheless, in the time division duplex system, the reciprocity of the wireless channel that is the basic principle of key generation is impaired by nonsimultaneous sampling and noise factors. Existing physical layer key generation approaches rely on hand-crafted feature extraction algorithms, which have high overhead or security issues and are impractical in real-world situations. This paper presents a novel physical layer key generation method to extract highly consistent keys from imperfect channel responses, which exploits channel reciprocity through deep learning. Specifically, we first design the Channel Reciprocity Learning Net (CRLNet), a neural network for efficiently learning channel reciprocity features from the wireless channel in TDD OFDM systems. Later, a new key generation scheme based on CRLNet is developed that can achieve a high key agreement rate. Experiments indicate that the CRLNet-based key generation scheme performs excellently in terms of key generation rate, key error rate, and randomness, confirming that our method has better performance and lower overhead than existing methods.
Physical layer key generation that exploits reciprocity and randomness of wireless fading channels has attracted considerable attention in recent years. Despite much research efforts in this field, the problem of wireless key generation at long distance and low data rate remains unknown and has not been studied. In this paper, we conduct extensive experiments and analysis in real indoor and outdoor environments to explore the feasibility of wireless key generation for long range (LoRa)-based network. Our experimental results show that: 1) the low transmission rate will lead to low channel reciprocity which makes wireless key generation significantly challenging and 2) when the requirement of high reciprocity is fulfilled, two nodes can generate the same secret key even when they are far away from each other (a few kilometers). Building on the strengths of existing secret key extraction approaches, we present LoRa-Key, the first complete key establishment protocol for LoRa network by exploring the shared randomness extracted from measured received signal strength indicator. LoRa-Key employs a number of signal processing techniques to improve key generation rate significantly. Moreover, we propose a novel compressive sensing-based reconciliation framework to reduce mismatch rate. Experimental results show that LoRa-Key can achieve key establishment rates of 18 bit/s in stationary scenario and 31 bit/s in mobile scenario. To the best of our knowledge, this is the first work that studies key generation protocol for LoRa network.
This article presents a novel low-complexity deep learning model for physical-layer secret key generation (PSKG), specifically designed to enhance wireless security in the Internet of Things (IoT). PSKG typically generates cryptographic keys by exploiting the reciprocal nature of wireless channels; however, this reciprocity is frequently compromised in time-division duplex (TDD) systems due to hardware imperfections and noise, which significantly complicate the key generation process. To effectively address these challenges, the proposed deep neural network (DNN) is developed to efficiently learn and enhance reciprocity features, even in the presence of imperfect channel state information (CSI). Furthermore, the study introduces a DNN-based PSKG method that strategically leverages the phase of channel frequency responses for key extraction. The results conclusively demonstrate that the proposed model substantially reduces the key disagreement ratio (KDR) and enhances randomness, thus providing a robust and practical solution for securing wireless communications in IoT environments.
Physical layer key generation that exploits reciprocity and randomness of wireless fading channels has attracted considerable attention in recent years. Despite much research efforts in this field, the problem of wireless key generation at long distance and low data rate remains unknown and has not been studied. In this paper, we conduct extensive experiments and analysis in real indoor and outdoor environments to explore the feasibility of wireless key generation for Long Range Wide Area Network (LoRaWAN). Based on the results, we present LoRa-Key, the first complete key establishment protocol for LoRaWAN by exploring the shared randomness extracted from measured Received Signal Strength Indicator (RSSI). LoRa-Key employs a number of signal processing techniques to improve key generation rate significantly. Experimental results show that LoRa-Key can achieve key establishment rates of 13 bit/s in stationary scenario and 21 bit/s in mobile scenario. To the best of our knowledge, this is the first work that studies key generation protocol for LoRaWAN.
Physical layer encryption is a promising solution to provide transmission security for Internet-of-Things (IoT). To realize physical layer encryption, channel state information (CSI) as a random source to generate secret keys should be exploited. The reciprocity of wireless channel between legitimate communication nodes which is typically assumed for time division duplex (TDD) systems, is of critical importance to ensure the consistency in generated keys at two nodes. However, in practical TDD systems, due to the asynchronous measurement of CSI, the receiver noise and the the difference in hardware, etc., the reciprocity of channel measurements between legitimate parties is hard to be guaranteed, which makes physical layer key generation approach infeasible. To address this issue, we propose a wireless key generation scheme based on deep learning techniques. Firstly, a feature extraction network, which consists of two auto-encoders trained jointly, is developed to extract the common feature from the different yet correlated channel observations at two legitimate nodes, which will be used as the source to generate keys. Then a quantizer with an adaptive guard band width is devised to output a binary secret-key sequence from the feature vector. Numerical results show that the proposed scheme achieves remarkable improvement in terms of key disagreement rate (KDR) and key generation rate (KGR) compared existing counterparts in the literature.
Channel reciprocity is a very crucial component in establishing a secret key generation scheme. Non-simultaneous measurement conditions and noise cause low channel reciprocity, which is indicated by measurements of channel parameters that are not perfectly identical in both users. The stages of the secret key generation scheme that directly quantifies the channel parameters will have an impact on the high difference in the bits produced. In this paper, we propose the use of Modified Polynomial Regression methods to enhance channel reciprocity that would lower the rate of key disagreement between two users. We also propose a combination of the Modified Polynomial Regression method with multilevel quantization to overcome the trade-off between performance parameters, i.e. the key disagreement rate and key generation rate. The test results conducted in an indoor wireless environment show that our proposed scheme has better performance compared to the existing scheme in terms of increasing reciprocity, key generation rate and decreasing key disagreement rate. The resulting secret key has also fulfilled the randomness requirements set by the National Institutes of Standards and Technology statistical suite.