Secure communication is essential for agricultural edge-IoT systems operating on resource-constrained wireless devices.Most RSSI-based secret key generation (SKG) methods improve only individual stages of the keygeneration pipeline and are rarely evaluated under realistic outdoor conditions.This paper presents HySKG, an endto-end RSSI-based SKG framework integrating parameter-tuned Moving Window Average preprocessing, Gray Code quantization, Cascade reconciliation, universal hash-based privacy amplification, statistical randomness evaluation, and AES-128 key extraction.The framework is implemented on Raspberry Pi 5 devices using IEEE 802.11n and evaluated across three outdoor propagation scenarios.HySKG achieves Pearson correlations of 0.937-0.993and a final Key Disagreement Rate of 0.00%, while retaining 3,072-3,328 bits of AES-compatible key material, corresponding to 24-26 AES-128 blocks, at an end-to-end Key Generation Rate of 25.99-26.66bps.The complete processing pipeline executes in less than 60 ms.Compared with five evaluated classical SKG baselines, HySKG produces approximately 3.7 times more retained key material under the same experimental conditions.NIST SP 800-22 testing indicates favorable statistical randomness characteristics for the evaluated key sequences, while the Leftover Hash Lemma analysis shows that the retained output lengths exceed the strict information-theoretic secrecy bounds.Therefore, the retained outputs are interpreted as statistically validated, AES-compatible key material rather than informationtheoretically secure secret keys.These results demonstrate the practical efficiency of the proposed integrated SKG framework for lightweight agricultural edge-IoT communication.
Secure communication is essential for agricultural edge-IoT systems operating on resource-constrained wireless devices.Most RSSI-based secret key generation (SKG) methods improve only individual stages of the keygeneration pipeline and are rarely evaluated under realistic outdoor conditions.This paper presents HySKG, an endto-end RSSI-based SKG framework integrating parameter-tuned Moving Window Average preprocessing, Gray Code quantization, Cascade reconciliation, universal hash-based privacy amplification, statistical randomness evaluation, and AES-128 key extraction.The framework is implemented on Raspberry Pi 5 devices using IEEE 802.11n and evaluated across three outdoor propagation scenarios.HySKG achieves Pearson correlations of 0.937-0.993and a final Key Disagreement Rate of 0.00%, while retaining 3,072-3,328 bits of AES-compatible key material, corresponding to 24-26 AES-128 blocks, at an end-to-end Key Generation Rate of 25.99-26.66bps.The complete processing pipeline executes in less than 60 ms.Compared with five evaluated classical SKG baselines, HySKG produces approximately 3.7 times more retained key material under the same experimental conditions.NIST SP 800-22 testing indicates favorable statistical randomness characteristics for the evaluated key sequences, while the Leftover Hash Lemma analysis shows that the retained output lengths exceed the strict information-theoretic secrecy bounds.Therefore, the retained outputs are interpreted as statistically validated, AES-compatible key material rather than informationtheoretically secure secret keys.These results demonstrate the practical efficiency of the proposed integrated SKG framework for lightweight agricultural edge-IoT communication.
This paper aims to develop and evaluate a lightweight and secure Secret Key Generation (SKG) system for static multi-user indoor wireless communication using Received Signal Strength Indicator (RSSI) data. The proposed system integrates fading interpolation, 5-bit multi-bit quantization, Cascade reconciliation, and SHA-128 hashing for privacy amplification. Experiments were conducted at5m,15m, and 25 m in a star topology with three legitimate users and one eavesdropper. At 5 m and 15 m, the system achieved high Key Generation Rates (KGR) of13.96%and14.52%, with Bit Mismatch Rates (BMR) of 86.73 % and 86.02 %. Cascade reconciliation retained up to 4120 bits from 5000-bit sequences, and SHA-128 ensured all keys were compressed to 128 -bit. NIST testing confirmed strong randomness (entropy 0.93) for legitimate users, while eavesdropper entropy remained low (0.81-0.84). AES-128 encryption successfully decrypted messages for valid pairs and failed for Eve, confirming system security. These results highlight the system's potential for secure and scalable communication in indoor IoT environments.
Secure communication in Internet of Things (IoT) devices can be improved through physical-layer secret key generation. In secret key generation, an information reconciliation process is required to match the bit sequences between the two communicating devices. However, this process often involves complex statistical computations that increase system overhead and require the transmission of bits over an insecure public channel. Therefore, this research presents a lightweight reconciliation algorithm based on minimal bit syndrome to reduce overhead and improve key security. The proposed method processes quantized Received Signal Strength (RSS) values into 8-bit block syndrome-based reconciliation mechanism and only the 8-bit syndromes from mismatched block are transmitted over the public channel. System testing was conducted under both static and dynamic scenarios in an indoor area using Raspberry Pi nodes. The results show that the proposed scheme demonstrates an improvement in the Key Generation Rate (KGR) by 2.2% and reduces the processing time by 11.6% in dynamic scenario and 8.3% in the static scenario. Also, the proposed scheme reduce the quantization and reconciliation time by 50% compared to previous research. Furthermore, the proposed method maintained a Key Disagreement Rate (KDR) of 0% across all block size and successfully passed the NIST randomness and complexity test, with all test parameters achieving p-values above 0.01. These result indicate that the proposed scheme improve the performance of existing secret key generation methods and it is suitable for implementation in IoT devices.
Physical-layer security (PLS) schemes use the randomness of the channel parameters, namely, channel state information (CSI) and received signal strength indicator (RSSI) to generate the secret keys. There has been limited work in PLS schemes in long-range (LoRa) wide-area networks (LoRaWANs) which hinder their widespread application. Limitations observed in existing studies include the requirement of high correlation between channel parameter measurements for secret key generation in the proposed schemes and the evaluation of the schemes has only been done in either fully indoor or outdoor environments. The real-world wireless sensor networks (WSNs) and LoRa use cases might not meet both requirements thus making the current PLS schemes inappropriate for these systems. By considering the limitations found in existing PLS schemes, this article proposes LoRA-LiSK, a practical and efficient shared secret key generation scheme for LoRa networks. Our proposed LoRa-LiSK scheme consists of several preprocessing techniques (timestamp matching, two sample Kolmogorov–Smirnov tests, and a Savitzky–Golay filter), multilevel quantization, information reconciliation using Bose–Chaudhuri–Hocquenghem (BCH) codes, and finally, privacy amplification using secure hash algorithm SHA-2. The LoRa-LiSK scheme is extensively evaluated on real WSN/IoT devices in practical application scenarios: 1) indoor to outdoor and 2) LoRa static and mobile outdoor links. It outperforms existing schemes by generating keys with channel parameter measurements of low correlation values (0.2–0.6), while still essentially achieving high key generation rates, and low key disagreement rates (10%–20%). The scheme updates a key in approximately 1 h using an application profile with high transmission rate compared to 3 h reported by existing works while still respecting the duty cycle regulation. It also incurs less communication overhead compared to the existing works.
With the rise of connected vehicular networks, Vehicle-to-Vehicle (V2V) communication has become a fundamental enabler for intelligent transportation systems. However, the open nature of wireless channels introduces risks such as eavesdropping and key compromise. To address these issues, we propose a secure V2V communication scheme based on Physical Layer Key Generation (PLKG) using LoRa technology. The proposed system utilizes Received Signal Strength Indicator (RSSI) measurements between two moving vehicles as the entropy source. To improve the channel correlation, a Savitzky-Golay filter is applied as a preprocessing step. Adaptive multibit quantization is then used to convert the RSSI into binary sequences, followed by the Cascade algorithm for error reconciliation. To enhance randomness, privacy amplification is applied using universal hashing and SHA-1. Finally, the generated key is used for AES-128 encryption. The system is implemented on Raspberry Pi 4B using AcSiP EK-S76SXB and RAK2245 LoRa modules. Experimental validation was conducted in three movement scenarios firstly both vehicle move from opposite-direction at10km/hfor300meters, secondly both vehicle move at the same-direction at10km/hfor800meters, and lastly same as the second scenario but with different speed at20km/h. Results shown that the system achieves a reliable key generation rate (KGR) of up to 176075.73 bit/s while maintaining low computation time, demonstrating its feasibility for secure real-time V2V communication.
Physical-layer secret key generation (SKG) offers a lightweight security approach for resource-constrained Internet of Things (IoT) devices. However, most existing methods are designed for direct links and do not explicitly address relay-assisted LoRa networks. This paper presents LoRa-PRIME, an SKG framework that combines channel probing, Savitzky–Golay preprocessing, modified Adaptive Dual-Threshold Quantization (ADQ), Bose–Chaudhuri–Hocquenghem (BCH) reconciliation, Periodic Reversal for Keying (PRK) randomness enhancement, and SHA-256 privacy amplification. Experiments were conducted using real LoRa devices under indoor line-of-sight (LOS), indoor non-LOS, and outdoor non-LOS conditions. Savitzky–Golay filtering increased channel correlation (e.g., from 0.7595 to 0.8443 in indoor LOS), supporting reliable key extraction. The 4-bit modified ADQ achieved higher key agreement rates (up to 0.9745) and key generation rates (up to 33.04 bps) compared with Sky-Glow, while reducing key generation time to 7.75 s/key. PRK improved randomness quality, increasing the number of NIST-compliant keys from 4 to 14 in indoor LOS and showing similar improvements in other environments. Complexity analysis indicates approximately linear O(N)scalability, with more than 85% of execution time dominated by channel probing, while signal processing and cryptographic stages introduce minimal overhead. The measured energy consumption is 6.619 J per key, lower than several reported LoRa-based SKG implementations. These results suggest that LoRa-PRIME provides an efficient and practical SKG solution for relay-assisted LoRa IoT deployments.
ABSTRACT With the rise of Industry 5.0, smart cities, and the ever‐expanding use of general wireless networks, ensuring seamless communication and robust data security has become a critical challenge. Generating secure secret keys (SKG) through wireless channels is particularly complex in environments where noise and wideband conditions introduce discrepancies and autocorrelation in channel measurements. These issues compromise cross‐correlation and randomness, leading to substantial bit disagreements, distinct keys at the transceivers, and unsuccessful SKG. This research begins by outlining the mathematical model of the signal preprocessing technique called multiscale principal component analysis (MSPCA). Subsequently, it explores the performance of key generation when employing the proposed scheme. A holistic system‐level framework for creating initial shared keys is presented, encompassing quantization methods such as uniform multilevel quantization (UMQ) and encoding methods such as 3‐bit Gray encoding. Monte Carlo‐based simulations in an indoor scenario evaluate system efficacy using metrics like Pearson correlation coefficient, bit disagreement rate (BDR), randomness, and complexity. The proposed scheme achieves a BDR lower than 0.01, a correlation coefficient greater than 0.95, and passes all National Institute of Standards and Technology (NIST) randomness tests, establishing it as a viable solution for securing wireless systems. In the context of Industry 5.0 and smart city infrastructures, where seamless communication and robust data security are paramount, the proposed SKG framework offers significant potential. With its ability to ensure secure and reliable communication, this scheme can underpin the development of advanced wireless systems that cater to the high demands of interconnected ecosystems, enhancing resilience and trust in critical applications.