State-of-the-art mechanisms against eavesdropping first encrypt all packet payloads in the application layer and then split the packets into multiple network paths. However, versatile eavesdroppers could simultaneously intercept several paths to intercept all the packets, classify the packets into streams using transport fields, and analyze the streams by brute-force. In this article, we propose a programming protocol-independent packet processors (P4)-based network immune scheme (P4NIS) against the intractable eavesdropping. Specifically, P4NIS is equipped with three lines of defenses to provide a softwarized network immunity. Packets are successively processed by the third, second, and first line of defenses. The third line basically encrypts all packet payloads in the application layer using cryptographic mechanisms. Additionally, the second line re-encrypts all packet headers in the transport layer to distribute the packets from one stream into different streams, and disturbs eavesdroppers to classify the packets correctly. Besides, the second line adopts a programmable design for dynamically changing encryption algorithms. Complementally, the first line uses programmable forwarding policies which could split all the double-encrypted packets into different network paths disorderly. Using a paradigm of programmable data planes-P4, we implement P4NIS and evaluate its performances. Experimental results show that P4NIS can increase difficulties of eavesdropping and transmission throughput effectively compared with state-of-the-art mechanisms. Moreover, if P4NIS and state-of-the-art mechanisms have the same level of defending eavesdropping, P4NIS can decrease the encryption cost by 69.85%-81.24%.
Encryption mechanisms improve the security of transmitting data. Nevertheless, attackers might silently eavesdrop packets to crack sensitive information or launch cyberattacks to damage the network performance of victims. In this paper, we propose a learning based adaptive network immune mechanism (LANIM) to prevent the eavesdropping attacks. Specifically, LANIM is equipped with three defense lines and one constraint. The first defense line focuses on making decisions about abnormal network conditions by the minimum risk machine learning algorithms. The second defense line is the encryption strategy which focuses on the intent and application. The programmable devices implement novel policies such as multipath transmission and packet encapsulation. LANIM inherits the existing countermeasures based on computational complexity, which is the third defense line. Besides, the policy dynamically updates with random seeds is the constraint of the cyberattack. The attackers have to conquer these three immune lines before the new policy update otherwise the offensive is shattered. We implement LANIM with the P4 language and the Smart Identifier Network (SINET) framework, evaluate the ability to resist the hazards of eavesdropping attacks and explore the trade-offs between security and performance.
27 September 2022 in Wireless Communications and Mobile Computing
Quantum Key Distribution (QKD) is a promising paradigm for Internet of Things (IoT) networks against eavesdropping attacks. However, classical quantum-based mechanisms are overweight and expensive for resource-constrained IoT devices. That is, the devices need to frequently exchange with the QKD controller via an out-band quantum channel. In this paper, we propose a novel Quantum-based Secure and Lightweight Transmission (QSLT) mechanism to ease the overweight pain for IoT devices against eavesdropping. Particularly, the mechanism predistributes quantum keys into IoT devices with SIM cards. Using one of the keys, QSLT encrypts or decrypts IoT sensitive data. It is noting that an in-band key-selection method is used to negotiate the session key between two different devices. For example, on one IoT device, the in-band method inserts a key-selection field at the end of the encrypted data to indicate the key’s sequence number. After another device receives the data, QSLT extracts the key-selection field and decrypts the data with the selected quantum key stored locally. We implement the proposed mechanism and evaluate its security and transmission performances. Experimental results show that QSLT can transmit IoT data with a lower delay while guaranteeing the security performance. Besides, QSLT also decreases power usage by approximately 58.77% compared with state of the art mechanisms.
Concurrent multipath transfer effectively improves network throughput and reliability by utilizing the unoccupied network links. However, the performance of concurrent multipath transfer degrades significantly in the heterogeneous Internet of Things. The reason is that the disorder and retransmission will increase sharply if the quality of links are greatly different. This paper proposes a multipath stateful forwarding mechanism in the granularity of flowlet, which is featured with planning the flowlet forwarding dynamically to improve the transmission success rate. Besides, our mechanism implements stateful forwarding by improving the state information processing capability of the data plane. It can adjust the size of the flowlet based on the network performance of the path, thereby improving the bandwidth resource utilization of the system. In particular, the proposed mechanism can sense network congestion status dynamically, detect network anomalies, and re-plan the forwarding strategy. Experimental results show that in the harsh environment (link quality varies greatly among the three test paths), the proposed mechanism can improve the throughput by 114.7% and reduce the percentage of out of order packets by 29.5% compared with the Round-Robin Scheduling mechanism.
This letter investigates the physical layer authentication (PLA) in dynamic environments, where the users have different mobility velocities. The performance of PLA in this scenario suffers from the inaccurate prediction of the channel statement information (CSI) owing to the Doppler shifts and multipath fading. To overcome the difficulty, a model-driven learning algorithm is developed to predict dynamic CSI. Specifically, the Bahdanau Attention Autoencoder (BAAE)-based PLA scheme is proposed to extract relevant channel features and mitigate inter-symbol interference and inter-carrier interference. Compared to existing deep learning-based PLA schemes, the proposed scheme addresses the core factors contributing to poor authentication accuracy at the algorithmic level, rather than brute-force learning from bigger datasets. Experiment results show that the proposed scheme presents more accurate predictions and lower complexity compared with the existing data-driven models across all tested signal noise ratio levels and velocities. Moreover, it is robust to maintain 99.6% authentication accuracy even when the users are moving with high velocities.
26 March 2020 in IEEE Wireless Communications Letters
Long identifier brings low packet forwarding efficiency in Internet of Things (IoT), whereas short identifier may suffer from the exhaustion of identifier space. Compared with fixed-length identifiers (e.g., IPv4 and IPv6), flexible identifiers are expected for balancing the packet processing efficiency with the various IoT scales. However, it is challenging to make IoT support the flexible identifier-based forwarding. In this letter, we firstly proposed a novel variable-length identifier (VLI) solution for interconnecting IoT networks. In particular, a VLI datagram header is designed to effectively support a flexible identifier field. Following a basic VLI header, one (multiple) extension header(s) can be added if require. Each extension header includes a fixed-size identifier field. According to the combination of multiple identifier fields, the variable-length identifier can be easily achieved, resolved and supported by the IoT nodes in a flexible way. Experimental results show that VLI can decrease the processing delay effectively.