Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Enjian Bai has a total of 51 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    FM3H-MH: An Efficient Large-Scale Privacy Amplification Scheme for Quantum Key Distribution

    by Qiankun Li, Enjian Bai, Yun Wu, Han Hai, Yuwen Cao

    Privacy amplification (PA) is a critical process in quantum key distribution (QKD) systems, essential for eliminating information leakage to eavesdroppers and extracting unconditionally secure keys. In practical high-speed QKD systems, a large input size enables the PA scheme to approach the asymptotic limit of the secure-key rate. Concurrently, the throughput of the PA scheme must keep pace with the GHz-level clock rates of modern QKD systems to prevent it from becoming a systemic bottleneck. To globally optimize these objectives, this paper proposes a scalable and highly efficient PA scheme combining a novel δ-almost-universal hash family (FM3H) with modular arithmetic hashing (MH), and proves the security of this hybrid scheme, within the information-theoretic and quantum-side framework. The proposed scheme achieves a throughput of 283 Mbps on a common mobile CPU with an input size of 1010 bits. NIST statistical tests and avalanche-effect measurements are reported as characterizations of the statistical properties of the output sequences. To our knowledge, this work presents the first FPGA-based implementation of a modular hashing-based PA scheme, thereby breaking the dependence of modular hashing schemes on general software platforms. In DV-QKD systems, the FPGA-based scheme achieves a throughput of 1.706 Gbps while maintaining low resource overhead.

  • Adaptive transmit antenna grouping scheme for signed quadrature spatial modulation

    by Yu Guan, Lexi Xu, Yuyang Peng, Enjian Bai, Xue Qin Jiang, Han Hai

  • Open Access

    High-Speed Privacy Amplification Algorithm Based on Two-Dimensional Cellular Automata in Quantum Key Distribution

    by Encheng Tian, Han Hai, Xue-Qin Jiang, Enjian Bai, Genlong Chen, Peng Huang, Guihua Zeng

    Privacy amplification is a critical component in quantum key distribution (QKD) to eliminate eavesdropper information and distill unconditionally secure keys. While conventional Toeplitz hash privacy amplification algorithms face challenges in computational complexity and hardware resource demands, existing one-dimensional cellular automata (1D CAs) algorithm lacks sufficient parallelism and exhibits limited diffusion under finite-size constraints. To overcome these limitations, a privacy amplification algorithm using two-dimensional cellular automata (2D CAs) for QKD is presented in this paper. The proposed algorithm decreases the computation complexity and increases the processing speed. Unlike conventional Toeplitz hash algorithms, the proposed algorithm utilizes the inherent parallelism of 2D CAs to enable simultaneous multi-bit confusion through cyclic row shifts and XOR operations. Furthermore, we prove that the 2D CAs-based algorithm is a universal hash family and satisfies the principle of privacy amplification. The randomness of the proposed algorithm was evaluated through the NIST test suite and an avalanche test, both of which indicated great performance. Finally, we implement the proposed algorithm in field-programmable gate array (FPGA). The experimental results on a Xilinx Artix-7 FPGA demonstrate that our scheme achieves high throughput and significantly reduces hardware resource consumption.


  • An efficient image encryption algorithm based on a five-dimensional conservative hyperchaotic system and gaussian transformation

    by Yaqi Fei, Enjian Bai, Yun Wu, Yuwen Cao

    With the rapid proliferation of Internet of Things (IoT) devices and edge computing systems, the demand for lightweight image encryption solutions has grown exponentially. This paper proposes an efficient image encryption algorithm based on a five-dimensional conservative hyperchaotic system and Gaussian transformation. First, a 5D memristor conservative hyperchaotic system is constructed, which exhibits complex dynamical behaviors including a broad hyperchaotic range and high sensitivity to initial conditions. The proposed system successfully passes all relevant statistical tests from the National Institute of Standards and Technology (NIST), demonstrating that the generated chaotic sequences exhibit strong randomness and meet rigorous statistical standards for cryptographic applications. Building upon this, the designed image encryption algorithm incorporates pixel scrambling and bit-level diffusion stages combined with Gaussian transformation. In the pixel scrambling stage, the image is divided into chunks, with blocks rearranged based on chaotic sequence indices. Rows and columns are further disrupted to scramble pixel positions. During the bit-level diffusion stage, pixel values are split into high four bits and low four bits. Both the high-bit planes and chaotic sequences undergo Gaussian transformation, and after geometric operations, they are recombined to complete the encryption process. Experimental results indicate that encrypted images with nearly uniform histograms, extremely low pixel correlation coefficients, information entropy approaching 8, and NPCR and UACI values close to ideal levels exhibit faster encryption speeds than existing methods, along with strong robustness. This study provides an efficient and secure solution for image encryption, particularly for scenarios requiring high confidentiality, and offers valuable insights for future research aimed at further optimizing encryption performance and security.Both the high-rank planes and chaotic sequences undergo Gaussian transformation, and after geometric operations, they are recombined.

  • DRL-Based Joint Beamforming and UAV Deployment Optimization for STAR-RIS-Assisted Over-the-Air Federated Learning

    by Yun Wu, Yitong Sun, Han Hai, Enjian Bai, Yunlong Yang

    In this letter, we propose a novel framework of the unmanned aerial vehicle (UAV)-mounted simultaneous transmitting and reflecting reconfigurable intelligent surface (STAR-RIS) communication for over-the-air federated learning (AirFL) systems. The STAR-RIS is deployed onboard the UAV to extend the coverage and improve the link quality between the base station (BS) and distributed clients. To maximize the number of participating clients under an aggregation error constraint, we formulate a joint optimization problem involving the beamforming of the BS and the STAR-RIS, and the UAV deployment. A dual-network hierarchical architecture based on deep reinforcement learning (DRL) is designed to tackle this strongly coupled optimization problem by integrating soft actor-critic (SAC) and double deep Q-network (DDQN). Specifically, the SAC optimizes the beamforming vectors, while the DDQN selects the optimal UAV hovering position from a discrete set of candidates. Simulation results demonstrate that incorporating UAV and STAR-RIS allows more clients to be selected in AirFL and significantly improves federated learning performance.


  • Joint Information Reconciliation-Privacy Amplification Scheme for CV-QKD

    by Kun Zhang, Jia Hou, Yan Feng, Xue-Qin Jiang, Enjian Bai, Peng Huang, Guihua Zeng

    In the conventional post-processing phase of Gaussian-modulated coherent state (GMCS) quantum key distribution (QKD) systems, the processes of information reconciliation (IR) and privacy amplification (PA) are typically carried out sequentially and separately. This approach incurs substantial computational complexity and may introduce potential security risks, such as key storage risks. In this paper, we propose a more efficient and secure post-processing scheme that integrates IR and PA into a single process by utilizing well-designed spinal codes. The proposed scheme significantly reduces computational complexity by employing the linear-complexity hash function instead of the superlinear complexity Toeplitz matrices used in conventional PA schemes, and effectively mitigates potential security risks associated with conducting IR and PA as separate processes. Experimental results show that the final secret key throughput of the proposed scheme is at least 250 times higher than that of the conventional PA scheme, with the improvement in throughput becoming even greater as the block lengths increase. More importantly, owing to the rateless property and the capacity-approaching performance of spinal codes, the GMCS QKD systems can achieve reconciliation efficiencies of higher than 97% across a broad range of signal-to-noise ratios from -22.5641 to 0.122 dB.

  • Open Access

    Model-Driven Meta-Learning-Aided Fast Beam Prediction in Millimeter-Wave Communications

    by Wenqin Lu, Xueqin Jiang, Yuwen Cao, Tomoaki Ohtsuki, Enjian Bai

    Beamforming plays a key role in improving the spectrum utilization efficiency of multi-antenna systems. However, we observe that (i) conventional beam prediction solutions suffer from high model training overhead and computational latency and thus cannot adapt quickly to changing wireless environments, and (ii) deep-learning-based beamforming may face the risk of catastrophic oblivion in dynamically changing environments, which can significantly degrade system performance. Inspired by the above challenges, we propose a continuous-learning-inspired beam prediction model for fast beamforming adaptation in dynamic downlink millimeter-wave (mmWave) communications. More specifically, we develop a meta-empirical replay (MER)-based beam prediction model. It combines empirical replay and optimization-based meta-learning. This approach optimizes the trade-offs between transmission and interference in dynamic environments, enabling effective fast beamforming adaptation. Finally, the high-performance gains brought by the proposed model in dynamic communication environments are verified through simulations. The simulation results show that our proposed model not only maintains a high-performance memory for old tasks but also adapts quickly to new tasks.


  • Active RIS Assisted Index Modulation for Wireless Communication System

    by Yun Wu, Xiao Zhang, Enjian Bai, Han Hai

    Index modulation and reconfigurable intelligent surfaces (RIS) have garnered significant attention in recent years due to their promising potential in advancing wireless communications. In this letter, we combine these two techniques and propose an amplification factor-indexed modulation (AFIM) scheme to improve the spectral efficiency in active RIS (A-RIS) aided systems, where the optimized amplification factor of A-RIS is modulated to transmit additional information bits. A closed-form expression for the average bit error probability of AFIM is derived. Furthermore, we extend AFIM to an A-RIS assisted hybrid index modulation (A-RIS-HIM) system, which incorporates reflecting modulation and receive spatial modulation. Besides, two low-complexity detectors based on greedy detection (GD) are proposed for the A-RIS-HIM system. Simulation results are provided to substantiate the effectiveness of the proposed scheme in terms of bit error rate.


  • Average Sum-Rate Maximization of Coupled Phase-Shift STAR-RIS-Assisted SWIPT-NOMA System

    by Yun Wu, Chenglong Zhang, Han Hai, Enjian Bai

    Coupled phase-shift simultaneously transmitting and reflecting reconfigurable intelligent surface (CPS-STAR-RIS) is regarded as an economically efficient alternative to STAR-RIS with an independent phase-shift model. In this letter, we investigate a simultaneous wireless information and power transfer (SWIPT) system with non-orthogonal multiple access (NOMA) assisted by CPS-STAR-RIS. To maximize the average sum-rate (ASR) of the system, we propose an alternating optimization (AO) scheme to jointly optimize the active beamforming vectors at the base station (BS) and the transmission and reflection coefficients at the CPS-STAR-RIS. Specifically, we first equivalently represent the ASR maximization problem by applying the weighted minimum mean square error (WMMSE) algorithm to address the coupling of variables. Subsequently, the successive convex approximation (SCA) algorithm and the penalty dual decomposition (PDD) algorithm are adopted to tackle the optimization challenges of the active beamforming vectors at the BS and the coefficients at the CPS-STAR-RIS, respectively. Simulations show that compared with the conventional RIS (CRIS) and the OMA schemes, the proposed scheme significantly enhances the ASR.