Arrhythmia poses a significant threat to human health. Accurate and real-time arrhythmia detection using electrocardiogram (ECG) signals is critical for medical diagnosis. With the rapid development of deep learning technologies, Deep Neural Networks (DNNs) have considerably improved the accuracy of arrhythmia monitoring. However, deploying these techniques on power-constrained wearable devices for real-time monitoring remains challenging. Lightweight designs have thus become essential for achieving low-power and high-throughput real-time detection with limited hardware resources. Binary Neural Networks (BNNs), as a promising lightweight method, exhibit significant potential in reducing hardware resource usage and enhancing inference speed. Nevertheless, existing binarized ECG detection accelerator designs have not yet conducted architectural Design Space Exploration (DSE), and therefore do not sufficiently balance the trade-off between throughput and power consumption. To address this, we propose a highly energy-efficient one-dimensional BNN accelerator for ECG detection. First, we develop a hardware-friendly BNN algorithm by eliminating fully connected layers and nonlinear computations. Subsequently, for the first time, this work utilizes Timeloop to analyze the optimal mapping of the BNN algorithm onto hardware, identifies the optimal accelerator dataflow, and conducts in-depth optimization of both the dataflow and the accelerator hardware circuits based on the Field-Programmable Gate Array (FPGA) architecture. Finally, we introduce a prediction mechanism for fine-grained power control by dynamically disabling redundant computational logic. Experimental results demonstrate that the proposed accelerator achieves a latency of 105.79$\mu $s, power consumption of 160 mW, throughput of 112.2 GOPS, and energy efficiency of 701.51 GOPS/W at 100 MHz. Compared with state-of-the-art Xilinx FPGA-based ECG detection accelerators, our design improves energy efficiency by$15.56\sim 3220.89$times.

Consensus protocols play a crucial role in determining the security and performance of blockchain systems, with committee-based consensus protocols being particularly important, especially in sharding consensus protocols. Anonymous election of committee nodes can mitigate DDoS attacks and bribery attempts. This approach can also be applied to sharding systems to mitigate the risk associated with a single vulnerable shard. However, current node secret selection schemes still present remaining issues. Single secret leader election schemes struggle to elect multiple leaders with equal anonymity, and existing secret committee election schemes lack adequate measures for tracking malicious nodes. To address these issues, we propose accountable secret committee election schemes that not only regulate the number of nodes but also maintain anonymity during the phases of leader proposal and verifier voting. Furthermore, our schemes enable the tracing of malicious nodes in a threshold way. In addition, we introduce two efficient threshold traceable membership proof schemes for both ad hoc and interactive scenarios. Unlike traceable ring signatures, our scheme can trace malicious nodes even after a single malicious behavior. Subsequently, we apply the accountable secret committee election scheme to sharding blockchains and devise a fully accountable anonymous consensus protocol. The experiment demonstrates that this protocol can elevate the difficulty of corrupting a single shard to the level of compromising the entire system, thereby significantly enhancing the security of the sharding system.

The communication efficiency optimization, censorship resilience, and generation of shared randomness are inseparable from the threshold cryptography in the existing Byzantine Fault Tolerant (BFT) consensus. The membership in consensus in a blockchain scenario supports dynamic changes, which effectively prevents the corruption of consensus participants. Especially in cross-chain protocols, the dynamic access to different blockchains will inevitably bring about the demand for member dynamic. Most existing threshold cryptography schemes rely on redefined key shares, leading to a static set of secret sharing participants. In this paper, we propose a general approach to coupling blockchain consensus and dynamic secret sharing. The committee performs consensus confirmation of both dynamic secret sharing and transaction proposals. Our scheme facilitates threshold cryptography membership dynamic, thus underlying support for membership dynamic of threshold cryptography-based BFT consensus schemes. We instantiate a dynamic HotStuff consensus to demonstrate the effectiveness of the scheme. After the correctness and security proof, our scheme achieves the secrecy and integrity of the threshold key shares while ensuring consensus liveness and safety. Experimental results prove that our scheme obtains dynamic membership with negligible overhead.