15 October 2021 in IEEE Access
This paper presents a comprehensive hardware accelerator architecture of YOLOv3-Tiny targeted for low-cost FPGA with a high frame rate, high accuracy, and low latency. The proposed accelerator implements all YOLO layers in hardware including zero pad layer, convolution layer, leaky ReLU layer, batch normalization layer, max-pooling layer, and up-sampling layer. The architecture is built based on data flow and control flow hybrid architecture. The data preparation and computation process work asynchronously using the data flow paradigm, while the overall governing process is controlled by proposed custom instruction set which adopts the principle of control flow paradigm. The principle of General Matrix Multiplication (GEMM) is adopted to compute the convolution process. We designed a GEMM processor using an optimum size of systolic array architecture. The systolic core is small and the overall architecture supports the multicore system, making it scalable to be implemented on larger size FPGAs. We also proposed a hardware architecture for mapping feature maps into matrix form for GEMM convolution which can save on-chip memory space. Lastly, we modified the original YOLO algorithm to further optimize it in our hardware. The modification includes reducing the bit precision to reduce memory space and bandwidth requirement, merging the normalization layer with the convolution layer to reduce arithmetic complexity, and adding a new DLQ layer to keep the bit size small while maintaining the accuracy. Based on the experimental results, our proposed design manages to achieve a frame rate of 8.3 FPS with the throughput of 31.5 GOPS, outperforming the same convolution computation that is performed by Ryzen 5 3600 CPU up to $69.3\times $ in latency. Moreover, our proposed design also has the smallest clock cycle ratio up to $1.75\times $ than other commercial accelerators. The system is useful and suitable for edge computing applications.
14 January 2022 in IEEE Access
With recent development in internet speed and reliability, cloud computing has become a more reliable solution for the user. In many cases where data privacy is critical, fully homomorphic encryption (FHE) can be a security solution for securing cloud computing. FHE enables computation on encrypted data, hence it ensures data privacy in case of cloud computing. One popular scheme of FHE is the BFV homomorphic encryption scheme, which is based on ring learning with error (RLWE) computation. The BFV scheme uses ring polynomials as the main object, hence its encryption, decryption, and evaluation require high-degree polynomial multiplication. In this paper, we present comprehensive design and implementation of a hardware architecture to accelerate encryption and decryption in BFV scheme. Our accelerator uses convolution approach for calculating a polynomial multiplication. To implement the convolution, we use a systolic array to calculate polynomial convolution followed by a simple delayed subtraction to calculate polynomial modulo reduction inside our accelerator’s core. Moreover, we use a built-in Gaussian pseudo-random number generator (PRNG) to generate Gaussian noise in the encryption operations. Finally, we implement the 1024 degrees BFV accelerator on the Xilinx PYNQ Z1 board and compare the encryption and decryption performances to other methods as well as a software implementation on Intel Core i7 with 8GB memory. Experimental results show that our accelerator outperforms the clock cycles of other methods with the same polynomial degrees 1024 up to $22\times $ . Moreover, our proposed Gaussian PRNG has better $2\times $ correlation compared to the rotation-only-based PRNG. Finally, our accelerator accelerates up to $9\times $ for encryption and $3.5\times $ for decryption as well as $6.8\times $ for overall compared to Microsoft SEAL on Intel Core i7 processor with 8GB memory. The proposed design is scalable for higher degrees polynomial multiplication and useful for security technology such as high-speed secure cloud computing, blind computing, and secure communication.
2 January 2019 in Wireless Personal Communications
SASinantya Feranti Anindya
4 December 2018 in IEEE Transactions on Very Large Scale Integration (VLSI) Systems
In this brief, a relaxation oscillator with comparator delay compensation is presented. Two digital compensation loops are introduced to compensate for the comparator delay. This relaxation oscillator is fabricated in the 0.18-μm CMOS process. The measured average oscillation frequency is 943.1 kHz. The whole oscillator consumes 5.2 μW under a 0.9-V supply. The measured average temperature coefficients (TCs) of the oscillation frequency with and without compensation are 93.88 and 618.17 ppm/°C, respectively. It achieves a TC improvement of 6.6×. The measured frequency variation is within ±0.7% from -10 °C to 100 °C by using the DCLs. The measured line sensitivity is 4.37%/V. The measured figure of merit (FOM 2 ) is 173.3 dB.
6 September 2022 in IEEE Transactions on Biomedical Circuits and Systems
The rapidly increasing number of COVID-19 patients has posed a massive burden on many healthcare systems worldwide. Moreover, the limited availability of diagnostic and treatment equipment makes it difficult to treat patients in the hospital. To reduce the burden and maintain the quality of care, asymptomatic patients or patients with mild symptoms are advised to self-isolate at home. However, self-isolated patients need to be continuously monitored as their health can turn into severe/critical condition within a short period of time. Therefore, a portable device that can remotely monitor the condition and progression of the health of these patients is urgently needed. Here we present a portable device, called Respinos, that can monitor multiparameter vital signs including respiratory rate, heart rate, body temperature, and SpO2. It can also operate as a spirometer that measures forced vital capacity (FVC), forced expiratory volume (FEV), FEV in the first second (FEV1), and peak expiratory flow Rate (PEFR) parameters which are useful for detecting pulmonary diseases. The spirometer is designed in the form of a tube that can be ergonomically inflated by the patient, and is equipped with an accurate and disposable turbine based air flow sensor to evaluate the patient's respiratory condition. Respinos uses rechargeable batteries and wirelessly connects to a mobile application whereby the patient's condition can be monitored in real-time and consulted with doctors via chat. Extensive comparison against medical-grade reference devices showed good performance of Respinos. Overall results demonstrate the potential of Respinos for remote patient monitoring during and post pandemic.