Integer Based Fully Homomorphic DSP Accelerator using Weighted-Number Theoretic Transform

Fully Homomorphic Encryption (FHE) has gained wide attention in cloud security as it allows computation on encrypted data. However, requires a huge key size, resulting impractical execution time. In this paper, we proposed an FHE hardware accelerator employing Weighted-Number Theoretic Transform (NTT) multiplier. NTT parameters are selected, way that the design is executable Digital Signal Processing (DSP) multiplier, to exploit its high clock rate. As kernel, general form, can be pre-computed and stored Look-up Tables (LUTs). The same LUTs also usable for weight-factor they both have symmetric periodicity properties. This optimization saved 70% of utilization. Next reduction within NTT. special prime moduli exploited accomplish simple operation, where inverse Montgomery multiplication replaced with shift subtraction. optimizations implemented encryption realized Kintex 7 platform. A magnitude 93.2% speedup improvement achieved Toy, compared benchmark software implementation. targeted full DSP implementation, higher frequency (249.19 MHz), while consuming lower resources.