A new approach for block-floating-point arithmetic

A new approach for implementing block-floating-point arithmetic is proposed. This approach intends to preserve the least-significant-bits (LSBs) to improve signal processing quality. The preservation of LSBs is automatically and perfectly done by hardware. Several simulation results of the proposed block-floating-point implementation have shown improved SNRs over conventional block-floating-point implementation as expected. For the same number of bits for each data representation in the memory, the SNRs better than floating-point are also observed