Memory-based VLSI architectures for digital filters: A survey

In memory-based computing, the inner product of the basic multiplication operation is pre calculated and stored in memories and fetched using the address which is one of the operand of multiplication. Thus, the conventional multiplication operation in the digital filter is replaced by memory-based computing thereby improving the area, delay and power efficiency features of the filters. Nowadays, since, the semiconductor memory has become cheaper, faster and efficient in terms of power dissipation, as a consequence, memory-based designs of digital filters are being implemented more and more. In this paper, we present a comprehensive review report on the types of memory-based computing and briefly elaborate various noteworthy architectures which are reported in literature. We have also presented a comparison of the performance of these architectures to make the relevant information readily available to the researchers for their future research work.