Method and apparatus for binary encoding of signals to implement dynamic precision digital MAC operations

摘要

Computer-implemented method of encoding a digital signal intended to be processed by a digital computing system, the method comprising the steps of: Receive (101) a sample of the digital signal quantized over a number Nd of bits, Decompose (102) the sample into a plurality of binary words of configurable size Np bits, Encode (103) the sample by a plurality of pairs of values, each pair comprising one of the binary words and an address corresponding to the position of the binary word in the sample, Transmit (104) the pairs of values to an integration unit to perform a MAC operation between the sample and a weighting coefficient. Figure 1