Home Sliding DFT Tutorial Weighted Sliding DFT Example Application Contact


Digital Engines LLC

Digital Engines is a US-owned Limited Liability Company founded in 2000 in Delaware to license innovative hardware and software concepts to the signal processing industry. Our latest IP is an enabling technology for sliding signal processing.

Sliding Signal Processor

The sliding signal processor, first introduced several decades ago, executes algorithms such as the discrete Fourier transform (DFT) on a sample-by-sample, sliding basis. A sliding processor is attractive for many applications because it responds immediately to changes in signal behavior (minimum latency), produces continual signal processor updates for predictive control systems, and responds faithfully to transient signals. These features offer significant benefits to systems used in speech recognition, industrial control, avionics, radar, data compression, securities trading, and many others.

Regrettably, the early sliding DFT filter cannot accommodate an amplitude weighting function in any straightforward manner, which is a serious shortcoming. Without an amplitude weighting function, the output channels from a signal processor tend to mix, which hampers signal detection and filtering.

Our weighted sliding DFT architecture enables sliding data systems by means of a unique algorithm and architecture combination that incorporates the weighting function within the DFT operation. It reduces the computational load by half, increases throughput by several times the speed of conventional architectures and generates a wide range of sample-by-sample filters and transforms.