This technology is a method to implement feedback loops in continuous-time digital processors (CT-DSPs) for improved signal processing.
Current CT-DSPs have advanced signal processing capabilities by being capable to continuously track event inputs. However, CT-DSPs have thus far only been implemented with feed-forward, transversal filters which have processing limitations. One potential reason for this is that high-order transversal filters request a large number of delay blocks, which consume a large amount of power and chip area. Currently, there are no efficient high-order filters that are both area- and energy-efficient, which are needed to expand signal processing capabilities.
This technology is a CT-DSP system implemented with separate timing and data paths. This CT-DSP system can apply an event-grouping and detection method that can be used in feedback systems with propagation delays. By implementing event-detection into a feedback loop of a continuous-time digital system, the system can automatically stop when there is no event in the system and commence operation when new events are detected by the system, increasing its energy-efficiency. As a result, this technology offers improved signal processing capabilities with increased selectivity, robust frequency response to delay line mismatches, and is efficient in its hardware use.
Yu Chen Ph.D. (work done in Yannis Tsividis’s lab)
IR CU16130
Licensing Contact: Greg Maskel