Reduced complexity adaptive filter
AssigneeZarlink Semiconductor Inc.
InventorGord Reesor
Life of this patent
Estimating the term…
Abstract
A FIR filter for use in an adaptive multi-channel filtering system, includes a first memory for storing data, and a second memory for storing filter coefficients. The second memory stores only non-zero valued coefficients or coefficients that are above a predetermined magnitude threshold such that the overall number of coefficients processed is significantly reduced.
Claims17
A FIR filter for use in an adaptive multi-channel filtering system, comprising:
- a first memory for storing data;
- a second memory for storing filter coefficients, said second memory storing only non-zero valued coefficients or coefficients that are above a predetermined magnitude threshold such that the overall number of coefficients processed is significantly reduced,
- wherein each storage location of said second memory includes a first address field for storing an address identifying a channel associated with the stored coefficient and a second address field for identifying the location of corresponding data within said first memory for that particular channel;
- and
- a processing unit for processing said data and said coefficients to form accumulated filter outputs.
A method of processing coefficients in a FIR filter in an adaptive multi-channel filter system, comprising storing said coefficients in a coefficient memory along with associated address data identifying a corresponding channel and data address in a data memory, using said address data to select an appropriate channel and data address for the associated coefficient, wherein each storage location in said coefficient memory includes a coefficient field for storing a particular coefficient, a first address field for storing data identifying a channel, and a second address field for storing data identifying the storage location in said data memory of the data corresponding to said particular coefficient, and processing said coefficients and associated data in said data memory to form accumulated filter outputs.
Dependencies are inferred from the claim text.