Industrial Technology Research Institute

Cheng-Shing Wu

An-Yeu Wu

I-Hsien Lee

Chien-Hao Chen

Meng-Hsin Ou

Po-Min Wang

Chia-Ho Pan

A packet classifier for a network and the method thereof are provided. The packet classifier includes at least one content locator and one matching engine. Each of the content locators performs at least one location operation. Each of the location operations outputs the sum of at least one accumulation item as an output offset. Each of the accumulation items is generated according to at least one content locator system parameter and/or the content data of a network packet. One of the accumulation items is a base value, and the base value is an offset defined by system parameters or the output offset of a location operations of one of the content locators. The output offset of the last location operation of one of the content locators is the location of a matching field of the network packet. This location is provided to the matching engine for a data matching operation.

Packet classifier for a network and method thereof

A symbol timing recovery method for DMT-VDSL is provided. The method converts the extracted sample to the frequency domain. Using the peaks and the company peaks in the frequency domain, the method determines the correct position of a peak response, thereby calibrating the sample extraction position. Thus, the symbol timing can be recovered.

Symbol timing recovery method for DMT-VDSL

Providing multi-path echo cancellation comprises providing a main echo signal corresponding to a main echo section of an echo response and providing a plurality of tail echo signals, each of the plurality of tail echo signals corresponding to respective tail echo sections of the echo response. Next, providing a multi-path echo canceller comprises summing the main echo signal and the plurality of tail echo signals to provide a synthesized echo.

Methods and systems for providing multi-path echo cancellation

A method for implementing a multiplier-less FIR filter is disclosed, in which the FIR filter performs the convolution of where his the k-th coefficient of the FIR filter. The dynamic range of the filter coefficients is compressed by a transformation: where parameters &#x3b1; and &#x3b2; are chosen depending on filter type, &#x2212;1&#x2266;&#x3b1;&#x2266;1, and m denotes iteration numbers of coefficient operation of transformation, so as to avoid the serious quantization error caused by the phenomenon of SPT distribution. Then, the compressed coefficients are quantized into SPT numbers, and the coefficients are optimized by removing redundant STP numbers.

Method for implementing a multiplier-less FIR filter

A CORDIC method and a CORDIC architecture applied in vector rotation are disclosed. An elementary angles set is extended by representing the elementary angles as the arctangent of the sum of two single signed-power-of-two terms to an extended elementary angles set. A combination of elementary angles is found from the extended elementary angles set such that the residue angle error can be minimized. A quantized scaling factor is used to scale the combination of elementary angles after being micro-rotated.

Growing community of inventors

Miaoli, Taiwan

Cheng-Shing Wu