Patents by Inventor Weishi Feng
Weishi Feng has filed for patents to protect the following inventions. This listing includes patent applications that are pending as well as patents that have already been granted by the United States Patent and Trademark Office (USPTO).
-
Patent number: 7392464Abstract: A data encoding system for a data stream comprises an interleaving module that receives the data stream as N bit data blocks and that reverses positions of at least two of the N bits of selected ones of the data blocks. A generating module generates P error checking bits for each of the N bit data blocks. An insertion module receives the P error checking bits from the generating module and inserts the P error checking bits into the corresponding data block received from the interleaving module.Type: GrantFiled: July 22, 2004Date of Patent: June 24, 2008Assignee: Marvell International Ltd.Inventors: Weishi Feng, Zhan Yu
-
Patent number: 7356072Abstract: A communication system comprises direct sequence spread-spectrum demodulator that receives first and second signals and that correlates a spreading code with a preamble of each data packet to produce a spreading codeword including a predetermined number of chips. A signal quality measurement device is responsive to the direct sequence spread-spectrum demodulator, measures signal quality values corresponding to each of the first and second signals for each data packet after correlation with the spreading code and that selects one of the first and second signals on the basis of the measured signal quality values.Type: GrantFiled: October 12, 2006Date of Patent: April 8, 2008Assignee: Marvell International, Ltd.Inventors: Weishi Feng, Hsiao Tang, Guorong Hu, Yungping Hsu
-
Patent number: 7322004Abstract: A Reed-Solomon decoder includes an inversionless Berlekamp-Massey algorithm (iBMA) circuit with a pipelined feedback loop. A first polynomial generator generates error locator polynomial values. A discrepancy generator generates discrepancy values based on the error locator polynomial values and the scratch polynomial values. Arithmetic units are used to generate the discrepancy values are also used to generate the error locator polynomial to reduce circuit area. A first delay circuit delays the discrepancy values. A feedback loop feeds back the delayed discrepancy values to the error locator polynomial generator. An error location finder circuit communicates with the iBMA circuit and identifies error locations. An error value computation circuit communicates with at least one of the error location finder circuit and the iBMA circuit and generates error values.Type: GrantFiled: April 20, 2006Date of Patent: January 22, 2008Assignee: Marvell International, Ltd.Inventors: Zhan Yu, Weishi Feng
-
Publication number: 20080016395Abstract: Apparatus having corresponding methods and computer programs comprise: a processor; a test interface that is in communication with the processor only when the test interface is enabled; a first memory to store firmware for the processor; and a second memory to store boot code for the processor, wherein when the processor is booted, the boot code causes the processor to read a portion of the firmware from a predetermined location in the first memory; wherein the test interface is enabled only when the portion of the firmware has a predetermined value.Type: ApplicationFiled: January 18, 2007Publication date: January 17, 2008Applicant: Marvell International Ltd.Inventor: Weishi Feng
-
Patent number: 7312727Abstract: A communications channel comprises a seed selector that selectively removes X M-bit symbols of user data from a seed set comprising Y M-bit symbols and that selects a scrambling seed from Y-X symbols remaining in the seed set, where X, Y and M are integers greater than one. A Hamming weight coding device that determines a Hamming weight of symbols of scrambled user data that are generated based on the user data and the selected scrambling seed and that selectively codes the symbols depending upon the determined Hamming weight.Type: GrantFiled: April 11, 2007Date of Patent: December 25, 2007Assignee: Marvell International Ltd.Inventors: Weishi Feng, Zhan Yu
-
Patent number: 7310765Abstract: A system for detecting errors in received input data includes a first error detection circuit. The first error detection circuit is configured to receive the input data. The input data includes at least one of data and data with errors. The first error detection circuit is configured to generate a first error detection sequence in a first order. The system includes a second error detection circuit. The second error detection circuit is configured to receive the first error detection sequence and an error sequence. The error sequence is received in a second order that is different from the first order when there is data with errors. The second error detection circuit is configured to generate a second error detection sequence that indicates whether the error sequence is generated correctly.Type: GrantFiled: February 4, 2005Date of Patent: December 18, 2007Assignee: Marvell International Ltd.Inventors: Weishi Feng, Liang Zhang, Zhan Yu
-
Patent number: 7272198Abstract: A method of decoding data includes receiving a symbol and determining a data rate that was used to encode the symbol. A set of correlator output signals are generated based on a first mode when a first data rate was used to encode the symbol and based on a second mode when a second data rate was used to encode the symbol. A maximum-valued signal in one of the set of correlator output signals is identified. The maximum-valued signal in one of the set of correlator output signals is modulated.Type: GrantFiled: September 22, 2006Date of Patent: September 18, 2007Assignee: Marvell International Ltd.Inventors: Guorong Hu, Yungping Hsu, Weishi Feng
-
Patent number: 7269778Abstract: A communications channel such as a data storage system removes unwanted bit patterns from user data without using run length limited coding on the user data. A buffer receives the user data. A data dependent scrambler communicates with the buffer and selects one of a plurality of scrambling sequences based the user data stored in the buffer or generates a scrambling sequence based on the user data stored in the buffer. A scrambling device communicates with the data dependent scrambler and scrambles the user data stored in the buffer with the selected scrambling sequence from the data dependent scrambler.Type: GrantFiled: April 25, 2003Date of Patent: September 11, 2007Assignee: Marvell International Ltd.Inventors: Weishi Feng, Pantas Sutardja
-
Publication number: 20070203947Abstract: A method for providing Internet services employing contact personal distance information by classifying contacts in a user's contact list with a certain classification. The method includes providing a contact list for each user that has an account with the service provider, and providing a personal distance classification for contacts in the contact list that identifies personal information between the user and the contacts in the contact list. The user can use the personal distance classification to allow the user to selectively communicate with the contacts in the contact list over the Internet using the classification. In one non-limiting embodiment, the personal distance classification includes classifying the contacts in the contact list as family members, friends or business associates.Type: ApplicationFiled: February 23, 2007Publication date: August 30, 2007Inventor: Weishi Feng
-
Patent number: 7249310Abstract: An error correcting Reed-Solomon decoder includes an error locator polynomial generator that generates an error locator polynomial and a scratch polynomial based on an inversionless Berlekamp-Massey algorithm (iBMA). An error location finder communicates with the error locator polynomial generator and generates error locations. An error values finder communicates with the error locator polynomial generator and generates error values directly from the error locator polynomial and the scratch polynomial.Type: GrantFiled: October 4, 2005Date of Patent: July 24, 2007Assignee: Marvell International Ltd.Inventors: Weishi Feng, Zhan Yu
-
Patent number: 7239648Abstract: A method and apparatus enables advanced signal processing in a wireless local area network (WLAN). First and second WLAN transceivers are provided with advanced signal processing capabilities. A maximum interframe period between data and an acknowledgement is required by the WLAN for compatibility. A duration of the interframe period is shorter than a duration that is required to perform the advanced signal processing. The first WLAN transceiver transmits a header and data. A first data field in the header is specified that enables the advanced signal processing. A second data field is specified that defines a data time period and an extension time period. The first WLAN transceiver transmits data during the data time period and dummy data during the extension time period. The second WLAN transceiver receives the header and initiates receiver processing during the extension time period.Type: GrantFiled: February 21, 2002Date of Patent: July 3, 2007Assignee: Marvell International Ltd.Inventors: Weishi Feng, Peter Loc
-
Patent number: 7234097Abstract: A communications channel that receives a user data sequence including N symbols and that supports host CRC includes a host bus interface (HBI) that generates cyclic redundancy check (CRCU) bits based on the user data sequence. A data dependent scrambler (DDS) receives the user data sequence and the CRCU bits and generates a scrambling seed. The DDS generates a scrambled user data sequence that is based on the user data sequence and the scrambling seed and generates a difference sequence.Type: GrantFiled: November 4, 2003Date of Patent: June 19, 2007Assignee: Marvell International Ltd.Inventors: Weishi Feng, Zhan Yu
-
Patent number: 7218255Abstract: A communications channel includes a buffer that receives symbols of user data including a plurality of M-bit symbols. A seed selector receives the M-bit symbols of the user data, selectively removes symbols of the user data from a seed set, and selects a scrambling seed from symbols remaining in the seed set. A scrambling device that communicates with the seed selector and the data buffer generates scrambled user data using the user data and the selected scrambling seed. A Hamming weight coding device determines a Hamming weight of symbols of the scrambled user data and selectively codes the symbols depending upon the determined Hamming weight.Type: GrantFiled: August 12, 2003Date of Patent: May 15, 2007Assignee: Marvell International Ltd.Inventors: Weishi Feng, Zhan Yu
-
Patent number: 7145969Abstract: An apparatus and a method for decoding data that has been Complementary Code Keying (CCK)-encoded at one of the set of first and second differing data rates. The method includes receiving a symbol, determining which of the first and second data rates was used to encode the symbol and applying the symbol to a first correlator to generate a set of correlator output signals. The first correlator generates the set of correlator output signals based on a first mode when the first data rate was used to encode the symbol and based on a second mode when the second data rate was used to encode the symbol. The method also includes identifying a maximum-valued signal in one of the set of correlator output signals and demodulating the maximum-valued signal in one of the set of correlator output signals.Type: GrantFiled: March 5, 2002Date of Patent: December 5, 2006Assignee: Marvell International Ltd.Inventors: Guorong Hu, Yungping Hsu, Weishi Feng
-
Patent number: 7099380Abstract: A method and an apparatus for selecting an antenna from an antenna diversity array in a wireless communication system are provided. The method includes the steps of modulating an incoming signal with a spread-spectrum type modulation, measuring a signal quality value for each antenna in the antenna diversity array, and selecting an antenna from the antenna diversity array on the basis of the measured signal quality values. The step of measuring a signal quality value for each antenna may include computing peak-to-average ratio values by dividing a peak sample value by an average sample value, where the average sample value is determined by averaging all of the predetermined number of sample values for each of the plurality of incoming spreading codewords, and the peak sample value is determined by selecting the maximum of all of the predetermined number of sample values for each of the plurality of incoming spreading codewords.Type: GrantFiled: November 16, 2001Date of Patent: August 29, 2006Assignee: Marvell International Ltd.Inventors: Weishi Feng, Hsiao Tang, Guorong Hu, Yungping Hsu
-
Patent number: 7051267Abstract: A Reed-Solomon decoder includes an inversionless Berlekamp-Massey algorithm (iBMA) circuit with a pipelined feedback loop. An error locator polynomial generator generates error locator polynomial values. A scratch polynomial generator generates scratch polynomial values. A discrepancy generator generates discrepancy values based on the error locator polynomial values and the scratch polynomial values. Multipliers used to generate the discrepancy values are also used to generate the error locator polynomial to reduce circuit area. A first delay circuit delays the discrepancy values. A feedback loop feeds back the delayed discrepancy values to the error locator polynomial generator and the scratch polynomial generator. An error location finder circuit communicates with the iBMA circuit and identifies error locations. An error value computation circuit communicates with at least one of the error location finder circuit and the iBMA circuit and generates error values.Type: GrantFiled: November 26, 2002Date of Patent: May 23, 2006Assignee: Marvell International Ltd.Inventors: Zhan Yu, Weishi Feng
-
Patent number: 7010739Abstract: An error correcting Reed-Solomon decoder includes a syndrome calculator that calculates syndrome values. An error locator polynomial generator communicates with the syndrome calculator and generates an error locator polynomial. An error location finder communicates with at least one of the syndrome calculator and the error locator polynomial generator and generates error locations. An error values finder communicates with at least one of the syndrome calculator, the error location finder and the error locator polynomial generator and generates error values using an error value relationship that is not based on the traditional error evaluator polynomial. The error locator polynomial generator is an inversionless Berlekamp-Massey algorithm (iBMA), which calculates an error locator polynomial and a scratch polynomial. The error value relationship is based on the error locator polynomial and the scratch polynomial.Type: GrantFiled: November 26, 2002Date of Patent: March 7, 2006Assignee: Marvell International Ltd.Inventors: Weishi Feng, Zhan Yu
-
Patent number: 7006016Abstract: A data encoding system for a data stream comprises a data dependent scrambler that receives the data stream including K m-bit symbols, that selects a seed based on the K m-bit symbols, that scrambles the K m-bit symbols using the seed and that outputs a codeword including the scrambled K m-bit symbols and the seed. A DC control module receives a plurality of the codewords from the data dependent scrambler, selectively inverts selected ones of the plurality of codewords to reduce a difference between a total number of zeroes and total number of ones in the plurality of codewords and outputs an encoded data stream.Type: GrantFiled: September 10, 2004Date of Patent: February 28, 2006Assignee: Marvell International Ltd.Inventor: Weishi Feng
-
Patent number: 6897792Abstract: A data dependent scrambler (DDS) for a communications channel that transmits a user data sequence having a plurality of symbols includes a scrambler that generates a scrambled user data sequence that is based on the user data sequence and a seed. A first encoder selectively interleaves adjacent symbols in the scrambled user data sequence if an all-zero symbol is produced by bit interleaving. The first encoder identifies a pivot bit that is adjacent to the all-zero symbol if interleaving is performed and replaces the all-zero symbol with an all-one symbol if the pivot bit is zero.Type: GrantFiled: November 5, 2003Date of Patent: May 24, 2005Assignee: Marvell International Ltd.Inventor: Weishi Feng
-
Patent number: 6868517Abstract: Method and apparatus for detecting errors in data read from a data storage medium include an error correction step/device which receives at least one of (i) data and (ii) data with errors, from the data storage medium, and outputs an error sequence in a first order in the case where data with errors is received. A first CRC step/device receives the at least one of (i) data and (ii) data with errors from the data storage medium, and outputs a CRC checksum in a second order different from said first order. A second CRC step/device receives both the error sequence and the CRC checksum, and outputs another CRC checksum indicative of whether the correction device or step has generated a correct error sequence. Preferably, a first CRC is coupled parallel to a Reed-Soloman decoder, and a second CRC is coupled in series with the first CRC and so as to receive the output of the R-S decoder.Type: GrantFiled: April 8, 2002Date of Patent: March 15, 2005Assignee: Marvell International Ltd.Inventors: Weishi Feng, Liang Zhang, Zhan Yu