Patents by Inventor Xuming Zhang
Xuming Zhang 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: 9641196Abstract: The present invention relates to communication field, disclosing a data block interleaving and deinterleaving method and apparatus for communication equipments. In the present invention, a recursive method for calculating interleaver or deinterleaver addresses for existing power line communication standards is proposed. The complex modulo operation is simplified to a series of Add-Compare-Subtract operations. Therefore, the hardware implementation complexity is significantly reduced.Type: GrantFiled: January 14, 2015Date of Patent: May 2, 2017Assignee: Hi-Trend Technology (Shanghai) Co., Ltd.Inventor: Xuming Zhang
-
Publication number: 20150263765Abstract: The present invention relates to communication field, disclosing a data block interleaving and deinterleaving method and apparatus for communication equipments. In the present invention, a recursive method for calculating interleaver or deinterleaver addresses for existing power line communication standards is proposed. The complex modulo operation is simplified to a series of Add-Compare-Subtract operations. Therefore, the hardware implementation complexity is significantly reduced.Type: ApplicationFiled: January 14, 2015Publication date: September 17, 2015Inventor: Xuming ZHANG
-
Patent number: 8966988Abstract: Ultra-miniature surface-mountable optical pressure sensor is constructed on an optical fiber. The sensor design utilizes an angled fiber tip which steers the optical axis of the optic fiber by 90°. The optical cavity is formed on the sidewall of the optic fiber. The optical cavity may be covered with a polymer-metal composite diaphragm to operate as a pressure transducer, Alternatively, a polymer-filled cavity may be constructed which does not need a reflective diaphragm. The sensor exhibits a sufficient linearity over the broad pressure range with a high sensitivity. The sensitivity of the sensor may he tuned by controlling the thickness of the diaphragm. Methods of batch production of uniform device-to-device optical pressure sensors of co-axial and cross-axial configurations are presented.Type: GrantFiled: April 5, 2012Date of Patent: March 3, 2015Assignee: University of MarylandInventors: Miao Yu, Hyungdae Bae, Xuming Zhang
-
Publication number: 20120210797Abstract: Ultra-miniature surface-mountable optical pressure sensor is constructed on an optical fiber. The sensor design utilizes an angled fiber tip which steers the optical axis of the optic fiber by 90°. The optical cavity is formed on the sidewall of the optic fiber. The optical cavity may be covered with a polymer-metal composite diaphragm to operate as a pressure transducer. Alternatively, a polymer-filled cavity may be constructed which does not need a reflective diaphragm. The sensor exhibits a sufficient linearity over the broad pressure range with a high sensitivity. The sensitivity of the sensor may be tuned by controlling the thickness of the diaphragm. Methods of batch production of uniform device-to-device optical pressure sensors of co-axial and cross-axial configurations are presented.Type: ApplicationFiled: April 5, 2012Publication date: August 23, 2012Applicant: UNIVERSITY OF MARYLANDInventors: Miao YU, Hyungdae BAE, Xuming ZHANG
-
Patent number: 8151648Abstract: Ultra-miniature surface-mountable Fabry-Perot pressure sensor is constructed on an optical fiber which utilizes a 45° angled fiber tip covered with a reflective layer which steers the optical axis of the fiber by 90°. The Fabry-Perot cavity is formed on the sidewall of the fiber and a polymer-metal composite diaphragm is formed on the top of the Fabry-Perot cavity to operate as a pressure transducer. The sensor exhibits a sufficient linearity over the broad pressure range with a high sensitivity. The sensitivity of the sensor may be tuned by controlling the thickness of the diaphragm. The sensor may be used in a wide range of applications, including reliable in vivo low invasive pressure measurements of biological fluids, single sensor systems, as well as integral spatial-division-multiplexing sensor networks. Methods of batch production of uniform device-to-device Fabry-Perot pressure sensors of co-axial and cross-axial configurations are presented.Type: GrantFiled: August 3, 2010Date of Patent: April 10, 2012Assignee: University of MarylandInventors: Miao Yu, Hyungdae Bae, Xuming Zhang
-
Patent number: 7925092Abstract: A hand-writing input device comprises: an input surface, a guidance device provided on the input surface along a predetermined track to guide a finger or pen along the track; a plurality of switch elements provided at particular positions on the track. When a particular character is inputted, the triggered switch element outputs the corresponding signal, and the combinations of the signals outputted by the switch elements correspond to the inputted characters. Since hand-writing input is performed by guiding a finger or pen along a predetermined track, only few switch elements are needed, thereby simplifying hand-writing recognition, reducing software requirements and increasing recognition accuracy.Type: GrantFiled: November 3, 2003Date of Patent: April 12, 2011Assignee: GM Global Technology Operations LLCInventors: Ziyan Wang, Xuming Zhang
-
Patent number: 7335901Abstract: A light delivery device (10) having a light source (12) and a variable aperture unit (18) is temporarily connected by a light guide (24; 24, 26) to a radiometer (38) for detecting irradiance of the delivered light. The light delivery device has a memory (30) for storing irradiance levels. The light delivery device is calibrated by adjusting the aperture to each of a series of predetermined settings, obtaining from the radiometer a corresponding series of delivered light irradiance levels measured thereby, storing the irradiance levels and aperture settings in memory, and applying a best fit algorithm to the irradiance measurements and aperture settings. Thereafter, a desired irradiance level can be set by selecting the best fit aperture setting. Output intensity levels may be measured at the same time as the irradiance levels and used to compensate for light source degradation when setting a desired irradiance level.Type: GrantFiled: December 21, 2005Date of Patent: February 26, 2008Assignee: EXFO Photonic Solutions Inc.Inventors: Roy Kayser, Graeme Rice, Xuming Zhang
-
Patent number: 7212582Abstract: In one embodiment, a device includes but is not limited to: a precoder having a precoder input and a precoder output; a spectral shaper, having a spectral-shaper transfer function and being operably coupled internal to said precoder; and a precoder filter, internal to the precoder, the precoder filter having a transfer function which provides substantially the inverse of the spectral-shaper transfer function.Type: GrantFiled: November 20, 2002Date of Patent: May 1, 2007Assignee: Mindspeed Technologies, Inc.Inventor: Xuming Zhang
-
Patent number: 7203241Abstract: Systems compensate for the presence of digital impairments using adaptive constellation techniques. Sample magnitudes are estimated from received samples transmitted during a training period using an algorithm based on minimizing the sum of squares of the estimated errors, afterwhich a suitable constellation is derived. In one embodiment, an optimum least-squares algorithm is employed. In an alternate embodiment, a least mean squares (LMS) or normalized LMS algorithm is performed.Type: GrantFiled: October 23, 2003Date of Patent: April 10, 2007Assignee: Silicon Laboratories Inc.Inventors: Xuming Zhang, Sverrir Olafsson
-
Publication number: 20070063967Abstract: A hand-writing input device comprises: an input surface, a guidance device provided on the input surface along a predetermined track to guide a finger or pen along the track; a plurality of switch elements provided at particular positions on the track. When a particular character is inputted, the triggered switch element outputs the corresponding signal, and the combinations of the signals outputted by the switch elements correspond to the inputted characters. Since hand-writing input is performed by guiding a finger or pen along a predetermined track, only few switch elements are needed, thereby simplifying hand-writing recognition, reducing software requirements and increasing recognition accuracy.Type: ApplicationFiled: November 3, 2003Publication date: March 22, 2007Inventors: Ziyan Wang, Xuming Zhang
-
Patent number: 7187719Abstract: A method and for optimizing bit rate throughput in a digital communication system is provided. The method includes establishing a relationship between signal to noise ratio and plural symbol rates for a particular constellation size. The method also includes determining noise power spectral density (N(f)), wherein N(f) is determined during a silent period of line probing; determining Xk(f), wherein Xk(f) is determined by turning on a remote station transmit signal, after N(f) has been measured and determining residual echo Ek(f), wherein Ek(f) is determined by turning on a central station echo canceller.Type: GrantFiled: June 18, 2003Date of Patent: March 6, 2007Assignee: Mindspeed Technologies, Inc.Inventor: Xuming Zhang
-
Patent number: 7173963Abstract: A technique for identifying the encoding law utilized by a central office codec may be implemented in a receive modem. The encoding law, which is typically dictated by the country in which the central office is located, is employed to generate a plurality of transmission levels during an initialization period associated with the modem system. The receive modem analyzes a number of these transmission levels to determine whether the levels have certain characteristics associated with the particular encoding law followed by the central office codec. When the receive modem detects the codec type, it may transmit a suitable identifier back to the transmit modem.Type: GrantFiled: July 2, 2003Date of Patent: February 6, 2007Assignee: Silicon Laboratories Inc.Inventors: Xuming Zhang, Zhenyu Zhou
-
Publication number: 20060139722Abstract: A light delivery device (10) having a light source (12) and a variable aperture unit (18) is temporarily connected by a light guide (24; 24, 26) to a radiometer (38) for detecting irradiance of the delivered light. The light delivery device has a memory (30) for storing irradiance levels. The light delivery device is calibrated by adjusting the aperture to each of a series of predetermined settings, obtaining from the radiometer a corresponding series of delivered light irradiance levels measured thereby, storing the irradiance levels and aperture settings in memory, and applying a best fit algorithm to the irradiance measurements and aperture settings. Thereafter, a desired irradiance level can be set by selecting the best fit aperture setting. Output intensity levels may be measured at the same time as the irradiance levels and used to compensate for light source degradation when setting a desired irradiance level.Type: ApplicationFiled: December 21, 2005Publication date: June 29, 2006Inventors: Roy Kayser, Graeme Rice, Xuming Zhang
-
Publication number: 20050170065Abstract: The present invention is related to a new soybean noodle. This noodle comprises 4 to 10 parts by weight of soybean powder and 2 to 4 parts by weight of wet bean-curd residue with water content of 60%˜80%. Moreover, the noodle may also comprise 0.5 to 2 parts by weight of soybean protein powder. The processing steps are as follows: peeling, defatting and crushing the soybean into powder of 70-120 mesh size; then mixing uniformly the treated powder with wet bean-curd residue and soybean protein powder; and finally extruding and aging the mixture to form finished noodle product using the extruder under the condition of high-temperature and high-pressure. This noodle features rich and balanced in nutrient. Besides rich plant protein, the said noodle comprises dietary fiber with the content up to 3%˜12%. The said noodle tastes delicious without soybean odor. When it is cooked, it is non-sticky and not easy to break. The noodle is suitable for all kinds of consumer groups.Type: ApplicationFiled: April 23, 2004Publication date: August 4, 2005Inventors: Cunku Xi, Yuhua Zhang, Xuming Zhang
-
Publication number: 20040258172Abstract: A method and for optimizing bit rate throughput in a digital communication system is provided. The method includes establishing a relationship between signal to noise ratio and plural symbol rates for a particular constellation size. The method also includes determining noise power spectral density (N(f)), wherein N(f) is determined during a silent period of line probing; determining Xk(f), wherein Xk(f) is determined by turning on a remote station transmit signal, after N(f) has been measured and determining residual echo Ek(f), wherein Ek(f) is determined by turning on a central station echo canceller.Type: ApplicationFiled: June 18, 2003Publication date: December 23, 2004Inventor: Xuming Zhang
-
Patent number: 6829296Abstract: The time domain equalizer filter of the present invention increases the data rate of a communications system while shortening a channel. Such a time domain equalizer filter is spectrally flat and the central tap is constrained to a non-zero real number. The error between the data filtered by the time domain equalizer filter and the data filtered by a target filter is reduced by adapting the time domain equalizer filter and the target filter. Adaptation of the time domain equalizer filter and the target filter may be accomplished by calculating new tap values for each filter while constraining the central tap of the time domain equalizer filter. In this way, the error between the data filtered by the time domain equalizer filter and the data filtered by the target filter may be reduced, so that a shortened channel may be provided. Accordingly, by shortening the channel and balancing the use of a spectrally flat time domain equalizer filter, the data rate of the communications system may be increased.Type: GrantFiled: September 20, 2000Date of Patent: December 7, 2004Assignee: Mindspeed Technologies, Inc.Inventors: Markos G. Troulis, Evangelos Petsalis, Xuming Zhang
-
Patent number: 6788843Abstract: The invention features the drawbridge assembly and its applications in optical switches, optical crossconnects, optical add/drop multiplexers and variable optical attenuators. In optical switches and optical crossconnects, an array of the drawbridge assemblies can be used to redirect the multiple input lights to multiple outputs. In add/drop multiplexers, the drawbridge assemblies can select the light channels to be added and dropped. In the attenuator embodiment, a vertical mirror is inserted into two fibers, the first one as the input and the second one as output. The drawbridge assembly controls the position of the vertical mirror for blocking a certain portion of the light and enabling the attenuation. The continuous change of the mirror position results in variable attenuation. A series of VOA form a multi-channel VOA system on a single substrate.Type: GrantFiled: August 19, 2002Date of Patent: September 7, 2004Assignee: Nanyang Technological UniversityInventors: Ai Qun Liu, Vadakke Matham Murukeshan, Xuming Zhang, Chao Lu
-
Patent number: 6785862Abstract: A convolutional interleaver includes an interleaver memory partitioned into a plurality of circular buffers, wherein each of the circular buffers has associated write pointers and read pointers, and wherein the interleaver is configured to selectively read symbols from an input vector and store the input symbols in the interleaver memory in accordance with the write pointers, and to selectively read symbols from the interleaver memory to form an output vector in accordance with the read pointers. In one aspect, symbols are written to the interleaver prior to reading; in another, the position of the write pointer corresponds to the position of the read pointer within the circular buffer, and symbols are read from said interleaver memory prior to writing. In another aspect, a de-interleaver applies the concepts and algorithms described above in an inverse manner.Type: GrantFiled: August 14, 2001Date of Patent: August 31, 2004Assignee: Mindspeed Technologies, Inc.Inventor: Xuming Zhang
-
Publication number: 20040096022Abstract: In one embodiment, a device includes but is not limited to: a precoder having a precoder input and a precoder output; a spectral shaper, having a spectral-shaper transfer function and being operably coupled internal to said precoder; and a precoder filter, internal to the precoder, the precoder filter having a transfer function which provides substantially the inverse of the spectral-shaper transfer function.Type: ApplicationFiled: November 20, 2002Publication date: May 20, 2004Applicant: Conexant Systems, Inc.Inventor: Xuming Zhang
-
Patent number: 6721279Abstract: Systems compensate for the presence of digital impairments using adaptive constellation techniques. Sample magnitudes are estimated from received samples transmitted during a training period using an algorithm based on minimizing the sum of squares of the estimated errors, after which a suitable constellation is derived. In one embodiment, an optimum least-squares algorithm is employed. In an alternate embodiment, a least mean squares (LMS) or normalized LMS algorithm is performed.Type: GrantFiled: February 2, 2000Date of Patent: April 13, 2004Assignee: PCTEL, Inc.Inventors: Xuming Zhang, Sverrir Olafsson