Patents by Inventor Christophe Mangin
Christophe Mangin 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).
-
Publication number: 20160014044Abstract: For allocating resources in a mesh communications network for setting up a data stream transmission from a talker device to a listener device via at least one node device, data paths being defined throughout the mesh communications network following a link-state routing protocol, each node device performs receiving, obtaining, determining, and checking. And when there are enough said resources, temporarily reserving and propagating. And when receiving for said data stream transmission a stream reservation response representative of a positive acknowledgement to said stream reservation request, effectively allocating and propagating.Type: ApplicationFiled: February 19, 2014Publication date: January 14, 2016Applicant: MITSUBISHI ELECTRIC CORPORATIONInventor: Christophe MANGIN
-
Patent number: 9166691Abstract: A device and method to allow to estimate drift induced by temperature variation in a network including a frequency splitter based on a Mach-Zehnder, or alike component, and to track it over time, by adding mirrors to the unused port of the Mach-Zehnder components. Doing so, both OLT and ONU are able to scan a band of frequencies. The frequency corresponding to the nominal wavelength of the component will be reflected on the other port on the same side of the Mach-Zehnder while other frequencies will go through the Mach-Zehnder to be reflected by the mirror and come back to the emitter. By measuring the reflected signal while scanning frequencies, the actual nominal wavelength of the component can be determined.Type: GrantFiled: December 9, 2011Date of Patent: October 20, 2015Assignee: MITSUBISHI ELECTRIC CORPORATIONInventors: Gwillerm Froc, Christophe Mangin
-
Patent number: 9136971Abstract: A method configuring an optical transmission interface of at least one first device transmitting optical signals on a path to a second device on a carrier wavelength substantially equal to a nominal wavelength of an optical band-pass filter present on the path, each first device including an optical reception interface to enable receiving optical signals from the second device, the second device including an optical reception interface to enable receiving optical signals output by the first optical band-pass filter and transmitted by any first device on a carrier wavelength substantially equal to the nominal wavelength, the optical transmission interface of each first device being initially configured for transmitting optical signals on a first carrier wavelength.Type: GrantFiled: February 20, 2012Date of Patent: September 15, 2015Assignee: MITSUBISHI ELECTRIC CORPORATIONInventors: Gwillerm Froc, Christophe Mangin
-
Patent number: 9130676Abstract: The invention relates to a method for performing a switch operation between a first Optical Line Terminal OLT and a second OLT in a Passive Optical Network PON, said first and second OLTs being connected to a plurality of Optical Network Units ONUs through at least one splitter. The invention enables to avoid the detection of a timestamp drift error during protection switching operation by modifying the threshold of the detection in the second OLT.Type: GrantFiled: January 17, 2013Date of Patent: September 8, 2015Assignee: MITSUBISHI ELECTRIC CORPORATIONInventor: Christophe Mangin
-
Publication number: 20130336657Abstract: A method configuring an optical transmission interface of at least one first device transmitting optical signals on a path to a second device on a carrier wavelength substantially equal to a nominal wavelength of an optical band-pass filter present on the path, each first device including an optical reception interface to enable receiving optical signals from the second device, the second device including an optical reception interface to enable receiving optical signals output by the first optical band-pass filter and transmitted by any first device on a carrier wavelength substantially equal to the nominal wavelength, the optical transmission interface of each first device being initially configured for transmitting optical signals on a first carrier wavelength.Type: ApplicationFiled: February 20, 2012Publication date: December 19, 2013Applicant: MITSUBISHI ELECTRIC CORPORATIONInventors: Gwillerm Froc, Christophe Mangin
-
Publication number: 20130279906Abstract: A device and method to allow to estimate drift induced by temperature variation in a network including a frequency splitter based on a Mac Zehnder, or alike component, and to track it over time, by adding mirrors to the unused port of the Mac Zehnder components. Doing so, both OLT and ONU are able to scan a band of frequencies. The frequency corresponding to the nominal wavelength of the component will be reflected on the other port on the same side of the Mac Zehnder while other frequencies will go through the Mac Zehnder to be reflected by the mirror and come back to the emitter.Type: ApplicationFiled: December 9, 2011Publication date: October 24, 2013Applicant: Mitsubishi Electric CorporationInventors: Gwillerm Froc, Christophe Mangin
-
Publication number: 20130195439Abstract: The invention relates to a method for performing a switch operation between a first Optical Line Terminal OLT and a second OLT in a Passive Optical Network PON, said first and second OLTs being connected to a plurality of Optical Network Units ONUs through at least one splitter. The invention enables to avoid the detection of a timestamp drift error during protection switching operation by modifying the threshold of the detection in the second OLT.Type: ApplicationFiled: January 17, 2013Publication date: August 1, 2013Inventor: Christophe MANGIN
-
Patent number: 8155108Abstract: A method for submitting a TCP segment stream to a in-line content analysis comprises: receiving from a TCP emitter a first TCP data segments stream corresponding to a given TCP connection between the TCP emitter and a TCP receiver; extracting and re-assembling a first byte stream from the TCP segment stream; passing a second byte stream, which is derived from the first byte stream, to a content analyser (200) adapted to perform a content analysis on the second byte stream; relaying to the TCP receiver a second TCP data segment stream with a byte stream content totally consistent with the byte stream passed to the content analyser. In order to keep the device (100) that implements the method as stealthy as possible, the second TCP data segment stream has the same number of data segments as the first TCP data segment stream, all segments in the second TCP data segment stream being of same size and having the same header as corresponding segments in the first TCP data segment stream.Type: GrantFiled: June 11, 2007Date of Patent: April 10, 2012Assignee: Mitsubishi Denki Kabushiki KaishaInventors: Christophe Mangin, Romain Rollet, David Mentre
-
Patent number: 8060593Abstract: A method of managing the allocation to TCP connections of memory blocks of a TCP re-assembly buffer in a TCP network device including a TCP Processing unit, comprises computing a function r(t) on a per connection basis, that measures the occupancy rate of the re-assembly buffer for the relevant TCP connection, and that is updated upon reception of any packet belonging to said TCP connection; and allowing the TCP connection to be closed and the corresponding memory blocks to be released based on a decision scheme that includes at least a first trigger set on the corresponding function r(t), which defines two different states for said function r(t).Type: GrantFiled: October 10, 2007Date of Patent: November 15, 2011Assignee: Mitsubishi Electric CorporationInventors: Christophe Mangin, Romain Rollet, David Mentre
-
Patent number: 7860108Abstract: There is disclosed a method of transferring input data packets from an upper layer to a lower layer of a stack of communication protocol layers, wherein the lower layer supports variable sized packets. One input data packet is received from the upper layer. This input data packet is segmented into several data segments the size of which is selected in a list of a given number N of predetermined segment size values. Signalling information is added to each data segment to form output data packets, which is delivered to a lower layer. Selection of the segment size value is performed according to an algorithm which minimizes the overall ineffective data and/or the overall signalling information. Thus, the transmission overhead generated by the segmentation scheme is reduced.Type: GrantFiled: October 20, 2004Date of Patent: December 28, 2010Assignee: Mitsubishi Denki Kabushiki KaishaInventors: Christophe Mangin, Romain Rollet
-
Patent number: 7804812Abstract: There is disclosed a method and a device of controlling transmission errors in a network comprising at least one transmitter, at least one receiver and a Radio Resource Management (RRM) unit for allocating transmission resource dedicated to the transmitter and/or the receiver. A FeedBacK (FBK) function is introduced to manage transmission of feedback information which is transmitted by the receiver to the transmitter to indicate transmission errors. The FBK function comprises a transmitter FBK instance and a receiver FBK instance. At first, the transmitter FBK instance transmits Protocol Data Units (PDUs) to the receiver FBK instance. The FBK function monitors transmission errors on the receiver FBK instance side and/or on the transmitter FBK instance side to determine a transmission quality level out of a set of given transmission quality levels.Type: GrantFiled: December 21, 2004Date of Patent: September 28, 2010Assignee: Mitsubishi Denki Kabushiki KaishaInventors: Christophe Mangin, Romain Rollet
-
Patent number: 7805538Abstract: A method and a device of address management in a network which comprises a plurality of devices, including at least one STAtion (STA), and at least one Address Resource Management (ARM) unit. Each of the devices includes a stack of communication protocol layers comprising at least an Upper Layer (UL) and a Lower Layer (LL). The UL of each of the devices has a respective UL address within the network. The ARM unit associates, respectively disassociates, some of the devices in at least one LL network. The devices of a given LL network are allowed to exchange together data packets emitted and received at the LL level. The ARM unit allocates, respectively releases, one given LL address per each associated device, respectively disassociated device, in a given LL network. Besides, the ARM unit handles one ARM Translation Table (TT) per LL network. The TT comprises one entry per device of a given LL network and each entry comprises the UL address and the LL address.Type: GrantFiled: December 10, 2004Date of Patent: September 28, 2010Assignee: Mitsubishi Denki Kabushiki KaishaInventors: Christophe Mangin, Romain Rollet
-
Patent number: 7613203Abstract: There is disclosed a method and a device for managing a transmission medium shared between a set of wireless network devices which comprises a plurality of stations and a Radio Resource Management (RRM) unit. The method is based on a time division scheme carried out by said RRM unit. Time is divided in a series of Frame Time Intervals (FTIs), which are split into a given number n of Time Intervals (TIs), each TI having a respective duration (D). The ownership of each TI is dedicated to a group of TI owner devices (TOG), selected among the set of devices. Each TOG comprises a given number k of transmitting devices, which are allowed to transmit over the transmission medium during the TI considered, and a given number m of receiving devices, which listen to the transmission medium during this TI, where m and k are strictly positive integer.Type: GrantFiled: October 13, 2004Date of Patent: November 3, 2009Assignee: Mitsubishi Denki Kabushiki KaishaInventors: Christophe Mangin, Romain Rollet
-
Patent number: 7525970Abstract: A method of dynamic transmission resource allocation that operates in a wireless network. The wireless network includes a plurality of mobile terminals each managing a given number of different data flows of protocol data units that are classified according to a set of flow types, a radio resource management unit for allocating transmission resources on a per protocol data unit basis, a plurality of schedulers operating according to a specific set of rules, and a selector for sharing the transmission resource among the mobile terminals.Type: GrantFiled: December 21, 2004Date of Patent: April 28, 2009Assignee: Mitsubishi Denki Kabushiki KaishaInventors: Christophe Mangin, Romain Rollet
-
Patent number: 7519084Abstract: There is disclosed a method of transmitting data packets from a transmitter to a receiver each having a stack of communication protocol layers which comprises a data Link Layer (LL), by using a Segmentation And Re-assembly (SAR) mechanism implemented within a SAR layer intermediate between the LL and an upper layer, as well as an error recovery mechanism implemented within the LL. The SAR mechanism performs, in a transmission direction, the function of segmenting Parent Packets (PPs) handled by the upper layer into consecutive segments of shorter length and the function of providing some segmentation information. The segmentation information comprises, for each segment, a two-level sequence numbering comprising a PP Sequence Number (PPSN) which identifies the PP which said segment belongs to, and a Segment Sequence Number (SSN) which identifies the rank of the segment within the PP.Type: GrantFiled: October 8, 2004Date of Patent: April 14, 2009Assignee: Mitsubishi Denki Kabushiki KaishaInventors: Christophe Mangin, Romain Rollet
-
Patent number: 7453799Abstract: There is disclosed a method and a device of method of dynamic ARQ windows management in a network for controlling data link transmission. The method comprises an ARQ management entity selecting a given operational control mode, and a transmitter sending one or more PDUs of a given data link to a receiver. The receiver stores received PDUs in a shared memory. A receiver ARQ instance sends back corresponding ARQ-ACK message including a specific information comprising selected operational control mode and/or a flow control data. It further removes acknowledged PDUs from the shared memory when delivering them to an upper layer. The transmitter ARQ instance continues or stops PDUs transmission upon reception of the ARQ-ACK message, based on the specific information comprised therein.Type: GrantFiled: February 23, 2005Date of Patent: November 18, 2008Assignee: Mitsubishi Denki Kabushiki KaishaInventors: Christophe Mangin, Romain Rollet
-
Publication number: 20080256247Abstract: A method of managing the allocation to TCP connections of memory blocks of a TCP re-assembly buffer in a TCP network device including a TCP Processing unit, comprises computing a function r(t) on a per connection basis, that measures the occupancy rate of the re-assembly buffer for the relevant TCP connection, and that is updated upon reception of any packet belonging to said TCP connection; and allowing the TCP connection to be closed and the corresponding memory blocks to be released based on a decision scheme that includes at least a first trigger set on the corresponding function r(t), which defines two different states for said function r(t).Type: ApplicationFiled: October 10, 2007Publication date: October 16, 2008Applicant: MITSUBISHI ELECTRIC CORPORATIONInventors: Christophe Mangin, Romain Rollet, David Mentre
-
Publication number: 20070291662Abstract: A method for submitting a TCP segment stream to a in-line content analysis comprises: receiving from a TCP emitter a first TCP data segments stream corresponding to a given TCP connection between the TCP emitter and a TCP receiver; extracting and re-assembling a first byte stream from the TCP segment stream; passing a second byte stream, which is derived from the first byte stream, to a content analyser (200) adapted to perform a content analysis on the second byte stream; relaying to the TCP receiver a second TCP data segment stream with a byte stream content totally consistent with the byte stream passed to the content analyser. In order to keep the device (100) that implements the method as stealthy as possible, the second TCP data segment stream has the same number of data segments as the first TCP data segment stream, all segments in the second TCP data segment stream being of same size and having the same header as corresponding segments in the first TCP data segment stream.Type: ApplicationFiled: June 11, 2007Publication date: December 20, 2007Applicant: MITSUBISHI ELECTRIC CORPORATIONInventors: Christophe Mangin, Romain Rollet, David Mentre
-
Publication number: 20050265389Abstract: There is disclosed a method of transmitting data packets from a transmitter to a receiver -each having a stack of communication protocol layers which comprises a data Link Layer (LL), by using a Segmentation And Re-assembly (SAR) mechanism implemented within a SAR layer intermediate between the LL and an upper layer, as well as an error recovery mechanism implemented within the LL. The SAR mechanism performs, in a transmission direction, the function of segmenting Parent Packets (PPs) handled by the upper layer into consecutive segments of shorter length and the function of providing some segmentation information. The segmentation information comprises, for each segment, a two-level sequence numbering comprising a PP Sequence Number (PPSN) which identifies the PP which said segment belongs to, and a Segment Sequence Number (SSN) which identifies the rank of the segment within the PP.Type: ApplicationFiled: October 8, 2004Publication date: December 1, 2005Inventors: Christophe Mangin, Romain Rollet
-
Patent number: 6967927Abstract: Data flows generated in accordance with a connected mode protocol and formatted in packets in accordance with a non-connected mode protocol are transmitted via an ATM network. A set of virtual circuits of the ATM network is assigned in advance to each pair of access points of the ATM network, without allocating transmission rate resources to them. When an access point receives a request, formulated in accordance with the connected mode protocol, to establish a connection between a source address and a destination address accessible via another access point of the ATM network, an available virtual circuit is selected from the set assigned to this pair of access points and a resource management cell containing a request to activate the virtual circuit is transmitted on the selected virtual circuit. When this cell is received by a node of the ATM network located on the selected virtual circuit, a transmission rate resource is allocated, if it is available, to the selected virtual circuit.Type: GrantFiled: November 21, 2000Date of Patent: November 22, 2005Assignee: France TelecomInventors: Olivier Dugeon, Fabrice Guillemin, Christophe Mangin