Abstract: An image super-resolution method based on a knowledge distillation compression model and a device thereof are disclosed. A small student network model is cascaded into a teacher network with high performance to better complete knowledge distillation, so that the performance of a student network can gradually approach the teacher network, and then the compression of a super-resolution network is completed. Using a distillation strategy of the present disclosure not only avoids manually designing feature conversion between different networks to align, but also greatly reduces the optimization difficulty of the student network. In order to alleviate the problem of inefficient distillation caused by a representation gap between teachers and students, the present disclosure regards a similarity relationship between layers of teachers as knowledge, so that students can learn the similarity relationship of teachers in their own space instead of directly imitating complex features of teachers.

Abstract: A signal processing method and system includes a baseband signal baseband signal processing module configured to perform slow envelope processing on a first signal, to obtain an envelope value E(n) of the first signal on which the slow envelope processing has been performed, obtain a phase value ?(n) based on E(n), where ?(n) and E(n) are in a linear relationship, and separate the first signal into a second signal and a third signal based on ?(n), where a phase difference between the second signal and the third signal is 2 ?(n), an amplifier configured to amplify the second signal and the third signal, and a synthesizer is configured to combine the amplified second signal and third signal to obtain a fourth signal.

Abstract: A signal processing system includes n paths of load modulation modules and a combination module, where then paths of load modulation modules are connected in parallel, an output end of each path of load modulation module is connected to an input end of the combination module, and n is an integer greater than 1; the n paths of load modulation modules include one path of main power amplification module and (n-1) paths of auxiliary power amplification modules, and the auxiliary power amplification modules are turned on when power values of signals received by input ends of the load modulation modules are greater than a first threshold; and the main power amplification module includes two outphasing power amplification units, and each path of auxiliary power amplification module includes two outphasing power amplifier arrays or one digital polar power amplifier array.

Abstract: A method for transmitting upstream data in a cable network is provided. The method includes a cable modem (CM) receiving a generic service identifier (G-SID) included in a corresponding information element (IE) in a data over cable service interface specification (DOCSIS) upstream bandwidth allocation MAP message. The G-SID is a unique identifier assigned to the CM. The method includes the CM transmitting the upstream data for one or more service flows of the CM via at least a portion of the CM bandwidth allocation.

Abstract: A signal processing method and system includes a baseband signal baseband signal processing module configured to perform slow envelope processing on a first signal, to obtain an envelope value E(n) of the first signal on which the slow envelope processing has been performed, obtain a phase value ?(n) based on E(n), where ?(n) and E(n) are in a linear relationship, and separate the first signal into a second signal and a third signal based on ?(n), where a phase difference between the second signal and the third signal is 2 ?(n), an amplifier configured to amplify the second signal and the third signal, and a synthesizer is configured to combine the amplified second signal and third signal to obtain a fourth signal.

Abstract: The communication method includes: sending, by a CMTS (cable modem terminals stem), a reference signal within a first time period by using a valid subcarrier in a first spectrum, where the first time period is a time period within which none of CMs (cable modems) connected to the CMTS sends a signal in the first spectrum; performing, by the CMTS based on an interfering signal received by the CMTS within the first time period, channel interference estimation on a channel that is used during signal reception of the CMTS and that occupies the first spectrum, where the interfering signal is a signal obtained after the reference signal passes through the channel; and canceling, by the CMTS based on a result of the channel interference estimation, interference from a first signal received by the CMTS to a second signal received by the CMTS.

Abstract: The present invention provides a power adjustment method, a network management system, a CMTS, and a CM. The power adjustment method includes: obtaining, by a network management system, a first SNRCM between two CMs connected to a CMTS, where the first SNRCM represents a signal ratio of a signal sent by the CMTS to a first CM and received by the first CM to a signal sent by a second CM to the CMTS and received by the first CM; and sending, by the network management system, indication information to the CMTS based on the first SNRCM, and a first relationship between an SNRCM and a transmit power at which the CMTS sends a signal to the CM.

Abstract: An echo cancellation training method in a cable modern is provided, comprising receiving an echo cancellation service ID (EC SID) assigned for echo cancellation training, the cable modem receiving the EC SID from a cable head end in an information element (IE) in a map (MAP or P-MAP) defining a time and frequency of upstream transmission opportunities for the cable modem, the EC SID establishing when the cable modem may perform the echo cancellation training in a downstream channel allocated to the cable modem, transmitting an echo cancellation test signal upstream to the cable head end via an upstream channel allocated to the cable modem and in a time slot specified by the received EC SID in the map, and measuring noise in one or more downstream channels allocated to the cable modem, with the noise resulting from the transmission of the echo cancellation test signal on the upstream channel.

Abstract: A method and an apparatus for determining a time offset are disclosed. The method includes: obtaining, by a device at a head end, a time-domain signal based on a received signal; and then determining a time offset based on values of peak-to-average ratios of a preset quantity of symbols starting from a qth symbol in the time-domain signal, where a peak-to-average ratio of the qth symbol is greater than a preset threshold. A new method for determining a time offset is provided, and takes advantages that the time-domain signal obtained by the device at the head end has stronger capabilities of resisting interference such as frequency offset and phase noise. The method for determining a time offset can be applied both to initial ranging and periodic ranging.

Abstract: Embodiments of the present invention provide a multicast profile switching method. In the method, using a multicast profile management module, a multicast profile switching request message is received from a cable modem termination system CMTS. A multicast profile having a maximum rate and commonly supported by all current multicast group members is determined. When the multicast profile having the maximum rate is different from a current multicast profile that is running, a multicast profile switching instruction is sent to the CMTS, where the multicast profile switching instruction is used to instruct the CMTS to perform multicast profile switching.

Abstract: A cable modem (CM), a hybrid fiber coaxial (HFC) network fault locating system, and a fault detection method are provided. The CM can not only detect intrusion noise outside a carrier and white noise, but also detect other fault problems, such as non-linear distortion, pulse noise, intrusion noise on a carrier, and linear distortion. The system includes a direction-selective apparatus. The apparatus can allow only an upstream signal entered from an output port of the apparatus to be transmitted to a CM to which a branch port of the apparatus is connected, without allowing an upstream signal entered from an input port of the apparatus to be transmitted to the CM to which the branch port of the apparatus is connected, so that once detecting a fault from the upstream signal, the CM can accurately determine, through locating, that the fault in an HFC network is located on an output port side of the apparatus.

Abstract: A method for transmitting upstream data in a cable network is provided. The method includes a cable modem (CM) receiving a generic service identifier (G-SID) included in a corresponding information element (IE) in a data over cable service interface specification (DOCSIS) upstream bandwidth allocation MAP message. The G-SID is a unique identifier assigned to the CM. The method includes the CM transmitting the upstream data for one or more service flows of the CM via at least a portion of the CM bandwidth allocation.

Abstract: The present disclosure provides example channel parameter estimation methods and related apparatuses. In one example method, a cable modem (CM) obtains a second data frame through a downlink channel between the CM and a cable modem termination system (CMTS). The second data frame is generated after a first data frame sent by the CMTS is modulated through the downlink channel, a zero-bit-loaded symbol of the first data frame includes no valid codeword, but includes a zero-bit-loaded NCP and an idle codeword identified by the zero-bit-loaded NCP, and the idle codeword includes a data sequence used for channel parameter estimation. The CM can estimate a channel parameter of the downlink channel based on a data sequence in an obtained zero-bit-loaded symbol of the second data frame and the data sequence in the zero-bit-loaded symbol of the first data frame.

Abstract: A method, system, and apparatus for calibration for cable modem (CM) echo cancellation (EC) training in full duplex (FDX) coaxial networks is provided. A CM receives a current resource block assignment message (RBA) from a cable modem termination system (CMTS), and determines if designated channels available for assignment as downstream (DS) channels or upstream (US) channels have changed assignment values from a previously received RBA, and if EC training has been performed on the designated channels for the current RBA. If the designated channels have changed assignment values and EC training has not been performed on the designated channels for the current RBA, the CM performs EC training on the designated channels for the current RBA.

Abstract: Embodiments of the present disclosure provide an FEC coding and decoding method and device, and a system. A transmit end determines a forward error correction FEC coding type according to a length of to-be-coded data in burst data, and performs coding according to the determined FEC coding type. A receive end determines a forward error correction FEC decoding type according to a length of to-be-decoded data in burst data, and performs decoding according to the determined FEC decoding type. The FEC coding and decoding method provided in the embodiments of the present disclosure improves utilization of a communication resource is improved, and saves a communication resource.

Abstract: Disclosed herein is a multi-stage echo cancellation scheme. The disclosed embodiments include an apparatus and method for monitoring and canceling echoes greater than 25 dB in a coaxial cable plant. The method includes obtaining echo channel estimate coefficients from a cable node. The method determines a location and strength of each partial echo in an impulse response using the echo channel estimate coefficients. Optionally, the method determines a frequency response of each partial echo in the impulse response.

Abstract: The present disclosure provides example channel parameter estimation methods and related apparatuses. In one example method, a cable modem (CM) obtains a second data frame through a downlink channel between the CM and a cable modem termination system (CMTS). The second data frame is generated after a first data frame sent by the CMTS is modulated through the downlink channel, a zero-bit-loaded symbol of the first data frame includes no valid codeword, but includes a zero-bit-loaded NCP and an idle codeword identified by the zero-bit-loaded NCP, and the idle codeword includes a data sequence used for channel parameter estimation. The CM can estimate a channel parameter of the downlink channel based on a data sequence in an obtained zero-bit-loaded symbol of the second data frame and the data sequence in the zero-bit-loaded symbol of the first data frame.

Abstract: The present invention provides a power adjustment method, a network management system, a CMTS, and a CM. The power adjustment method includes: obtaining, by a network management system, a first SNRCM between two CMs connected to a CMTS, where the first SNRCM represents a signal ratio of a signal sent by the CMTS to a first CM and received by the first CM to a signal sent by a second CM to the CMTS and received by the first CM; and sending, by the network management system, indication information to the CMTS based on the first SNRCM, and a first relationship between an SNRCM and a transmit power at which the CMTS sends a signal to the CM.

Abstract: A cable modem (CM), a hybrid fiber coaxial (HFC) network fault locating system, and a fault detection method are provided. The CM can not only detect intrusion noise outside a carrier and white noise, but also detect other fault problems, such as non-linear distortion, pulse noise, intrusion noise on a carrier, and linear distortion. The system includes a direction-selective apparatus. The apparatus can allow only an upstream signal entered from an output port of the apparatus to be transmitted to a CM to which a branch port of the apparatus is connected, without allowing an upstream signal entered from an input port of the apparatus to be transmitted to the CM to which the branch port of the apparatus is connected, so that once detecting a fault from the upstream signal, the CM can accurately determine, through locating, that the fault in an HFC network is located on an output port side of the apparatus.

Abstract: The present disclosures describes a communication method, a CMTS, and a CM. The communication method includes: sending, by a CMTS, a reference signal within a first time period by using a valid subcarrier in a first spectrum, where the first time period is a time period within which none of CMs connected to the CMTS sends a signal in the first spectrum; performing, by the CMTS based on an interfering signal received by the CMTS within the first time period, channel interference estimation on a channel that is used during signal reception of the CMTS and that occupies the first spectrum, where the interfering signal is a signal obtained after the reference signal passes through the channel; and canceling, by the CMTS based on a result of the channel interference estimation, interference from a first signal received by the CMTS to a second signal received by the CMTS.