Abstract: A training method based on an improved protograph neural decoder includes the following steps: a to-be-trained decoding network is constructed based on an initial variable sub-network layer, an initial check sub-network layer and a preset shuffled belief-propagation (BP) sub-network layer; the initial variable sub-network layer, the initial check sub-network layer and the preset shuffled BP sub-network layer are updated and trained by calculating log-likelihood ratio (LLR) based on a preset mean square error loss function and a preset decoder objective function to obtain a target protograph neural decoder; and the preset mean square error loss function is configured to calculate a loss value between output information of the check sub-network layer and the preset shuffled BP sub-network layer. The target protograph neural decoder includes an optimized variable sub-network layer, an optimized check sub-network layer and an optimized shuffled BP sub-network layer. A training device is also provided.

Abstract: Disclosed are radionuclide microspheres and a preparation method therefor and an application thereof. The radionuclide microspheres include at least one or more radionuclides and are microspheres loading the radionuclides. The preparation method includes: by using an improved emulsion polymerization process, forming a prepolymer intermediate by a structural monomer, a functional monomer, and a vinyl crosslinker; adding one or more radionuclides for coordination polymerization with the prepolymer intermediate to form nuclear particles; and adding one or more small molecule monomers or high molecular materials for secondary polymerization to obtain the radioactive microspheres. The radionuclide microspheres are negatively charged porous microspheres which can load both chemotherapeutic drugs and polypeptide biological drugs. The microspheres may be implanted via vascular intervention and percutaneous puncture and may be used for treating solid malignant tumors such as liver cancer and lung cancer.

Abstract: A training method based on an improved protograph neural decoder includes the following steps. A to-be-trained decoding network is constructed based on an initial variable sub-network layer, an initial check sub-network layer and a preset shuffled belief-propagation (BP) sub-network layer. The initial variable sub-network layer, the initial check sub-network layer and the preset shuffled BP sub-network layer are updated and trained by calculating log-likelihood ratio (LLR) based on a preset mean square error loss function and a preset decoder objective function to obtain a target protograph neural decoder. The preset mean square error loss function is configured to calculate a loss value between output information of the check sub-network layer and the preset shuffled BP sub-network layer. The target protograph neural decoder includes an optimized variable sub-network layer, an optimized check sub-network layer and an optimized shuffled BP sub-network layer. A training device is also provided.

Abstract: A method for optimizing a Polar-RNNA quantizer of MLC NAND flash based on deep learning comprises the following steps: Step S1: transforming an MLC flash detection task into a deep learning task, and obtaining three hard-decision read thresholds based on a neural network; Step S2: expanding six soft-decision read thresholds based on the three hard-decision read thresholds; Step S3: constructing an LLR mapping table, and obtaining new LLR information of MLC flash based on the LLR mapping table; Step S4: symmetrizing an MLC flash channel, and performing density evolution; and Step S5: optimizing the soft-decision read thresholds based on a genetic algorithm to obtain an optimal quantization interval. According to the invention, polar codes can be directly used for MLC flash channels without the arduous work of MLC flash channel modeling, so that the reliability of MLC flash is effectively improved.

Abstract: A method for optimizing a Polar-RNNA quantizer of MLC NAND flash based on deep learning comprises the following steps: Step S1: transforming an MLC flash detection task into a deep learning task, and obtaining three hard-decision read thresholds based on a neural network; Step S2: expanding six soft-decision read thresholds based on the three hard-decision read thresholds; Step S3: constructing an LLR mapping table, and obtaining new LLR information of MLC flash based on the LLR mapping table; Step S4: symmetrizing an MLC flash channel, and performing density evolution; and Step S5: optimizing the soft-decision read thresholds based on a genetic algorithm to obtain an optimal quantization interval. According to the invention, polar codes can be directly used for MLC flash channels without the arduous work of MLC flash channel modeling, so that the reliability of MLC flash is effectively improved.

Abstract: The present disclosure provides a method, switch and system for updating a route, relating to computer network technologies. The method comprises detecting, by a current device, link states of at least two neighbor devices; when the current device detects that a link state of a first neighbor device is unreachable, setting, by the current device, a state of a routing item with an initial state being reachable among all routing items corresponding to the first neighbor device in an internal routing table to be unreachable; and detecting, by the current device, whether states of all routing items corresponding to each target subnet in the internal routing table are all unreachable, respectively; if yes, sending a route updating message to another neighbor device except the first neighbor device.

Abstract: Taught herein is an integragted narrow bandpass filter array and a method of its fabrication. The filter array is a Fabry-Perot type of filter array, wherein the pass band changes with the thickness of the spacer layer. The integrated filter array comprises a substrate, a lower mirror stack, a spacer array, and an upper mirror stack. The spacer array is an array of varied thicknesses formed using a combinatorial deposition technique. The spacer array is used to control the pass band of each mini-size narrow bandpass filter and realizes the integration of narrow bandpass filters with different pass bands on a single substrate. The merit of this technique lies in its fabrication efficiency and finished product rate which are much higher than for conventional methods. The filter array is completely matched with detector arrays and functional in most of the important optical ranges.

Abstract: Taught herein is an integragted narrow bandpass filter array and a method of its fabrication. The filter array is a Fabry-Perot type of filter array, wherein the pass band changes with the thickness of the spacer layer. The integrated filter array comprises a substrate, a lower mirror stack, a spacer array, and an upper mirror stack. The spacer array is an array of varied thicknesses formed using a combinatorial deposition technique. The spacer array is used to control the pass band of each mini-size narrow bandpass filter and realizes the integration of narrow bandpass filters with different pass bands on a single substrate. The merit of this technique lies in its fabrication efficiency and finished product rate which are much higher than for conventional methods. The filter array is completely matched with detector arrays and functional in most of the important optical ranges.