21 results about "Cross layer" patented technology

A method of dynamically routing packets to a destination node performed by a computing device is disclosed. The method includes: (1) detecting a status of a plurality of links to the destination node across a plurality of communications modalities; (2) determining a set of links to use for routing packets to the destination node based on the detected statuses; and (3) sending packets to the destination node via the determined set of links. A related computer program product, apparatus, and system are also disclosed.

The invention relates to a vanadium dioxide-based multiband dynamic adjustable terahertz metasurface, which comprises a cross-like layer, a first dielectric layer, a first thin film layer, a double-opening metal layer, a second dielectric layer and a second thin film layer which are sequentially arranged from top to bottom, the cross-like layer, the first thin film layer and the second thin film layer are made of vanadium dioxide; the double-opening metal layer is annular, and the annular surface of the double-opening metal layer is provided with a first opening part and a second opening part which penetrate through the annular surface. Vanadium dioxide serves as materials of the cross-like layer, the first thin film layer and the second thin film layer, phase change characteristics of vanadium dioxide at different temperatures are utilized, and the first opening part and the second opening part formed in the double-opening metal layer are combined, so that the metasurface can achieve the wave absorbing function and the frequency selection function at the same time. Moreover, the metasurface provided by the invention is simple in structure and easy to process, and the production cost can be effectively reduced.

The invention discloses a software defect positioning method based on feature crossing and a KAN network. Firstly, different types of features are extracted from a defect report and a source code file, then cross correlation of the features is carried out through a cross layer, and a hidden relation between the features is mined; the KAN is further used for learning a parameterized nonlinear activation function, the process of fitting a polynomial function by the feature cross network is converted into the process of fitting a plurality of one-dimensional functions, dimension disasters are overcome, changes of complex functions are accurately captured and adapted, complex high-order feature interaction information in defect positioning is accurately captured, and the defect positioning accuracy is improved. Therefore, the defect positioning performance is improved. And finally, inputting a result output by the KAN network into a full connection layer, calculating a final similar score, and carrying out descending sorting according to the final similar score to obtain a defect positioning result. The method improves the defect positioning precision, has advantages in the aspect of calculation efficiency, and is especially suitable for key business application scenes with strict time requirements.

Provided is a ferroelectric field effect transistor including a source region, a drain region, a channel provided between the source region and the drain region, a ferroelectric layer provided on the channel and including a ferroelectric material including an oxide of a first element, a gate-interposed layer provided on the ferroelectric layer and including a paraelectric material including an oxide of a second element different from the first element, and a gate electrode provided on the gate-interposed layer, wherein the gate-interposed layer includes a first interposed layer adjacent to the ferroelectric layer, and a second interposed layer adjacent to the gate electrode, the first interposed layer includes a mixture of the first element and the second element, and a ratio of the first element in the first interposed layer may be greater than a ratio of the first element in the ferroelectric layer.

Various embodiments include methods and neural network computing devices implementing the methods for performing quantization in neural networks. Various embodiments may include equalizing ranges of weight tensors or output channel weights within a first layer of the neural network by scaling each of the output channel weights of the first layer by a corresponding scaling factor, and scaling each of a second adjacent layer's corresponding input channel weights by applying an inverse of the corresponding scaling factor to the input channel weights. The corresponding scaling factor may be determined using a black-box optimizer on a quantization error metric or based on heuristics, equalization of dynamic ranges, equalization of range extrema (minima or maxima), differential learning using straight through estimator (STE) methods and a local or global loss, or using an error metric for the quantization error and a black-box optimizer that minimizes the error metric with respect to the scaling.

A process layer management and display method supporting multi-chip stacking design comprises the following steps: analyzing a layout file, generating a process layer list and a core particle list of all core particles, and distributing a corresponding shadow layer for each core particle; the shadow layer comprises a layer number and a layer name of the process layer, and a definition for controlling visible, selectable and color display of the core particles or the process layer; searching all core particles containing corresponding process layers from the core particle list and the process layer list according to input layer numbers and layer names, and performing highlight display; when the editing area is switched, the process layer of the target core particle is set to be in an active state in the process layer list; and storing the editing operation data of the process layer in the current active state into an independent layout file. An independent shadow layer is bound for each core particle, and the number and the name of the original process layer are reused, so that the correct editing behavior is ensured; the core particles and all process layers in the core particles are managed and displayed through coarse and fine granularity, so that the management and display problems of cross layers among different processes are solved.

The application discloses a kind of anonymous routing methods of double-layer satellite network for delay-sensitive service, it is applied to VLEO / LEO double-layer satellite network, this network is based on software-defined network SDN architecture implementation, including data plane, control plane and application plane;Data plane includes VLEO layer, VLEO / LEO cross layer and LEO layer;Application plane is deployed with the routing method of the application upper;The method comprises: constructing layered decoupling routing calculation framework, to divide routing decision into including LEO layer inner routing, VLEO / LEO cross layer routing and VLEO layer inner routing multilayer path selection;Using the collaborative computing capability of SDN controller and satellite node, multilayer path selection is reconfigured and weight disturbance in real time, generates optimal routing path.The application can effectively reduce end-to-end delay, balance network load, improve the real-time performance, robustness and security anonymity of double-layer satellite network.

The application discloses a kind of unmanned material transport systems and methods, including intelligent equipment cluster and remote centralized dispatch system, and intelligent equipment cluster is connected with remote centralized dispatch system;Intelligent equipment cluster includes horizontal transport intelligent equipment for the same layer transport of material, intelligent construction elevator for the cross layer transport of material, and stacking robot for the sorting of material;Remote centralized control system includes front-end interface for issuing tasks, supervising material process and equipment state and back-end server for task scheduling, task path planning, task conflict monitoring and management, and the information collected is transmitted to front-end interface for display, and front-end interface is connected with back-end server;Intelligent equipment cluster and back-end server are communicated bidirectionally through real-time communication module.The application can automatically complete material sorting and transportation, can reduce construction site manual operation, improve equipment efficiency, and reduce transportation cost.

Presented are systems and methods for implementing a cross layer network control. A wireless communication node can receive or send information communicated between the wireless communication node and an application.

The invention relates to a small target detection method and system based on adaptive texture perception and frequency domain-cross-layer collaborative optimization, and the method comprises the steps: providing a cross-scale texture guide module, independently capturing the spatial distribution characteristics of heterogeneous features through a multi-scale texture extraction unit, and constructing scene-level texture prior through an additive fusion mechanism; a frequency domain detail enhancement module is provided, features are mapped to a complex frequency spectrum space through real number fast Fourier transform, dynamic reweighting is carried out on medium-high frequency components by using an adaptive weight vector, and target contour information is explicitly enhanced in combination with inverse transform and a residual connection mechanism; a cross-layer feature stabilization module is designed, context clues of different granularities are captured through lightweight multi-scale receptive field extension, and a dynamic channel gating modulation mechanism is introduced to perform active intervention on a feature evolution path. Compared with the prior art, the method has the advantages of characteristic enhancement, stable positioning, balanced detection precision and operation efficiency and the like.

Aspects of this disclosure relate to methods of cross layer rate selection and related controllers and systems. A media access control scheduler can select a rate for transmitting data for a wireless terminal based on a queue stability metric and a predicted transmission rate associated with the wireless terminal. Information identifying the rate can be sent to an application server. The rate can then be used by application server to send data for the wireless terminal to an access point via an access point controller, in which the access point can wirelessly serve the wireless terminal.

To provide a wiring board in which a risk that both of adjacent wiring layers short-circuit is reduced.SOLUTION: A wiring board includes: a first wiring layer; an insulation layer that contains a filler formed onto the first wiring layer; a via hole that penetrates the insulation layer, and exposes an upper surface of the first wiring layer; and a second wiring layer that is electrically connected to the first wiring layer so as to fill the via hole, and is extended to the upper surface of the insulation layer from the via hole. The insulation layer contains: a first insulation layer that coats the first wiring layer; and a second insulation layer that is laminated onto the first insulation layer. The second insulation layer is thinner than the first insulation layer, and an amount of the filler contained in the second insulation layer is smaller than the amount of the filler contained in the first insulation layer. In a cross-sectional view, an inner surface of the via hole is inclined to the upper surface of the first wiring layer. In the cross-sectional view, an inclination angle to the upper surface of the first wiring layer of the inner surface of the via hole positioned in the first insulation layer is larger than the inclination angle to the upper surface of the first wiring layer in the inner surface of the via hole positioned in the second insulation layer.SELECTED DRAWING: Figure 1

The utility model belongs to the technical field of heat exchangers, and provides a cross laminated heat exchanger, which comprises a heat exchanger core body, the heat exchanger core body comprises a first heat exchange component and a second heat exchange component, the first heat exchange component comprises a plurality of groups of first heat exchange assemblies arranged in an array, and the second heat exchange assembly comprises a plurality of groups of second heat exchange assemblies arranged in an array. The second heat exchange component comprises a plurality of second heat exchange assemblies arranged in an array mode. The heat exchanger core is formed by intersecting and stacking the multiple first heat exchange assemblies and the multiple second heat exchange assemblies, the composition structure is simple, the production difficulty and the production cost are reduced, the specification of the heat exchanger can be adjusted by increasing or decreasing the same number of the first heat exchange assemblies and the second heat exchange assemblies, and the production efficiency is improved. And meanwhile, the multiple sets of first heat exchange assemblies are communicated to form the first flow channel, the multiple sets of second heat exchange assemblies are communicated to form the second flow channel, the first flow channel and the second flow channel are independently formed, and the oil-water isolation sealing effect is good.

A kind of high-density server PCB board manufacturing method, comprising the following steps;Step S1, providing n layer circuit board;Step S2, n layer circuit board is formed respectively electric signal layer and ground layer;Step S3, through drilling, copper sinking, electroplating, ground hole and signal via hole that are formed in L1, L2, L3 ……Ln are passed through n layer circuit board;Step S4, between L1 and L2, Ln-1 and Ln layer by laser, electroplating makes L1 and L2, Ln-1 and Ln layer form blind hole;Step S5, between L1 to L3, Ln-2 to Ln layer by laser, electroplating is formed across L2 and Ln-1 cross layer blind hole, and cross layer blind hole is conducted between L1 and L3, Ln-2 and Ln layer.

The present application relates to an information processing method, apparatus, device and storage medium, and relates to the field of Internet technologies. The method includes: obtaining a first pooling vector of a target user; the first pooling vector is a vector generated in the process of processing the user features of the target user through the two-tower structure of an information processing model to obtain a user tower vector of the target user; inputting the first pooling vector into a feature automatic cross layer of the information processing model to obtain a first cross-equivalent vector corresponding to the target user; splicing the first cross-equivalent vector and the user tower vector to obtain a first target vector of the target user; processing the first target vector and a second target vector of target information in a target manner to obtain a vector processing result; and processing the target information in a target processing manner based on the vector processing result. Through the above method, the information interaction between the user side and the information side can be improved, and the processing effect of the target information can be enhanced.

A method of dynamically routing packets to a destination node performed by a computing device is disclosed. The method includes: (1) detecting a status of a plurality of links to the destination node across a plurality of communications modalities; (2) determining a set of links to use for routing packets to the destination node based on the detected statuses; and (3) sending packets to the destination node via the determined set of links. A related computer program product, apparatus, and system are also disclosed.

A method of dynamically routing packets to a destination node performed by a computing device is disclosed. The method includes: (1) detecting a status of a plurality of links to the destination node across a plurality of communications modalities; (2) determining a set of links to use for routing packets to the destination node based on the detected statuses; and (3) sending packets to the destination node via the determined set of links. A related computer program product, apparatus, and system are also disclosed.

The present invention discloses a lightweight brain tumor segmentation network for multimodal magnetic resonance imaging and a segmentation method thereof, belonging to the technical field of brain tumor image segmentation; it includes an encoder and a decoder. The encoder includes a first feature processing layer, a second causal cross layer, a third causal cross layer, and a fourth feature fusion layer; the decoder includes a first output layer, a second feature processing layer, a third feature decoding layer, and a fourth output layer; the encoder adopts a cross-layer connection method to realize the information fusion of feature maps of the same scale between different levels and transmits it to the decoder, and the decoder splices the features flowing out of the encoder to perform cross-processing of shared features. By adopting a cross-layer connection method between the decoder and the encoder, the present invention realizes the information fusion between different levels of feature maps of the same scale; the causal cross-connection module shares parameters to promote the information exchange and fusion between features, better captures the common features between modalities, and thus improves the segmentation accuracy of the brain tumor segmentation network.

The invention discloses a string connection type hollow-core optical fiber and an optical transmission system, and belongs to the technical field of optical fibers. The optical fiber comprises: a hollow core region (101); an outer cladding (301); the multiple negative curvature medium units (201) are distributed around the hollow core area and located on the inner side of the outer cladding layer; a cell wall (601); a cell space (701) commonly defined by the negative curvature medium cells and the cell walls; the entity crossing layers (401) are arranged in the unit spaces and stretch across the unit spaces, and the two ends of the entity crossing layers are connected to the unit walls. Different from a traditional nested structure, the cross-over layer is adopted to replace a nested pipe, the two ends of the cross-over layer are fixedly connected to the unit walls to form bidirectional support, and the machining yield and the structural stability are remarkably improved; the number of air chambers is reduced and the manufacturing process is simplified; the cross-layer configuration of more layers can be realized, the limitation of the number of layers of a traditional nested structure is broken through, the transmission loss is effectively reduced, and the working bandwidth is expanded.

The invention discloses a defect analysis method, a storage medium and a terminal. The defect analysis method comprises the following steps: providing a machine learning model; training and testing the machine learning model, and after the machine learning model passes the test, inputting the characteristic parameters of each to-be-detected integrated circuit layer of the to-be-detected wafer into the machine learning model, and outputting the position of the key defect integrated circuit layer with the key defect and the position of the potential defect integrated circuit layer causing the key defect based on a machine learning model. Based on a machine learning model, the position of a key defect integrated circuit layer which will generate key defects and the position of a potential defect integrated circuit layer which causes the key defects in the manufacturing process of a wafer to be detected can be rapidly predicted, and the potential defects associated with cross layers are identified and predicted through artificial intelligence. The method has the advantages that manual misjudgment and manual time consumption are reduced, the production efficiency is improved, a better rate can be obtained in a shorter time, and reworking in subsequent tape-out caused by cross-layer defects is avoided as far as possible.

The invention discloses a frequency domain selection cross-layer fusion target detection method based on DCT (Discrete Cosine Transform). In order to solve the problem that in an existing YOLO series target detection network, in the multi-time down-sampling process, target detail information is prone to being lost, a frequency domain selection cross-layer fusion module DCTFSCLF based on discrete cosine transform (DCT) is introduced into the YOLO series network, and a frequency domain selection cross-layer fusion module DCTFSCLF based on the DCT is introduced into the YOLO series network. The module performs multi-scale DCT on an input image, extracts medium-high frequency features according to a preset frequency selection mode, and fuses the medium-high frequency features with spatial domain features of different scales in a neck network in a cross-layer mode, thereby enhancing the expression ability of target details and edge information. According to the method, mainstream detection networks such as YOLOv5, YOLOv8 and YOLO11 can be flexibly embedded, the detection precision and the recall rate are effectively improved while the real-time performance is ensured, and the method has good universality and engineering application value.