Patents
Literature
Patsnap Eureka AI that helps you search prior art, draft patents, and assess FTO risks, powered by patent and scientific literature data.

189results about How to "Reduce resource overhead" patented technology

Auditing data distributed storage method based on multi-layer encryption strategy and related product

The invention relates to the technical field of audit data storage, in particular to an audit data distributed storage method based on a multi-layer encryption strategy and a related product, and the method comprises the following steps: dividing audit data into a plurality of grades, and adopting different encryption strategies for the audit data according to different grades to obtain encrypted audit data; dynamically fragmenting the encrypted audit data based on a fragmentation rule to generate a plurality of data fragments; the data fragments are distributed and stored in a plurality of distributed storage nodes; carrying out integrity verification on the stored data through Hash calculation, Merkle tree verification and a verification information storage mechanism; according to the method, a multi-layer encryption strategy is combined, flexible encryption protection aiming at data with different sensitivities is realized, meanwhile, the encryption efficiency of common data is considered, and the fault-tolerant capability and the dynamic monitoring and migration mechanism of distributed storage nodes are improved through the combination of fragmentation and redundancy coding technologies; and stable operation of the system under the condition of node failure or overhigh load is ensured.
Owner:CHENGDU BIG DATA GRP CO LTD

Design method for long-distance high-precision general-inductance integrated detection waveform

ActiveCN119583283AAchieving Simultaneous PerceptionHigh distance speedMulti-frequency code systemsHigh level techniquesTime domainCarrier signal
The invention discloses a long-distance high-precision sensing integrated detection waveform design method, which comprises the following steps of: 1, determining the number Ncomb of transmission combs according to the number requirement of sensing waveforms required to be supported by a system at the same time; step 2, generating an expandable wide-range sensing signal according to the sensed distance demand, the expandable wide-range sensing signal lasting for L symbols; 3, determining a short period Tps, a short period repeated transmission frequency Lp, a long period Tpl and a long period repeated transmission frequency Np according to the speed detection requirement of the moving target and typical communication time slot configuration; step 4, completing frequency domain signal mapping in a pre-detection stage, and completing subcarrier mapping of adjacent L OFDM symbols according to a frequency domain continuous mode; step 5, completing time domain signal mapping in a pre-detection stage, and repeatedly transmitting the frequency domain mapping signal in the step 4 for Lp times according to a configured Tps period; step 6, completing frequency domain signal mapping in the detection stage, and completing subcarrier mapping of L adjacent OFDM symbols according to the selected combing mode; and step 7, completing time domain signal mapping in the detection stage, and repeatedly transmitting the frequency domain mapping signal in the step 5 for Np times according to the configured long period Tpl. According to the invention, the problems of limited sensing distance and ambiguity of the communication waveform for sensing can be solved, and high-precision sensing of distance and speed of far and near targets in a wide range can be met.
Owner:SOUTHEAST UNIV

Perception method, communication device and computer readable storage medium

The invention discloses a sensing method, a communication device and a computer readable storage medium. The method comprises the following steps: measuring a first sensing signal; and sending a first sensing report in response to at least one of the measurement results of the first sensing signal meeting a reporting condition. The scheme provided by the invention is beneficial to optimizing the configuration of the sensing signal.
Owner:BEIJING SPREADTRUM HI TECH COMM TECH CO LTD

Self-adaptive model partitioning method and system applied to distributed training

The invention discloses a self-adaptive model partitioning method and system applied to distributed training, belongs to the technical field of deep learning, and aims to solve the technical problem of how to realize distributed training of a model by combining deep reinforcement learning and a Bayesian optimization algorithm for distributed training. Comprising the following steps that a Q network is constructed based on a deep neural network, state information, actions and a reward function are defined, and the state information comprises feature vectors of a slice model and training time, a resource utilization rate and inter-node communication overhead when distributed training is conducted on the slice model through distributed computing nodes; the action is a fragmentation strategy adopted by the intelligent agent under the current state information, the reward function R is used for evaluating the effect of the current fragmentation strategy, and the Q network takes the state information as input and predicts and outputs the Q value of each action possibly adopted by the intelligent agent; and carrying out multiple times of iterative training on the deep enhanced adaptive model to obtain a final fragmentation strategy and parameters of the Q network.
Owner:SHANDONG INSPUR SCI RES INST CO LTD

Method for accelerating sparse multi-scalar multiplication by using image processor

The invention discloses a method for accelerating sparse multi-scalar multiplication by using an image processor, which belongs to the technical field of parallel computing, and comprises the following steps: managing scalar vectors participating in operation and point vectors in a group by a CPU (Central Processing Unit), and applying for sufficient memory on a GPU (Graphics Processing Unit) to store the contents; the CPU and the GPU work asynchronously, the GPU carries out bucket distribution, bucket summation and bucket merging calculation, and finally a calculation result in a window is obtained. According to the method, firstly, the sparsity of the scalar is fully utilized, the performance is excellent especially when large-scale data is processed, and the performance is remarkably improved; secondly, the design of the multi-GPU parallel computing architecture gives full play to the advantages of GPU hardware resources, and the load balance is finely regulated and controlled by the method, so that the resource overhead of the system is reduced; finally, the adopted parallel processing technical scheme enables MSM operation to be efficiently expanded to a multi-node multi-GPU environment, and the efficient calculation requirement of a large-scale zero-knowledge proof system is met.
Owner:HUAZHONG UNIV OF SCI & TECH

Database consistency comparison and compensation method, device, equipment, medium and product

The invention discloses a database consistency comparison and compensation method. The method comprises the following steps: acquiring a plurality of databases to be verified; sorting each piece of data in each to-be-verified database according to the primary key, and calculating a hash value of each piece of data in each sorted database to obtain a plurality of preprocessed databases; comparing the plurality of preprocessed databases by adopting a comparison algorithm based on the primary key and the hash value to determine a comparison result; the comparison result is a first result or a second result; the first result represents that the plurality of preprocessed databases have no difference; the second result represents that the plurality of preprocessed databases have differences; the second result comprises a difference conclusion and comparison information; and generating a difference file based on the comparison information, and compensating the plurality of preprocessed databases based on the difference file to obtain a plurality of compensated databases. The detection efficiency is improved, and the storage cost and the resource overhead are reduced.
Owner:AGRICULTURAL BANK OF CHINA

Configuration plug-in implementation method based on Extism framework

The invention discloses a plug-in configuration implementation method based on an Extim framework, which is characterized in that configuration information is independent from a traditional external configuration center or a configuration file by using a Wasm plug-in, a complex configuration center system does not need to be additionally deployed, the resource overhead of the system is reduced, hot update of configuration is realized by packaging configuration parameters in the Wasm plug-in, and the configuration efficiency is improved. The application can detect and load the new-version plug-in and a naming and recording mechanism for the plug-in version without restarting in the running process, plug-in version management under the multi-pod environment is achieved, it is ensured that the plug-in of the same version is used in each pod, dynamic adjustment and flexible calling of configuration parameters are achieved through an interface provided by an Extism framework, and therefore the configuration efficiency is improved. The application can read and modify the configuration parameters at any time according to the service requirements, so that the configuration can be quickly customized in a changeable service environment, and the self-adaptive requirements of various service logics in a complex scene are met.
Owner:SHENZHEN AOSI NETWORK TECH CO LTD

Method for obtaining time difference of two channels of Beidou GEO satellite

The invention discloses a method for obtaining time difference of two channels of a Beidou GEO satellite, and belongs to the field of Beidou navigation. The method comprises the following steps: constructing a local clock counter and clock interruption which are irrelevant to a satellite signal by using a time mark carried by an RDSS downlink signal, intra-week second and week counting information analyzed from telegraph text information, and time information such as bit counting and chip counting which are obtained after capturing and tracking the Beidou satellite signal; the method comprises the following steps: sampling a local clock by using a satellite frame to obtain a local clock count, calculating an accurate frame count of a current satellite in each time interruption so as to calculate a satellite transmitting moment of each receiving channel, and selecting transmitting moments of two channels to carry out subtraction so as to obtain a Beidou RDSS satellite dual-channel time difference. According to the method, hardware logic calculation is replaced by a software calculation mode, so that the baseband hardware resource overhead is reduced, and the reliability, convenience and expandability of obtaining the satellite dual-channel time difference can be improved.
Owner:SHENZHEN YUANDONG HUAQIANG NAVIGATION & ORIENTATION CO LTD

Method and system for loading data required by convolutional neural network accelerator operation processing unit

The invention belongs to the related technical field of integrated circuit design, and particularly relates to a method for loading data required by an accelerator operation processing unit of a convolutional neural network, which comprises the following steps of: controlling a cache sequence of weight data and feature map data through a state machine, when controlling the cache of the feature map data, firstly performing input feature map format conversion, the method comprises the following steps: performing sub-segment division on external channel-by-channel input feature map data with the bit width of A bit, storing the external channel-by-channel input feature map data with the bit width of A bit in a grouping dislocation storage mode, and converting the stored feature map data into N paths of parallel data with the bit width of B bit (B = A / N) in a special reading mode, the B * N data flow can complete feature map data loading of N storage units at the same time, the problem that the off-chip data access bit width is not matched with the data bit width needed in the operation processing unit is solved, and the bandwidth utilization rate of the convolutional neural network accelerator is improved.
Owner:HUAZHONG UNIV OF SCI & TECH

Dynamic link regulation and control method for large-scale satellite network simulation platform

The invention belongs to the field of satellite network simulation, and particularly relates to a link dynamic regulation and control method for a large-scale satellite network simulation platform. The method is characterized by comprising the following steps: an eBPF program searches simulation starting time of current equipment from an eBPF Map by taking an equipment index as a key; judging whether simulation is started or not, and if not, searching the simulation starting time of the current equipment again; if the simulation is started, calculating a timestamp according to the system time and the simulation starting time, and searching link state information of the current equipment at the current moment from the eBPF Map by taking the timestamp and the equipment index as keys; and the eBPF program executes link regulation and control logic based on the eBPF according to the link state information. According to the method, the system resource overhead during formal simulation is reduced, large-scale dynamic link state regulation and control with finer time granularity can be supported, and the simulation logic of the satellite node during formal simulation cannot be influenced.
Owner:COMP NETWORK INFORMATION CENT CHINESE ACADEMY OF SCI

Unmanned aerial vehicle-mounted investigation and rapid interference method

The invention relates to an unmanned aerial vehicle-mounted investigation and rapid interference method. The method comprises the steps of searching a frequency band where a radar target is located; after the radar frequency band is successfully locked, the equipment processes the analysis and interference functions of the radar signal in parallel, and judges and executes a corresponding interference strategy according to an analysis result; interference strategies are selected step by step according to stages: a device scans radar signals at an interval of 200MHz to complete a target search stage; when the center frequency of the radar signal is locked, entering a target tracking stage; after locking, entering a target optimization stage; if the center frequency changes, entering a target re-search stage, and trying to re-lock the signal; in each stage, the equipment adopts different interference modes: a full forwarding interference mode is used in a target search stage, a slice interference mode is used in a target tracking stage, and a random interference mode is switched in a target re-search stage. According to the method, the dependence on an external independent reconnaissance system and resource scheduling is remarkably reduced, and the load requirements of miniaturization and light weight of the unmanned aerial vehicle are met.
Owner:CHONGQING UNIV OF POSTS & TELECOMM

Hardware acceleration method and system based on neural network feature extraction

The invention belongs to the related technical field of image processing, and discloses a hardware acceleration method and system based on neural network feature extraction, and the system comprises a coding convolution processor, a shared feature map cache, a feature point detection structure, a feature image pixel point selector and a descriptor generation structure which are operated according to an assembly line mode. The coding convolution processor carries out convolution on the input image and outputs and stores the input image to the shared feature map cache; the feature point detection structure generates feature points through convolution and post-processing; the feature image pixel point selector generates a pixel region according to the feature point information; the descriptor generation structure only reads feature point related pixel region data from the shared feature map and calculates a descriptor. The end-to-end full hardware acceleration design avoids external memory access, the pipeline design improves the operation efficiency, a feature point pixel region can be obtained in advance, and only the pixel region is processed through a feature image pixel selector and a descriptor generation structure, so that the resource overhead can be reduced, and the system energy efficiency can be improved.
Owner:HUAZHONG UNIV OF SCI & TECH

Playing interface processing method and device, program product and electronic equipment

The invention provides a playing interface processing method and device, a program product and electronic equipment, and relates to the technical field of computers. The method comprises the following steps: displaying a playing interface setting interface; the playing interface setting interface is used for setting editable interface elements in the playing interface; in response to a setting instruction in the playing interface setting interface, determining display information of at least one editable interface element; and displaying the playing interface according to the display information of the editable interface element. According to the method and the device, the user can set the playing interface in a self-defined manner, the personalized aesthetic requirements of the user can be met, the audition experience is improved, the workload of a program developer is reduced, and the resource overhead of a server or a client is reduced.
Owner:HANGZHOU NETEASE CLOUD MUSIC TECH CO LTD

RTOS inert processor context switching method

The invention discloses an RTOS inert processor context switching method, and relates to the technical field of electric digital data processing. According to the RTOS inert processor context switching method provided by the invention, task analysis is carried out in a software mode, special hardware does not need to be used for supporting, and the chip size and hardware resource overhead can be reduced. Moreover, the inert context switching logic is added in the context switching function of the real-time operating system to read the content of the second array and realize the dynamic conversion of storage and recovery operations, so that the defect when no hardware support exists can be overcome, and the system performance is improved.
Owner:CHIPEASY TECHNOLOGY (SHANGHAI) LTD

Ultra-high sensitivity signal generating and receiving method based on Ku frequency band wide beam antenna

The invention provides an ultrahigh sensitivity signal generating and receiving method based on a Ku frequency band wide beam antenna, and relates to the technical field of satellite communication. The signal generation method comprises the following steps: encoding data information to be sent by adopting a three-dimensional TPC encoder; carrying out spread spectrum modulation on the coded information by using an ultra-long spread spectrum code; and processing the chip after spread spectrum modulation, outputting a Ku frequency band signal, and sending the Ku frequency band signal to a satellite. The signal receiving method comprises the following steps: receiving a satellite signal, and capturing a spreading code; completing optimal sampling recovery and carrier phase recovery; and carrying out iterative decoding on the signal after carrier phase recovery through a decoder to obtain decoded data, and completing signal receiving. According to the invention, the duration of burst signals can be effectively reduced, low-complexity receiving of weak-power and extremely-low-signal-to-noise-ratio signals is realized, and the method is particularly suitable for a burst satellite communication system which is low in communication rate, extremely low in signal power, large in system frequency deviation and has a one-way communication requirement.
Owner:THE 54TH RESEARCH INSTITUTE OF CHINA ELECTRONICS TECHNOLOGY GROUP CORPORATION

Privacy protection federated aggregation method based on multi-client inner product function proxy re-encryption

The invention provides a privacy protection federated aggregation method based on multi-client inner product function proxy re-encryption, which comprises the following four steps: an initialization stage, a local training stage, a federated aggregation stage and a parameter updating stage, and specifically comprises the following steps: the initialization stage: step 110, a key generation center (KGC) inputs a security parameter lambda and calls a group generation algorithm; according to the multi-client inner product function proxy re-encryption-based privacy protection federated aggregation method provided by the invention, the computing load is mainly in a parameter aggregation stage of the central server, so that the computing power advantage of the central server can be fully utilized, and the resource overhead of the clients is reduced; in a parameter transmission process, an inner product function encryption mode is adopted to encrypt model parameters, and meanwhile, a semi-honest central server is ensured not to contact plaintext information of any parameter through a proxy re-encryption thought, so that the security of related parameters is ensured, and the privacy of local data is further ensured.
Owner:HEBEI XIONGAN YITAI ZHIFANG MEDICAL TECH CO LTD

Fractional time delay rapid compensation method based on piecewise polynomial fitting interpolation

The invention discloses a decimal time delay rapid compensation method based on piecewise polynomial fitting interpolation, and relates to the technical field of small base stations, MIMO macro base stations or enterprise-level WIFI, comprising an interpolation filter with a fixed coefficient and the following steps: S1, after power-on initialization, establishing a cell carrier wave service signal, logic or software to realize I / Q two-path time delay detection, i / Q signals to be synchronized are generated, and decimal time delay mu is obtained through calculation; s2, selecting a branch needing time delay compensation, and configuring an input parameter mu of a corresponding interpolation filter; s3, the delay operation unit D0 calculates and outputs d = x3-x2-x1 + x0; the delay operation unit D1 calculates and outputs e =-x3 + 3x2-x1-x0; the delay operation unit D2 outputs x2; and S4, calculating and outputting y = (d / 2 * mu + e / 2) * mu + x2, and completing the synchronization of the I / Q two-path time delays, and solving the problem of I / Q asynchronization caused by a relatively fixed decimal time delay scene existing in the I / Q two-path signals on the receiving link.
Owner:HANGZHOU ZHONGKE YIXIN MICROELECTRONICS TECHNOLOGY CO LTD

RDMA rapid congestion control method and system

An RDMA rapid congestion control method and system relates to the technical field of networks, and comprises the following steps: when a network node device receives a target explicit congestion notification message ECN, generating a target congestion notification message CNP based on a source IP, a destination IP and a destination queue number QP or a source QP number in the target ECN message, and sending the target congestion notification message CNP to a source server; when a source end server determines that a target CNP message is a host end CNP message based on a notification type field in the target CNP message, if a target source end QP number corresponding to a source IP, a target IP and a host end QP number in the target CNP message is inquired from a preset source and host end QP mapping table, the host end QP number in the host end CNP message is replaced with the target source end QP number, a new target CNP message is formed, and the target CNP message is sent to the source end server. And executing an original congestion control speed regulation algorithm based on the new target CNP message. According to the invention, the efficiency of congestion notification can be improved, the resource overhead is reduced, and the universality is improved.
Owner:FIBERHOME TELECOMMUNICATION TECHNOLOGIES CO LTD

Digital network switch caching method, device and equipment for scientific data center

The present application provides a method, device and equipment for caching a digital networking switch for a scientific data center. The method is applied to at least one switch in a cascaded cache structure topology network, including: receiving multiple response data from a digital object warehouse; calculating the cache priority of multiple response data according to a multi-dimensional cache function to obtain the cache priority of each response data, and the multi-dimensional cache function is determined according to the cache scenario; determining the response data to be cached from the multiple response data according to the cache priority of each response data; according to the cache replacement algorithm, using the response data to be cached to replace the cache response data in the switch, so as to cache the response data to be cached to the switch, and the cache replacement algorithm is determined according to the hit rate of the lower-level device. In this way, the method reduces the resource overhead of the switch in the digital networking and realizes load balancing, thereby improving the data transmission efficiency.
Owner:PEKING UNIV

Data communication management method, electronic equipment and storage medium

The invention provides a data communication management method, electronic equipment and a storage medium, and the communication management method comprises the steps: generating a first data operation task in response to a received first data processing operation, and searching a first data operation instruction corresponding to the first data operation task in a preset first action model set table; and / or, in response to the received second data processing operation, generating a second data operation task, and searching a second data operation instruction corresponding to the second data operation task in a preset second action model set table; and in response to the found first data operation instruction and / or the second data operation instruction, performing data communication based on the first data operation instruction and / or the second data operation instruction. By means of the method, real-time balance between real-time data reading and operation instruction issuing is achieved, resource overhead generated by task scheduling is reduced, and the real-time performance and efficiency of data communication are improved.
Owner:HUIZHOU HUINENG JINGDIAN TECH CO LTD

Hardware accelerator for robot grabbing

The invention provides a hardware accelerator for robot grabbing. The hardware accelerator comprises a top layer controller, n computing array units, a buffer unit and a router unit. The top layer controller is used for managing sequential logic of a top layer; the buffer unit is controlled through reading and writing of enabling and addressing signals, each computing array unit is composed of m high-bit multiplication unit groups, the m high-bit multiplication unit groups of the same computing array unit share activation, and different computing array units share weights; and the calculation array unit is responsible for processing tile-based convolution or deconvolution of input feature mapping, and outputting as the input of the next layer through the output buffer. The grabbing detection network processed and quantized by the hardware accelerator has lower processing delay, lower power consumption and area, and higher peak energy efficiency and peak area efficiency; more quantization schemes with different bit widths are supported; and the precision is not lost while the resource overhead and the calculation and storage complexity are reduced.
Owner:SUN YAT SEN UNIVERSITY SHENZHEN +1

Code block segmentation method and device, network equipment and medium

The invention provides a code block segmentation method and device, network equipment and a medium, and the method is applied to a sending end, and comprises the steps: carrying out the semantic segmentation of a to-be-sent transmission block, and obtaining a plurality of feature sequences; wherein in the plurality of feature sequences, the lengths of at least two feature sequences are different; determining the coding rate of each feature sequence according to the importance levels of the plurality of feature sequences; according to the coding rates corresponding to the plurality of feature sequences respectively, coding each feature sequence to obtain coded data of the transmission block; and sending the coded data to a receiving end. Differential transmission is carried out according to the importance degree of the semantic features, the problem that semantic feature differential coding is difficult to support effectively due to the fact that code rates and code lengths of all code blocks in the same transmission code are the same in the prior art is solved, the transmission efficiency of a future semantic communication system is improved, and resource overhead is reduced.
Owner:ZGC INSTITUTE OF UBIQUITOUS-X INNOVATION & APPLICATIONS

Ebpf-based rdma data injection attack detection method and system

The application discloses an eBPF-based RDMA data injection attack detection method and system, the method comprises the following steps: a sending terminal monitors the RDMA use behavior of a user state program through an eBPF-based monitoring program, and writes the monitored RDMA operation data into a data buffer; a super process reads the RDMA operation data in the data buffer, calculates a hash value of the RDMA operation data, and sends the hash value as additional data to a receiving terminal; the super process on the receiving terminal calculates a hash value of the received RDMA operation data, and compares the calculated hash value with the hash value carried in the additional data to realize data verification on the RDMA operation data. The application aims to realize accurate detection of RDMA data injection attacks under the premise of compatibility with general RDMA network cards, improve the real-time performance of software protection, and reduce resource consumption.
Owner:NAT UNIV OF DEFENSE TECH

Method and device for inserting pseudo sequence in data modulation

A method and apparatus for inserting a pseudo sequence during data modulation in a communication system according to a channel state and a link state, etc. are disclosed. In an embodiment, a method for data modulation by a wireless communication node of a wireless communication system includes: inserting at least one first pseudo sequence into a first data sequence in a first data block; and inserting at least one second pseudo sequence into a second data sequence in a second data block; wherein a first length of the first pseudo sequence is different from a second length of the second pseudo sequence, and wherein the second pseudo sequence includes the first pseudo sequence or is included in the first pseudo sequence.
Owner:ZTE CORP

Signal sending method and apparatus

Embodiments of this application provide a signal sending method and an apparatus. The method includes: A first device sends a first signal to a second device on N orthogonal frequency division multiplexing OFDM symbols, where the N OFDM symbols include N1 first symbols, N is an integer greater than or equal to 1, and N1 < N. The first device sends a second signal to a third device on at least one of the N1 first symbols. According to the signal sending method and the apparatus provided in this application, a technology similar to radio frequency identification RFID and wake-up radio WUR can be introduced to a communication system, and network resource overheads can be reduced.
Owner:HUAWEI TECH CO LTD

Composite operational circuit in integrated circuit

The invention provides a composite operation circuit in an integrated circuit. The composite operation circuit comprises a first register set, a second register set, a third register, a multiplication / binary operation module and an addition module. The first register group is used for caching a plurality of filter coefficients or a plurality of convolution coefficients. And the second register group caches a plurality of to-be-filtered data which are input in sequence. And the third register receives and caches the input binary data to be subjected to binary convolution. The multiplication / binary operation module and the addition module perform multiplication and addition operation based on the data to be filtered and the filtering coefficient when performing filtering operation, and perform binary operation and addition operation based on the convolution coefficient and the binary data to be subjected to binary convolution when performing binary convolution operation. In this way, the filtering operation and the binary convolution operation are carried out through the time division multiplexing first register and the addition module, and the resource overhead of hardware is reduced.
Owner:MEMSIC SEMICON WUXI

Heterogeneous GPU cluster scheduling method based on two deadline perception

The invention relates to a scheduling method which schedules deep learning jobs through a primitive-dual framework and a dynamic programming algorithm in a heterogeneous GPU cluster and can meet different deadline requirements of the jobs. According to the method, on-line analysis and historical data are combined in a heterogeneous GPU cluster environment, so that the overhead of obtaining the job throughput is reduced; according to the method, job scheduling and job placement are decoupled, the complexity of job scheduling is reduced, meanwhile, a simple and effective job placement scheme is designed, and the degree of resource fragmentation is effectively reduced; a scheduling target is modeled in a scheduling module, reward value functions are respectively designed for strict deadline operation and soft deadline operation, an optimization problem is constructed by taking maximization of an overall reward value as the scheduling target, and strict deadline requirements and soft deadline requirements of the operation are considered at the same time; and reconstructing the target optimization function into an integer linear programming problem, and solving by using a primitive-dual framework and a dynamic programming algorithm to ensure the theoretical effectiveness of the solution.
Owner:CHENGDU UNIV OF INFORMATION TECH

ETC system resource dynamic allocation method based on virtualization technology

The invention provides an ETC system resource dynamic allocation method based on a virtualization technology, and the method comprises the following steps: constructing a plurality of virtualization operation spaces based on the virtualization technology, and carrying out the multi-stage virtualization resource configuration based on a dynamic resource allocation technology; resources are configured for the backup virtualization operation space, and the backup virtualization operation space is updated based on a time sequence prediction method; and the server recycles idle system resources to establish a test virtualization space. According to the method, more accurate resource allocation of the application components in the system is realized in a multi-stage resource allocation mode, the fault deactivation time can be remarkably shortened in combination with the redundancy design of the virtual backup, the resource consumption level of the virtual backup is reduced according to the updating strategy, and the recycled idle resources can be fully utilized. Furthermore, a test virtualization space is additionally arranged to ensure that daily work and test work are carried out separately, the overall work is not influenced, and the test work can be completed by using idle resources.
Owner:GUIZHOU QIANTONGZHILIAN TECH CO LTD

Motion control system and method based on multiple bionic visual neurons

The embodiment of the invention provides a motion control system and method based on multiple bionic visual neurons, and the system comprises an image collection module which is used for collecting an environment video in real time, and obtaining image information; the visual detection module is used for calculating and outputting a visual motion trend through LGMD and DSN according to the acquired image information, and the visual motion trend comprises early warning information and a translation direction of an obstacle object; and the motion execution module comprises a three-state hysteretic neuron TSHN and is used for carrying out scheduling through the set neuron output priority and the visual motion trend, outputting corresponding motion instructions in sequence and controlling the robot to move. By adopting the embodiment of the invention, various visual motion trends can be quickly extracted and fused from a dynamic environment, and corresponding behavior responses are automatically triggered according to different visual motion scenes, so that visual motion sensing neurons in different modes are correctly scheduled and allocated, and an obstacle avoidance task of a robot is completed.
Owner:GUANGZHOU UNIVERSITY

Method and device for connecting base station and core network device

The present application provides a connection method and device between a base station and a core network device. The method includes: the base station sends a first S1 link establishment request to a first core network device and a second core network device respectively, receives a first S1 link establishment response message sent by the first core network device, and uses the first core network device to process services; receives a first S1 link establishment failure message sent by the second core network device, and periodically sends a second S1 link establishment request to the second core network device until a second S1 link establishment response message sent by the second core network device is received, and uses the second core network device to process services. The base station does not need to maintain the S1 link between the backup device, which reduces the resource overhead of the base station.
Owner:CHENGDU TD TECH LTD