98 results about "Bandwidth distribution" patented technology

The present application relates to the technical field of large-scale data processing, and particularly relates to a parallel computing method and system suitable for large-scale data processing. The system comprises a task management unit, a distributed load balancing module, an elastic expansion architecture, an intelligent communication optimization module, and a resource monitoring unit, wherein the task management unit divides large-scale data into a plurality of sub-tasks by means of a task decomposer, and distributes the sub-tasks to computing nodes by means of a task scheduler and a priority distributor; the distributed load balancing module achieves global load balancing by means of a load sensing unit, a dynamic adjustment unit and a balance optimization unit; the elastic expansion architecture dynamically adjusts system resources by means of a node manager, a resource pool controller and an expansion decision-making device; the intelligent communication optimization module optimizes inter-node communication by means of a communication path planning unit, a bandwidth distribution unit and a delay compensation unit; and the resource monitoring unit monitors the system performance in real time by means of a performance collector, a state analyzer and an anomaly detector.

The invention discloses a GPU Box hardware decoupling system based on a Retimer card and a PCIe Switch chip, and belongs to the technical field of computer hardware architecture and high-speed interconnection. According to the system, a Retimer card and a PCIe Switch chip are integrated in an independent GPU Box, and a decoupling link of a CPU server and a GPU acceleration card is constructed; the Retimer card realizes 30-meter long-distance PCIe signal transmission and breaks through physical distance limitation; the PCIe Switch chip pools GPU resources through a dynamic routing and MRIOV technology, supports flexible allocation of computing power by multiple servers, and realizes Peer-to-Peer direct connection communication between GPUs. Aiming at a large model reasoning scene, the system optimizes KV cache bandwidth allocation and video memory and memory cooperative scheduling, so that the 100B parameter model reasoning throughput is greatly improved; and meanwhile, the usability of the system is greatly improved through fault isolation and hot plug design. According to the method, the problems of physical binding of the CPU and the GPU, limited transmission distance, rigid resource allocation and the like in a traditional architecture are solved, and the method is suitable for large-scale AI calculation and distributed GPU cluster deployment.

The invention relates to the technical field of machine learning, and particularly discloses a spatial local XGBoost machine learning model based on multi-scale geographic weighting, and the model comprises a feature decoupling module which is used for constructing a two-channel input structure of geographic features and non-geographic features, and achieves the feature decoupling and fusion through a multi-scale spatial weight matrix; the bandwidth allocation module dynamically determines the bandwidth scale in the XGBoost tree splitting process, and establishes a dynamic weight bound by tree levels; the constraint gain module generates split points with scale marks; a contribution decoupling module extracts split features and associates geographic and non-geographic features to realize salinization prediction and contribution index extraction; and the verification optimization module optimizes model parameters through a multi-scale thermodynamic diagram and spatial autocorrelation analysis. The model can capture geospatial effects of different scales, improves the spatial local prediction precision, and is applied to soil salinization analysis scenes with spatial heterogeneity.

The invention relates to the technical field of image transmission, in particular to a remote image transmission system and method based on a compression algorithm, and the method comprises the steps: 1, image blocking and feature extraction; 2, performing block importance dynamic classification; 3, differential compression and layered packaging are carried out; 4, channel adaptive transmission control: dynamically adjusting a transmission strategy according to a real-time channel bit error rate, and when the bit error rate exceeds a threshold value, pausing non-key data layer transmission and migrating non-key blocks to a key data layer according to high compression strength; calculating a bandwidth allocation factor based on the bit error rate and the non-key block proportion, and dynamically controlling the sending amount of a non-key data layer; and step 5, hierarchical reconstruction of the receiving end. By combining a plurality of technologies such as image partitioning, dynamic classification, differential compression, channel adaptive control and receiving end reconstruction, the bandwidth utilization rate is effectively optimized while the image quality is ensured.

The invention relates to the technical field of data processing, and discloses a gateway service integrated management method and system based on an FTTR demand. The method comprises the following steps: carrying out topology scanning on a test signal of an optical time domain reflectometer to obtain a room-level FTTR topology mapping table, carrying out feature extraction on service traffic of each room according to deep packet inspection to obtain a room function identification tag library, and carrying out attenuation compensation on a corresponding room optical port by an optical power meter to obtain an FTTR link quality evaluation matrix, and bandwidth allocation is carried out through a reinforcement learning algorithm to obtain a room-level QoS strategy configuration set, and building data interaction is carried out based on an MQTT protocol to obtain an FTTR gateway service integrated management scheme. According to the invention, the intelligent level of FTTR network room-level management and the resource allocation efficiency are improved. According to the invention, the real-time performance of optical fiber link quality monitoring and the collaboration of building network management are improved.

The embodiment of the invention provides a server bandwidth allocation method and device, equipment and a storage medium. The method comprises the following steps: acquiring state data of an expansion board card; determining a power supply level signal according to the state data; determining an output enabling state according to the power supply level signal; combining the state data, the power supply level signal and the output enabling state to determine hardware feature information; determining configuration information corresponding to the hardware feature information; determining bandwidth identification information matched with the configuration information based on a preset bandwidth allocation strategy; and based on the bandwidth identification information, bandwidth allocation is carried out on the expansion board card, the bandwidth can be automatically identified and reallocated, the same expansion board card can adapt to various configuration combinations, and the utilization rate and the flexibility of the server are remarkably improved.

The invention discloses a data transmission method and device, a nonvolatile storage medium and electronic equipment. The method comprises the steps that a plurality of fragmentation requests sent by a plurality of clients are acquired, and each fragmentation request at least comprises network layer data and application layer behavior data; determining a network quality score of the client according to the network layer data and the application layer behavior data of the client; according to the network quality score, the plurality of clients are divided into a plurality of grades, and the grades are in positive correlation with the priority of fragment transmission; determining the maximum connection number corresponding to the clients of the plurality of levels, and determining a fragment transmission strategy based on the maximum connection number; and controlling fragmentation transmission of the client according to the fragmentation transmission strategy. According to the method and the device, the technical problems of non-uniform resource allocation and incapability of effectively coping with the dynamic change of the network state caused by the static property and the singleness of a related bandwidth allocation strategy in a video fragment on-demand scene are solved.

The invention discloses a multilayer space dynamic topology adaptive communication network construction method, and belongs to the technical field of communication networks. The method is realized through five steps of constructing a multi-layer fusion architecture of space-based, air-based and ground-based networks, instantiating network slices of power service perception, performing dynamic topology abstraction and resource pre-mapping based on virtual cell division, performing power and bandwidth collaborative allocation based on reinforcement learning, and deploying multi-domain fault diagnosis and predicting a closed loop. According to the invention, power and bandwidth collaborative allocation based on deep reinforcement learning not only can optimize the total capacity of the system, but also can realize dynamic balance of the network load. The system can automatically adjust the power and bandwidth distribution according to the real-time network state in the operation process, and the self-adaptive mechanism enables the network to efficiently cope with the demand changes of different power businesses.

The invention relates to the technical field of industrial automation control and robots, in particular to a heterogeneous high-real-time multi-axis robot cooperative motion control system and method.The system comprises an SoC main control layer, an FPGA driving layer and a hardware interconnection layer, the SoC main control layer adopts a heterogeneous SoC chip, an ARM processor is responsible for track planning and task scheduling, an FPGA logic deployment reconfigurable PID calculation engine is responsible for FPGA logic deployment, and the FPGA driving layer is responsible for FPGA logic deployment. A multi-axis position ring and speed ring algorithm is realized through the time division multiplexing processing unit; the FPGA driving layer comprises a plurality of FPGA coprocessors, and an FOC current loop IP core is instantiated to execute bottom layer vector control related operation; the hardware interconnection layer adopts an EPPI protocol supporting topology awareness, dynamic bandwidth allocation and a ping-pong buffering mechanism, and efficient transmission of a control instruction and feedback data is achieved. Through hierarchical control and hardware collaborative design, the problems that an existing multi-axis control system is large in calculation delay, insufficient in synchronization precision and low in hardware resource utilization rate are effectively solved.

The present disclosure provides a bandwidth allocation method, an electronic device, and a computer readable storage medium. The bandwidth allocation method comprises: determining an overall wide area network (WAN) interface bandwidth of a network according to capability information of a WAN interface of each node currently added to the network, wherein nodes comprise a master node and a slave node, and the WAN interface is an interface where the node is connected to the WAN; and allocating traffic to at least one egress node according to the current overall WAN interface bandwidth of the network and a bandwidth distribution condition of each node.

The application discloses a network card management system and method and electronic equipment, and relates to the technical field of network card management. The system comprises a mainboard, a network card and a preset cable. The mainboard is provided with a mainboard power supply circuit, a logic controller and a connector. The connector comprises a power supply pin connected to the mainboard power supply circuit and at least one bandwidth identification pin. The inside of the mainboard side plug of the preset cable is provided with a target circuit connection structure, and the power supply pin and the bandwidth identification pin are directly electrically connected when the cable is connected. The logic controller automatically determines the network card bandwidth allocation mode by collecting the level signal of the bandwidth identification pin and generates a configuration signal to control the network card work. The technical problem that the network card bandwidth configuration in the related art depends on software setting is solved. The automatic identification and configuration of the network card bandwidth are realized through the hardware connection topology, and the system deployment efficiency and reliability are significantly improved.

The invention provides a bandwidth allocation method and device, electronic equipment and a storage medium, and the method comprises the steps: determining an initial bandwidth allocated to each switch by each processor, and enabling each processor to be any one of a plurality of processors of each computing node; each switch is any switch in a plurality of switches with a target number of switches of a predetermined integer, and the plurality of processors are respectively connected with each switch; when the initial bandwidth is a non-integer multiple of the preset transmission channel bandwidth, rounding the initial bandwidth through a rounding strategy, and determining a target bandwidth allocated to each switch by each processor; and determining the receiving total bandwidth of each switch based on the target bandwidth, and adjusting the target bandwidth based on the receiving total bandwidth so as to control each processor to allocate the adjusted target bandwidth to each switch, compared with the related technology, the problem that the bandwidth allocation of the processor is difficult to design due to the bandwidth granularity can be avoided. And meanwhile, the bandwidth allocation efficiency is remarkably improved.

The invention discloses an audio stream transmission method and processing of a self-adaptive network, and the transmission method comprises the following steps: calculating a network state value P representing a network state based on collected network environment data, determining the audio type of audio stream data to be transmitted, and transmitting the audio stream data to be transmitted; and determining transmission demand data of each piece of audio stream data based on the audio type, obtaining a comprehensive weight value A meeting the transmission requirement based on the transmission demand data, and allocating a corresponding transmission bandwidth to each piece of audio stream data according to the network state value P and the comprehensive weight value A. According to the invention, the priority and bandwidth allocation of the audio stream can be dynamically adjusted according to real-time network feedback, the adaptability of the current network environment is considered, the system can more flexibly allocate bandwidth resources, and the audio stream with high priority can be preferentially transmitted, so that the quality and reliability of audio stream transmission are improved.

Disclosed in the present application are a bandwidth construction method, a bandwidth prediction method, and a related apparatus. The bandwidth construction method comprises: first, acquiring first bandwidth data and task scheduling information, wherein the first bandwidth data comprises n bandwidth sampling values of a first memory access component of a vehicle in a first driving scenario, m operation tasks associated with the first driving scenario are run on the first memory access component, and the task scheduling information is used for indicating the runtime sequence of the m operation tasks; and then, on the basis of the first bandwidth data and the task scheduling information, obtaining second bandwidth data, wherein the second bandwidth data is used for indicating time-domain distribution information of the bandwidth of each operation task among the m operation tasks in the first driving scenario. In this way, the construction of fine-grained, operation-task-level bandwidth distribution on a memory access component is realized, such that the bandwidth of the memory access component can be predicted more accurately, thereby not only facilitating the utilization rate of bandwidth resources, but also improving the performance of an autonomous driving system.

The application relates to the technical field of Internet of Vehicles, and discloses a car-machine multi-screen cooperative display and interaction method, which comprises the following steps: maintaining a global session and a member view through a centralized session arranger, realizing session access and dynamic member management of multi-screen devices, and determining multi-screen related parameters; constructing a full-order execution rule based on a logical clock, the multi-screen related parameters and an operation identifier, and assigning a global serial number to each multi-screen operation; obtaining a focus of an operation request, and determining feedback parameters through a soft lock and a priority preemption strategy; periodically hashing and checking based on a shadow state and a domain Merkle tree, positioning state differences according to the feedback parameters, and performing incremental repair through an incremental patch; and adopting WebRTC layered coding and a weighted bandwidth allocation strategy, preferentially guaranteeing control signaling and a main driving screen experience, and dynamically adapting in a bandwidth-limited scene. The application can solve the problems of multi-screen interaction inconsistency and insufficient real-time performance, thereby improving the multi-screen cooperative experience in a vehicle.

The invention provides a cloud gateway message processing method and device and a medium, and relates to the technical field of networks. The method comprises the following steps: establishing a plurality of output sub-interfaces with different service qualities according to value-added services of users, wherein each output sub-interface obtains bandwidth allocation of cloud gateway equipment according to the service quality of the output sub-interface; setting a plurality of message processing actions and establishing a plurality of service routing instances according to the value-added service of the user, wherein each service routing instance is associated with a plurality of message processing actions and an output sub-interface; establishing a service matching instance according to the value-added service use rule of the user, wherein the service matching instance comprises a matching rule of using a service routing instance by different terminals of each user; and using the service matching instance to match the service routing instance, and using the matched service routing instance to process the user terminal message. According to the application, more customized value-added services can be provided for multiple users and multiple terminals through one cloud gateway device.

The present application provides a bandwidth allocation method and device, an electronic device, and a storage medium, which determine the initial bandwidth allocated by each processor to each switch, wherein each processor is any one of the multiple processors of each computing node, each switch is any one of the multiple switches of a predetermined integer target number of switches, and multiple processors are connected to each switch respectively; when the initial bandwidth is a non-integer multiple of the preset transmission channel bandwidth, the initial bandwidth is rounded by a rounding strategy to determine the target bandwidth allocated by each processor to each switch; based on the target bandwidth, the total receiving bandwidth of each switch is determined, and based on the total receiving bandwidth, the target bandwidth is adjusted to control each processor to allocate the adjusted target bandwidth to each switch. Compared with related technologies, the present application can avoid the problem of difficult design of processor bandwidth allocation due to bandwidth granularity. At the same time, it significantly improves the efficiency of bandwidth allocation.

The present application relates to the field of communications, provides a communication method and a communication apparatus, facilitating the reduction of the signaling overhead of a trigger frame. The method comprises: receiving a trigger frame, wherein the trigger frame comprises first information and second information; the first information is used for indicating a discrete bandwidth distribution mode in a first frequency domain range when the first frequency domain range comprises a discrete RU; the second information is used for indicating a discrete bandwidth distribution mode in a second frequency domain range when the second frequency domain range comprises a discrete RU; and an indication granularity of the discrete bandwidth distribution mode in the first frequency domain range is different from an indication granularity of the discrete bandwidth distribution mode in the second frequency domain range, thereby reducing the indication overhead.

This utility model discloses a communication device with dynamic bandwidth physical isolation for Ethernet, including a main control circuit and a communication circuit. The main control circuit and the communication circuit are connected to realize external network communication. The communication circuit is connected to an isolation circuit, which includes a photoelectric conversion chip M1. Its optical signal input pin is connected to an external fiber optic network to receive optical signals, and its electrical signal output pin is connected to the communication circuit to output electrical signals. This device constructs a physical isolation channel through the photoelectric conversion chip, and uses optical signal transmission to achieve physical layer isolation of Ethernet communication, effectively blocking network attacks and data leakage paths. The dynamic bandwidth characteristic can adaptively adjust bandwidth allocation according to transmission needs, improving network resource utilization. It is suitable for scenarios with high requirements for network security and bandwidth stability, such as finance and government, providing a physical isolation solution for Ethernet communication that combines security and flexibility.

The invention relates to a multi-source disturbance self-calibration method and system for a rotary filtering type monochromator, and belongs to the technical field of optical instrument filtering calibration. The method comprises the following steps: converting a multi-source disturbance signal response characteristic into a disturbance energy spectrum basis vector, and extracting spectrum bandwidth distribution to obtain a spectrum self-disturbance calibration parameter; luminous flux preposition constraint is carried out, the convergence state of a real-time spectrum observation value under the disturbance change condition is detected, confidence coefficient calibration is carried out on the disturbance contribution weight on which the self-calibration operation depends, and a filtering disturbance compensation control strategy is generated; constructing a filtering performance evolution model, identifying spectral wavelength shift and resolution data after disturbance self-calibration, verifying the filtering degradation period of the monochromator, and outputting a filtering calibration feedback report; and setting a critical interference stable threshold, establishing a filtering virtual regulation and control mechanism, enhancing the compensation calibration weight of a filtering disturbance compensation control strategy, and generating a filtering disturbance self-calibration scheme.

This disclosure provides methods, components, devices and systems for techniques for tone mapping for 60 MHz distribution bandwidth. Some aspects more specifically relate to techniques for tone mapping for the 60 MHz distribution bandwidth that support distributed transmissions. In some examples, the wireless device may transmit obtain data for transmission in a physical layer convergence protocol (PLCP) protocol data unit (PPDU) over a PPDU bandwidth that exceeds 80 megahertz (MHz) and comprises an 80 MHz distributed transmission subblock, and the 80 MHz distributed transmission subblock may be punctured providing a 60 MHz distribution bandwidth. Tones of the 60 MHz distribution bandwidth may be shifted by tone shift values to align with a boundary of valid tones defined in the punctured 80 MHz distributed transmission subblock of the PPDU bandwidth.

The invention discloses a bandwidth allocation and loss self-compensation cooperative control method, system and device and a medium, and the method comprises the steps: S1, obtaining electric power communication service data, optical fiber line state data and external environment data through multi-source data collection; s2, constructing and training an AI model according to the collected multi-source data, and constructing and training a multi-source data fusion model, a reinforcement learning bandwidth allocation model and a loss trend pre-judgment model; s3, outputting results based on each model constructed and trained by the AI model, synchronously generating a bandwidth allocation strategy and loss compensation parameters, and executing the bandwidth allocation strategy and the loss compensation parameters; and S4, feeding back and inputting the communication state data subjected to bandwidth-loss cooperative regulation and control into each model in construction and training, and updating each model parameter to optimize subsequent regulation and control. According to the invention, dynamic coordination of bandwidth and loss compensation is realized, the hysteresis of discrete regulation and control is solved, the service demand change and the line loss fluctuation are synchronously responded, and the communication link stability and the resource utilization rate are improved.

The invention relates to the technical field of big data transmission interaction, and particularly discloses a multi-channel DMA implementation method based on an FPGA, unnecessary data handling and CPU intervention are eliminated by integrating a hardware acceleration unit on a data transmission path, an intelligent scheduler ensures low-delay response of a high-priority channel, and the data transmission efficiency is improved. Meanwhile, fair bandwidth allocation of all channels is guaranteed through a weighted fair queue algorithm, the total throughput is maximized, multiple times of interruption completion are combined into one time, the interruption response and processing burden of a CPU is greatly reduced, software interaction is simplified, CPU resources are released to process other important tasks, a single high-performance transmission engine is adopted for serving multiple channels, and the system performance is improved. Compared with a scheme of instantiating an independent engine for each channel, precious FPGA logic resources are saved, high-performance and low-delay data transmission is realized, data processing delay is greatly reduced, and flexibility, expandability and maintainability of the system are enhanced.

The invention relates to the technical field of engineering data processing, in particular to a priority-based engineering supervision data processing method, which comprises the following steps of: receiving a data packet sent by a multi-source supervision terminal, and analyzing a value proof voucher embedded in the packet; the value proof voucher is generated by the terminal based on a data service attribute index; according to the real-time bandwidth capacity and the value certification voucher of each data packet, an improved bandwidth auction algorithm is operated, a bandwidth allocation voucher is output, and the improved bandwidth auction algorithm introduces a two-dimensional bidding mechanism and credit degradation punishment for the resource mismatching problem in the project supervision data; and issuing the bandwidth allocation voucher to a corresponding terminal. According to the invention, the adaptive capacity of the system for resisting resource abuse and scheduling congestion is improved. An end-to-end bandwidth allocation closed-loop control and fusing retransmission mechanism is realized, and the transmission link stability and the scheduling efficiency are remarkably improved.

The invention relates to a high-voltage switch fault diagnosis system based on non-intrusive sound wave monitoring. The system comprises: a separation module for performing blind source separation on a flowing acoustic wave signal to obtain a plurality of acoustic wave source signals and corresponding energy ratios; the calculation module is used for calculating the kurtosis of the amplitude sequence of each sound wave source signal and the standard deviation of a first-order difference value, and determining a fundamental frequency component and a plurality of harmonic frequency components based on a spectral analysis result; the first determining module is used for determining an SF6 gas acoustic wave signal according to the kurtosis of the amplitude sequence of each acoustic wave source signal, the standard deviation of the first-order difference value, the spectrum distribution, the energy ratio, the fundamental frequency component and the multiple harmonic frequency components; the second determining module is used for determining a leakage fault coefficient according to the bandwidth distribution of the plurality of harmonic frequency components and the distribution of the high-frequency components; and the output module is used for outputting early warning information according to a comparison result of the leakage fault coefficient and a fault threshold value. According to the invention, accurate identification of the SF6 gas sound wave signal can be realized.

The invention provides a telephone traffic control method and device based on an IMS (IP Multimedia Subsystem) and a medium, and relates to the technical field of communication. Historical data are analyzed, a flow reference interval of voice and video telephone traffic is calculated, current voice and video telephone traffic flow values are obtained, and current network load, time delay and bandwidth occupation conditions are combined, so that the traffic control efficiency is improved. The method comprises the following steps of: calculating a corresponding service quality index according to a current traffic value, evaluating a traffic deviation value of voice and video based on the current traffic value and a reference interval, respectively determining correction coefficients of voice and video traffic in combination with the service quality index, and determining a priority adjustment sequence of the bandwidth by comparing the two correction coefficients. By dynamically calculating the flow reference interval and the correction coefficient of the voice and video traffic, the priority adjustment of bandwidth allocation in the communication network is optimized, and the service quality and the user experience are improved.

The application relates to the technical field of data processing, and discloses a gateway service comprehensive management method and system based on FTTR demand. The method comprises the following steps: obtaining a room-level FTTR topology mapping table through topology scanning of an optical time domain reflectometer test signal, extracting features of service traffic of each room based on deep packet detection to obtain a room function identification label library, compensating attenuation of corresponding room optical ports by using an optical power meter to obtain an FTTR link quality evaluation matrix, performing bandwidth allocation by using a reinforcement learning algorithm to obtain a room-level QoS strategy configuration set, and obtaining an FTTR gateway service comprehensive management scheme based on MQTT protocol building data interaction. The application improves the intelligent level and resource allocation efficiency of room-level management of the FTTR network. The application improves the real-time performance of optical fiber link quality monitoring and the collaboration of building network management.

The invention relates to the technical field of data analysis, can be applied to business system platforms of financial science and technology, medical treatment and health and the like, and discloses a dynamic bandwidth adjustment method, device and equipment based on bucket-level resources and a medium. Obtaining high-priority resource buckets and common resource buckets, collecting system resource states of the resource buckets, performing health degree analysis on the high-priority resource buckets and the common resource buckets, and generating health degree scores of the resource buckets; and performing weighted averaging on the health degree score of the high-priority resource bucket and the health degree score of the common resource bucket according to a preset weight to obtain a global health degree, and performing different dynamic bandwidth allocation and dynamic speed-limiting adjustment on the high-priority resource bucket and the common resource bucket according to the global health degree to obtain a target bandwidth adjustment result. According to the invention, the differentiation and adaptive management of the bandwidth can be realized according to the service importance and the load change.

The application discloses a signal frequency parameter estimation method and device, and relates to the technical field of signal processing. The method comprises the following steps: acquiring a first center frequency, a first bandwidth and a signal-to-noise ratio, and determining a center frequency distribution range and a bandwidth distribution range in combination with a frequency interval; performing band-pass filtering processing and autocorrelation processing on an original signal, and then performing Fourier transform and frequency domain correction processing to obtain a target frequency domain signal; determining a second center frequency in the center frequency distribution range based on the target frequency domain signal, and then determining a corresponding bandwidth lower / upper limit distribution range; and determining a bandwidth start / end frequency in the bandwidth lower / upper limit distribution range based on the target frequency domain signal to obtain a second bandwidth. The application reduces the influence of noise interference through filtering and autocorrelation processing, adaptively limits the distribution range to avoid the limitation of Fourier transform length, and realizes accurate positioning of the frequency parameter through a difference method, so that the accuracy of frequency parameter estimation is significantly improved.

Embodiments of the present application relate to the technical field of computer systems, and provide a computer system for bandwidth allocation, a bandwidth allocation method, and a product. When a motherboard of the computer system is powered on, a monitoring module, upon identifying a change in applications of extension units plugged into the motherboard, generates, on the basis of bandwidth configuration information corresponding to the extension units, a bandwidth allocation information table corresponding to a central processing unit; after the motherboard of the computer system is powered on, a baseboard management controller generates an option configuration file on the basis of the bandwidth allocation information table; after a main unit of the computer system is triggered, a basic input / output system is used for configuring any connection port of the central processing unit on the basis of the option configuration file during a boot process, so as to allocate bandwidth resources of the central processing unit to the extension units. The embodiments of the present application aim to automatically allocate bandwidth resources in a computer system.