64results about How to "Guaranteed resource utilization" patented technology

The invention discloses a game theory-based radio resource optimization method in an LTE-A relay system. The method mainly resolves the problem of compromise between system throughput and user fairness when relays are introduced to a cell. The method comprises the steps that SINR values of all users in the cell are calculated, whether the users are connected to a base station or the relays is determined according to the SINR values of the users, the reachable data rate and the minimum required rate of each user are calculated, the SINR dispersity of each micro cell is calculated, and the cell balance index meeting cell throughput maximization and user fairness optimization is obtained through the game theory. According to the game theory-based radio resource optimization method, the SINR dispersity of the users is used as the main parameter of fairness, and fairness of the users in the center of the cell and the users on the edge of the cell are improved obviously; the throughput, the fairness and the game theory are combined, so that system resources are used more effectively, the throughput is maximized under the condition that the fairness is guaranteed, and whole performance of the system is enhanced.

The invention provides a resource pool allocation method and device, which can improve the resource utilization rate of the system while avoiding unnecessary collisions. The resource pool allocation method comprises the following steps: a first user obtains a first resource pool configuration signaling, wherein the first resource pool configuration signaling comprises a configuration signaling ofa first user sending resource pool and a configuration signaling of a second user sending resource pool; and the first user configures the first user sending resource pool according to the configuration signaling of the first user sending resource pool, wherein the first user sending resource pool comprises a first user overlapping part resource pool and a first user non-overlapping part resourcepool; and the first user overlapping part resource pool is the overlapping part of the first user sending resource pool and the second user sending resource pool configured by the configuration signaling of the second user sending resource pool.

The invention belongs to the cloud computing technical field and relates to a multi-QoS constrained cloud computing task scheduling method. According to the method, required QoS descriptions of all tasks submitted by users are given, and QoS service providing ability descriptions of all resources when the resources are distributed are given; and a scheduling algorithm tries to search a QoS optimal task scheduling scheme according to the QoS descriptions, so that the QoS optimal task scheduling scheme can be obtained. With the multi-QoS constrained cloud computing task scheduling method adopted, problems existing in multi-QoS constrained cloud computing task scheduling can be solved. The method can be applied to cloud computing task scheduling.

The method includes steps: calculating backup LSPs of all main LSPs in network; if a LSP wins an advantage of former backup LSP, then the LSP is determined as current backup LSP; otherwise the former backup LSP is as current backup LSP. The invention provides a device for selecting backup LSP. The device includes calculation module, comparison module, and current backup LSP determination module. The feature of the invention is that network holds minimal bandwidth in use for recovering path so as to raise use ratio of network bandwidth.

The invention relates to the technical field of wireless local area network communication, in particular to a commercial WIFI (Wireless Fidelity) user dynamic bandwidth limiting method and system. The commercial WIFI user dynamic bandwidth limiting method comprises the following steps: 1) detecting a bandwidth utilization rate of commercial WIFI equipment; 2) setting a user bandwidth triggering threshold according to the bandwidth utilization rate; and 3) when a bandwidth usage amount of a user exceeds the user bandwidth triggering threshold, setting a bandwidth speed limiting value for a corresponding user. Through adoption of the commercial WIFI user dynamic bandwidth limiting method and system, the bandwidth speed limiting value of the user is adjusted dynamically according to the overall bandwidth utilization rate of the equipment, so that the network usage experience of the user can be ensured as much as possible, and the problem of network congestion can be solved well.

The invention relates to a compatible method of a 64-bit Linux operating system for 32-bit application software. The compatible method comprises the steps of designing a 32-bit sub-system in a nucleus of a 64-bit operating system of a Loongson platform, judging a to-be-run program, selecting and calling a corresponding system, or judging the format of the application program through a pointer length, further selecting a system corresponding to the program, and calling an loader, a linker and a runtime library corresponding to the bit number of the program for execution. The compatible method of the 64-bit Linux operating system for the 32-bit application software has the beneficial effects that as application programs which cannot be deployed in a Loongson 64-bit Linux operating system such as a Linux platform within a short time or application programs which are relatively complicated to transfer from a 32-bit Linux operating system to the 64-bit Linux operating system and even unsuitable for a 64-bit platform can be completely run in the 32-bit sub-compatible sub-system of the 64-bit Linux operating system, resource utilization rate and running efficiency are guaranteed, system compatibility is improved, the number of applicable programs is increased, and popularization of application of the 64-bit Linux operating system is promoted.

The invention relates to a resource distribution method and device oriented to multi-level elastic application of a data center and belongs to the technical field of multi-level cloud application. The resource distribution method comprises the following steps: firstly, obtaining a multi-level application request put forward by a user; secondly, ranking bandwidth demands of various layers in a descending order; next, distributing virtual machine resources to each layer according to the bandwidth demand of each layer in an order from high to low, specifically including the steps of firstly, calculating the feasible vectors FV1 of all distribution schemes according to a tree topology under a current cloud platform to the number of virtual machines needed according to the request of the present layer, and secondly, obtaining the optimal distribution scheme of each chain to the present layer by use of a recursive method; next, calculating the number of virtual resources operating on demand and reserving the virtual resources in the data center; and finally, outputting the optimal distribution scheme to the multi-level application request of the user. Compared with the prior art, the resource distribution scheme of the resource distribution method takes a resource utilization rate and resilience into account, and therefore, the efficiency of the whole system is improved while the resilience is guaranteed.

The invention discloses a method for testing heartbeat which comprises the following processes: an idle task is arranged to be used for sending heartbeat messages; the idle task is called to send the heartbeat messages to a collection center processor unit (CPU). The invention also provides a heartbeat testing system and the CPU to be applied in a symmetrical multi-processing (SMP) system to realize the state detection of the CPU in the SMP system. The initial priority of the idle task adopted by the invention is low priority and the priority of the idle task can be adjusted according to actual requirements, thus ensuring that the heartbeat testing has a relatively high real-time performance under the situation that the utilization rate of internet resources in the SMP system is guaranteed.

The invention discloses a method for establishment of an audio fingerprint database. The method includes: extracting audio fingerprints and key values thereof; setting up a corresponding relation between the audio fingerprints with same key values and the key values; distributing the key values and the corresponding audio fingerprints into servers; setting up a corresponding relation between the key values and the servers to set up the audio fingerprint database. The invention further discloses equipment for establishment of the audio fingerprint database and a method and equipment for retrieval of the audio fingerprints according to the audio fingerprint database. Due to the fact that the audio fingerprints and the key values associated with the audio fingerprints are distributed into the servers in a balanced manner, the audio fingerprints and the key values associated with the audio fingerprints in each inquire request can be precisely provided for the corresponding server to complete calculation, and accordingly resource utilization rate is guaranteed to the maximum extent; by means of statistics of access frequency of each key valve of each server during system operation, load balance is optimized, and system throughput capacity is improved.

The invention discloses a virtual wireless network resource allocation method based on channel pricing. The virtual wireless network resource allocation method is a matching/Stackelberg hierarchical game based on channel pricing. The user satisfaction based on the stream bandwidth, the system bandwidth and the slice power are respectively taken as reward functions, a joint optimization model for the three-layer architecture virtual network is established, and solving is performed by adopting a matching/Stackelberg hierarchical game. A game lower layer defines an MVNOm-InPn pair and a one-to-one matching game between the MVNOm-InPn pair and UEs, a slice power price based on channel average information is defined, upper-layer game information is coupled, the number of iterations is reduced,and a rejection-receiving algorithm with low calculation complexity is given, so that a user selects the optimum under the condition that slice power is equal. The game high layer is an InPs and Stackelberg game, and a power pricing and allocation strategy based on channel local information is given, so that the system utility and the resource utilization rate are optimal. And double-layer circulation and stabilization conditions are given.

The invention discloses a batch computation method and device for geographic position distances. The batch computation method for geographic position distances comprises the following steps of for a given first position point set A and a given second position point set B, respectively obtaining geohash codes of position points in the set A and the set B; aiming at any obtained geohash code i, determining a subset Ai formed by position points, corresponding to the code, in the set A and determining a subset Bi formed by position points, corresponding to the code, in the set B; and aiming at the obtained geohash code i, computing a distance of a position point pair corresponding to each element in Ai*Bi. According to the method and device disclosed by the invention, the unnecessary position point pair distance computation can be decreased and the processing speed can be enhanced; and the method and device can be effectively suitable for distributed processing environment.

The invention discloses a routing and spectrum allocation method and device for a data center optical network. The method comprises the following steps of: arranging each group of service traffic intoa service traffic matrix T; inputting the service flow matrix T into the trained n * (n-1) routing allocation neural network between the data centers to obtain the routing category of the service flow between the n * (n-1) data centers, with n being an integer; inputting the service traffic matrix T into the trained n * (n-1) spectrum allocation neural networks between the data centers to obtainthe spectrum category of the service traffic between the n * (n1) pairs of the data centers, with n being an integer; and integrating the routing category and the spectrum category of the service traffic between the data centers by n * (n-1). High-speed allocation of resources is realized while a certain resource utilization rate is ensured, the deep learning technology has outstanding performancein solving complex problems, the complex relationship between input and output can be autonomously learned through the strong learning ability of the neural network, the forward calculation complexity is low, the consumed time is short, and the efficiency is very high.

The invention relates to a fairness job scheduling method based on a reservation mechanism, and belongs to the technical field of computers. The invention provides a fairness job scheduling method based on a reservation mechanism in order to solve the low computing resource utilization rate, prolonged job waiting time and response time and lack of consideration of fairness due to job starvation and resource fragmentation of a traditional job scheduling method. According to the method, a scheduling model is established based on the characteristics of independent task applications, and the fairness is ensured by using a dynamic adjustable priority and a preemption mechanism, that is, any job cannot be executed after a job with the priority higher than that of the job. The reservation mechanism is used to ensure the maximum resource utilization rate, so that idle resources can be used in advance by jobs with low priorities and low resource requirements, and resources are reserved for jobswith high priorities and high resource requirements.

The embodiment of the invention provides a resource allocation indication method, a resource allocation acquisition method, a base station and a user terminal. The resource allocation indication method is applied to the base station, and comprises the following steps: determining frequency domain resource allocation granularity; wherein the frequency domain resource allocation granularity is greater than one resource block RB; dividing the BWP according to the size of the bandwidth part BWP and the frequency domain resource allocation granularity to obtain N units; sending scheduling information to the user terminal; wherein the scheduling information comprises a frequency domain resource allocation indication domain, and the frequency domain resource allocation indication domain is used for indicating a unit allocated to the user terminal in the N units. According to the embodiment of the invention, the resource utilization rate can be well ensured.

The invention discloses a side-cloud collaborative electric power Internet of Things intelligent inspection method and system. The method comprises the following steps: collecting information collected by an unmanned aerial vehicle; convolution operation being carried out on the collected information to extract feature information, and a 5G core network being utilized to transmit the feature information and the total number of the serial numbers; performing full-connection operation by utilizing the feature extraction information and the serial number, and judging abnormal features; performingcategory identification on the information of the abnormal features; and outputting the abnormal information and the information category to complete inspection. According to the invention, the hardware resource difference of each level is fully utilized, a task allocation scheme with unified efficiency and capability is realized, the computing capability of the whole system is accelerated, the hardware resource utilization rate and the inspection work efficiency of all equipment are ensured, the problem of large calculation amount of a unilateral server is solved, and thus better economic benefits are brought.

The invention discloses a fuel conveying device for a hot-water boiler. The fuel conveying device comprises a fixing block, a conveying box and a crushing box, wherein the conveying box is arranged above the fixing block, and the crushing box is arranged above the conveying box; a rolling shaft is arranged in the crushing box, and a blade is arranged at the outer side of the rolling shaft; a baffle is arranged below the rolling shaft; a spacer is arranged below the baffle; a return pipe is arranged at the outer side of the crushing box, and the lowest end of the baffle is communicated with thereturn pipe; and a conveying motor is arranged on the return pipe, and the outlet of the return pipe is connected with the crushing box. The invention discloses a fuel conveying device for a hot-water boiler. The crushed fuel can be filtered through formation for filter holes, so that resource utilization is guaranteed; the spacer is made from a rubber material and gives buffering for the fuel, so that the conveying box is effectively protected; the device is more convenient to move through arrangement for universal wheels; and the combustion condition of the fuel can be effectively observedthrough an observation port in a combustion chamber.

The invention discloses a method for elastic virtual subnet mapping in a wireless mesh network and mainly solves the problems of high mapping cost and low link mean availability of the existing wireless mesh network in resource allocation and virtual request mapping processes. The technical scheme comprises the following steps of (1) comprehensively considering minimization of load balancing and virtual request mapping cost of an underlying physical network in the wireless mesh network and establishing a combined dynamic optimization problem model; (2) decomposing a combined dynamic optimization problem into two continuous static sub-problems; (3) describing the two continuous static sub-problems, namely target functions and constraint conditions of sub-channel resource allocation sub-problems and virtual network mapping sub-problems of a physical layer and an MAC layer; and (4) solving the two continuous static sub-problems to obtain an optimal resource allocation matrix and an optimal link mapping variable matrix. According to the method, the link mapping cost is reduced, the link mean availability is improved, and the method can be used for virtualization of the wireless mesh network.

The embodiment of the invention discloses a method and device for task management. By application of the technical scheme, by combination with the actual input data size of tasks and current quota use rate of a group of a cluster where the tasks are located, and adaptability adjustment is performed on corresponding resources. Thus, in a large computing cluster, computing resources are distributed individually distributed for running tasks, it is guaranteed that the tasks obtain reasonable resources, task execution efficiency is improved, the resource utilization rate of the computing cluster is ensured, and task concurrency is improved.

The invention provides a heterogeneous cascade network QoS closed-loop mutual coupling system and method. The system comprises a LAN QoS control module, a WAN QoS control module and a cascade engine, the LAN QoS control module, the WAN QoS control module and the cascade engine respectively comprise a control plane entity unit and a data plane entity unit, and the control plane entity units of the LAN QoS control module, the WAN QoS control module and the cascade engine respectively comprise a user attribute memorizer SPR module, a P-CSCF module, a PDF module and a PEP module; the data plane entity units of the LAN QoS control module, the WAN QoS control module and the cascade engine respectively comprise a QoS queue module and a channel queue module. With the system and method, the QoS priority on the LAN side and the WAN side can be dynamically regulated, QoS closed-loop mutual coupling is achieved, the QoS of a cascade network is guaranteed, the QoS is improved, and the utilization rate of resources is guaranteed at the same time.

The invention discloses a resource distribution method and device and a storage medium, and belongs to the technical field of cloud computing. The method comprises the following steps: determining bidding information of a plurality of bidders requesting to acquire virtual resources; based on the bidding information of each bidder in the plurality of bidders, obtaining a target bidder from the plurality of bidders through a resource auction model, the target bidder being a bidder for maximizing social welfare, and the resource auction model being an auction mechanism for describing the maximization of social welfare; and allocating corresponding virtual resources to the target bidder. In the application, the resource auction model is an auction mechanism describing social welfare maximization, and the target bidder determined from the plurality of bidders through the resource auction model is the bidder enabling the social welfare maximization, so that the resource utilization rate is ensured. Meanwhile, as long as the bidder who needs to acquire the virtual resources has the chance to bid, the possibility that any bidder acquires the virtual resources is ensured, and cross-base-station sharing of the resources is realized.

The invention discloses an aggregate proportioning door with the adjustable opening degree, which comprises a stone material cabin, an emptying door steam cylinder, a nylon sleeve and emptying doors, wherein the stone material cabin is arranged above the emptying doors, and the emptying door steam cylinder is arranged between the stone material cabinet and the emptying doors. The aggregate proportioning door is characterized in that the number of the emptying doors is two, the two emptying doors are respectively a large emptying door and a small emptying door, the diameter of the large emptying door is smaller than the diameter of the small emptying door, and the nylon sleeve is arranged on the emptying door steam cylinder. The aggregate proportioning door has the beneficial effects that through the arrangement of the large emptying door and the small emptying door, the metering accuracy of the aggregate can be greatly improved, in addition, the rough metering process and the precise metering process are separated, and the maximum resource utilization can be ensured.