43results about How to "Save CPU time" patented technology

The present invention provides a method and apparatus for a bandwidth adaptive image compression / decompression scheme. In one embodiment, the present invention uses a special protocol between the sender and the receiver to measure the latency of the connection. This protocol and its result are sent and received at an interval based on a dynamic feedback loop algorithm. Based on the results of the protocol, a compression scheme is chosen. This scheme uses CPU time conservatively, and also transmits the most interesting data first. In another embodiment, the present invention throws away data that is repetitious, especially when the connection is down for a short period of time. In yet another embodiment of the present invention, measurements are taken for the perceptual degradation of the image for various compression schemes, and the results are supplemented with the results of the protocol to choose a viable compression / decompression scheme.

The invention relates to a numerical method, and provides and solves a novel two-dimensional Euler equation in Lagrange form to solve inverse problems of solid wall face geometric shape design. The invention provides a mapping mode for deducing the Euler equation on a Lagrange plane, computational grid is simplified, and numerical dissipation of convective terms is lowered furthest. Using the numerical method of using the Euler equation in Lagrange form to solve the inverse problems of one kind can simultaneously obtain solution of flow field physical quantity and solution of the solid wall face geometric shape design.

The invention discloses a dynamic prediction-based network driver layer data packet receiving method, which comprises the following steps of establishing a blank chain table and a circulation sequence, and creating a kernel thread and a kernel timer; receiving network equipment interruption, executing an interruption processing program, and stopping receiving network equipment interruption immediately after entering the interruption processing program; judging whether a network equipment interruption instruction is a receiving instruction or a wrong instruction; if the instruction is a receiving instruction, judging whether a data arriving bit in a receiving descriptor of the receiving instruction is zero or not, if not, judging whether the number of data packet caches in the blank chain table is lower than a threshold or not, if not, obtaining a new data packet cache from the blank chain table and judging whether obtaining is successful or not, if not, counting the length of effective data in the data packet cache corresponding to the current receiving descriptor, and discarding the effective data. By the method, the number of the pre-distributed data packet caches can be calculated to adapt to the change of the network flow, and the phenomenon of consuming too much memory when the data flow is lower is avoided.

The invention relates to a multi-channel software-hardware coordinated calling method for the purpose of narrowing the gap between the increasing number of sophisticated spacecrafts for controlling computer serial communication and limited software-hardware resources of embedded systems. The method comprises following steps of reusing hardware multichannel series channels; sending, caching and reusing, receiving cache independence, setting channel numbers via software to achieve channel switch, and setting send data lengths of corresponding channels and receive data lengths; reusing receive interruption and overtime interrupt of serial ports, adopting to same receive interruption and overtime interrupt by multiple channels of series ports and occupying external interrupt resources of the same CPU; and carrying out software-hardware parallel work in order to achieve serial communication of multiple channels. The serial communication of multiple channels is finished by main processes of software tasks, interruption processes of software and hardware processing. The multi-channel software-hardware paralleled serial communication processing method helps to achieve multi-channel software-hardware paralleled serial communication in order to save computer hardware resources, and reduce CPU time occupied by multi-channel serial communication and is applied for controlling a computer of a certain type.

The invention discloses an optimization method for improving Bluetooth positioning accuracy, and the method comprises the steps: setting a cache container to obtain and store gateway data between eachgateway and a positioning label in a time period at a time, enabling the gateway data to become a calculation period, and enabling the gateway data to be a signal intensity value between each gatewayand the positioning label; Grouping the gateway data according to the positioning tags, selecting the gateway with the highest signal intensity value from the list of the target positioning tags as areference gateway point, storing the reference gateway point in a cache container, and selecting the gateway with the signal intensity value only smaller than that of the reference gateway point fromthe list of the target positioning tags as an auxiliary gateway point; And converting the intensity values between the reference gateway point and the target positioning tag and the intensity valuesbetween the auxiliary gateway point and the target positioning tag into distance values, and calculating the coordinate value of the target positioning tag according to the coordinate positions of thereference gateway point and the auxiliary gateway point. By combining the physical characteristics of the Bluetooth positioning tag and the Bluetooth positioning gateway, the positioning accuracy ofthe Bluetooth positioning tag is improved.

The invention discloses a multi-resolution WENO format and ILW boundary processing combined fixed-point rapid scanning method, and the method comprises the following steps: converting a steady hyperbolic conservation law problem into a time-dependent hyperbolic conservation law problem, and processing an object surface boundary by using a novel five-order ILW boundary processing method; discretizing a space part of the hyperbolic conservation law equation by using a finite difference multiple resolution weighting basic oscillation-free format; discretizing a time part in the control equation by using a three-order Runge-Kutta method and a fixed-point rapid scanning method; obtaining an approximate value of each point of the next time layer according to a space-time full-discrete method, and obtaining a numerical simulation result when the residual error of the conservation variable in the calculation region tends to be stable. According to the method, the advantage that the high-orderprecision of the calculation format near the boundary of the object plane is not reduced can be kept, the use number of space templates is effectively reduced, parameters needing to be manually adjusted are reduced, implementation is simpler, and implementation is easy.

A method for solving RAM (Random Access Memory) fragments generation during drawing of a Windows GDI (Graphics Device Interface) comprises the following steps: creating a class when the system is constructed, and creating a global static unique memory bitmap object and two global static unique memory equipment context objects; when the system starts initiation, defining all the window classes in the initiation memory bitmap objects and memory equipment context objects as derived classes, and performing painting interface operation in the OnPaint () method of the window classes through directly invoking the memory equipment context objects. According to the invention, when the WindowMFCGDI is adopted to perform GUI drawing, the memory equipment context objects and the memory bitmap objects, both of which are used during painting of each window, are not required to be created by the WindowMFCGDI; instead, the global unique memory equipment context objects and the memory bitmap object, both created in position, can be used, so that memory is not required to be created and allocated for each painting, CPU time occupied by frequent creating is shortened, and the stability and speed of system running are increased.

The invention provides a method for optimizing the performance of an encryption card based on DEVICE MAPPER. The API of the encryption card is integrated into the universal block device kernel framework Device Mapper so that a dm-crypt encryption block device module suitable for the encryption card can be realized, and furthermore, an encrypted file system can be established for protecting confidential data at a disk storage level; the encryption block device module comprises two modes, namely synchronous block encryption and asynchronous block encryption of the domestic encryption algorithm sm1, and a method for registering the algorithm sm1 to a kernel password manager is provided. Compared with the prior art, the method for optimizing the performance of the encryption card based on DEVICE MAPPER is characterized in that the increase of the speed is realized by virtue of parallel operating of a plurality of cards, and in the aspect of system resource utilization rate, principally, the previous encryption card working pattern is changed from waiting mode into an asynchronous working mode, and therefore, the time of the cpu is greatly saved.

The conversion device includes an input for receiving data corresponding to an image to be displayed. The received data is in a JPEG decoder output data format A processor is included for reconstructing and writing the image to be displayed into the image memory, in a display module expected input data format. The bandwidth of the image memory is greater than one byte. The processor is fully hardwired and includes a first logic stage for writing the received data byte by byte into an intermediate memory at chosen addresses such that the written data form a sequence of data in the display module expected input data format, and a second logic stage for reading the written data in the intermediate memory, forming successive packets of read data having a size corresponding to the bandwidth, and successively writing the packets into the image memory at chosen addresses such that the written packets together form all the lines of the image.

The invention discloses a keyboard circuit capable of improving key-press scanning speed. In the keyboard circuit, one ends of each column line and row line are connected with a power supply through a current limiting resistor; the other ends of the column line which are multiplexed by a PNP (positive-negative-positive) switching transistor and the other ends of the m column lines which are multiplexed by an NPN (negative-positive-negative) switching transistor are connected correspondingly with the I / O interfaces, namely P1.1, P1.2,...,P1.m after being multiplexed; meanwhile, the half of therest column lines which are multiplexed by the PNP switching transistor and the other half which are multiplexed by the NPN switching transistor are correspondingly connected with I / O interfaces, namely P[1.(m+1)], P[1.(m+2)],...,P[1.(m +n) / 2]; the base of each switching transistor is connected to the I / O interface P[1.0] through the current limiting resistor; the column line and the row line of each key-press are connected with the ground wire through a combination switch; all the column lines are connected to the inputs of interrupt Input port through an AND circuit unit. The keyboard circuit provided by the invention has the advantages that when the sum of the column lines and the row lines of the key circuit is 2N, the required number of the I / O interfaces is N+1, and when the sum of the column lines and the row lines is 2N+1, the required number of the I / O interfaces is N+2 and the values of column and the row can be obtained after scanning for two times.

The invention discloses a network-on-chip branch defining task mapping method. The method comprises the steps of determining a task graph of an application needing to be designed, a delay limiting value of two task node connecting edges defined based on a transmission bandwidth requirement and a selected NoC topological structure; and starting from the root node, carrying out path traversal on each tree-shaped searched branch, mapping task nodes of unmapped sub-task nodes to routing nodes which are not occupied in the NoC array until all the task nodes in the task graph are mapped to the routing nodes in a one-to-one correspondence manner, and finishing mapping to form a task mapping scheme. According to the method, branch polling is constrained by adopting the data transmission quantity between the task nodes as a limit, so that the CPU time for developing a mapping scheme is reduced, the network load of the system is balanced, and the performance of an application system based on heterogeneous NoC can be greatly improved.

The present invention relates to a redundancy filling method of semiconductor manufacturing process. The method involves providing a circuit pattern, generating a density report for the circuit pattern that identifies a feasible area for dummy insertion, simulating a planarization process with the density report and identifying a hot spot on the circuit pattern, inserting a virtual dummy pattern in the feasible area and adjusting the density report accordingly, and thereafter simulating the planarization process with the adjusted density until the hot spot is eliminated. The present invention provides a method for reducing the dummy metal count in a circuit design thereby saving mask making time, CPU runtime, and data storage memory. This will help achieve design timing closure quicker and easier.

A double focal spot X-ray tube is used to acquire a set of two images PIα, PIβ for a given gantry position from slightly different view positions. The stereo or binocular disparity (BD) of imaged structures is used to estimate the object depth in view direction, which in turn is used to discriminate between objects inside IO and outside EO the body. Respective structures are virtually removed from the images PIα, PIβ.

The invention relates to the router design and research field of the mobile wireless self-organized network, which is more particularly characterized by including the flowing steps sequentially: the initial TTL values of the IP messages sent by all nodes are set to be 64. When the node sends an IP message, a router stored to a target IP is only sent to a next corresponding hop, or floods the messages to all the neighbors. When the node receives the IP message, the retransmitting hop times passed by the message can be analyzed according to the current TTL value of the message, and the shortest path to the source can be determined by combining the corresponding routing items of the node. The non-shortest routing produced in the process of flooding can be deleted for overtime as no corresponding IP message is transmitted. The invention realizes a novel rapid and simple Ad Hoc path-finding method with low protocol bandwidth consumption.

The invention discloses a monitoring method and system for MPEG video transmission quality, and relates to the field of video transmission quality monitoring. The DF calculation process of this method includes: after the monitoring node segments the video encoding data payload of the IP data packet in MPEG format, writes it into the MR file; For file segmentation, calculate MR according to the PCR and traversal offset of two file segments; after the monitoring node calculates the virtual cache VBpre and VBpost before and after receiving the IP packet, calculate DF according to the VBpre and VBpost calculated within the specified time; The MLR calculation process includes: the monitoring node calculates the MLR according to the SEQ field or the CC field of the IP data packet. The invention can accurately and efficiently calculate the DF and MLR of the video stream, and monitor the transmission quality of the MPEG video through the DF and the MLR.

The invention discloses an optimization method for improving Bluetooth positioning accuracy, and the method comprises the steps: setting a cache container to obtain and store gateway data between eachgateway and a positioning label in a time period at a time, enabling the gateway data to become a calculation period, and enabling the gateway data to be a signal intensity value between each gatewayand the positioning label; Grouping the gateway data according to the positioning tags, selecting the gateway with the highest signal intensity value from the list of the target positioning tags as areference gateway point, storing the reference gateway point in a cache container, and selecting the gateway with the signal intensity value only smaller than that of the reference gateway point fromthe list of the target positioning tags as an auxiliary gateway point; And converting the intensity values between the reference gateway point and the target positioning tag and the intensity valuesbetween the auxiliary gateway point and the target positioning tag into distance values, and calculating the coordinate value of the target positioning tag according to the coordinate positions of thereference gateway point and the auxiliary gateway point. By combining the physical characteristics of the Bluetooth positioning tag and the Bluetooth positioning gateway, the positioning accuracy ofthe Bluetooth positioning tag is improved.

An image reconstruction apparatus and related method. The amount of out-field-of view material for a CT scanner (IMA) with a given field of view (FoV) in a bore (B) is established. Based on the measurement, a hybrid-image reconstructor (RECONX) is configured to switch between different reconstruction algorithms.

The invention provides a method for correcting the low-level cut height of an infrared occultation sensor based on a non-absorbing spectrum. The detection spectral data above the height; Step 2: Analyze the detection spectral data to determine the low-level cut height range to be corrected; Step 3: Through the HITRAN spectral absorption library, analyze the effective observation spectrum of GF-AIUS (750-4100cm ‑1 ) within the position of the completely non-absorbing band; step 4: high-pass filter the spectrum of the full effective band, extract the spectral data of the non-absorbing band within the corresponding cut height range, and calculate the scaling factor; step 5: according to the scaling factor, Correct the spectrum on the lower cut height; Step 6: Carry out atmospheric correction on the lower cut height. The beneficial effects of the present invention are: the rapid height-cut correction method of the present invention is simpler, faster and more accurate.