40 results about "Forward recursion" patented technology

Decoding signals represented by a trellis of block length N divided into windows of length L includes a step of decoding a forward recursion from a point P that is before the beginning of a window up to the beginning of the window. P is chosen at a sufficient distance from the beginning of the window such that forward recursion determines a known state metric at the beginning of the window. A next step includes decoding the window using forward recursion from the known state at the beginning of the window up to the end of the window to define a set of known forward recursion state metrics which are stored. A next step includes decoding using backward recursion starting from a known state at the end of the window and moving backward. A next step includes calculating a soft output at each stage of the backward recursion using the stored forward recursion state metrics, and branch metrics at each stage, and outputting the soft output for that stage in a LIFO format.

The invention provides a multi-target positioning method of a bistatic multi-input multi-output radar, comprising the following steps of: (1) transmitting mutually orthogonal phase-coded signals by M transmitting array elements, and receiving the phase-coded signals by N receiving array elements, wherein the distances of the M transmitting array elements and the N receiving array elements are all of half wavelengths; (2) carrying out matched filtering on the received phase-coded signals by a matched filter of a receiver of each receiving array element; (3) carrying out multistage Wiener filtering on a matched signal data covariance matrix space, and carrying out forward recursion to obtain a signal subspace; (4) carrying out high-resolution DOA (Direction of Arrival) estimation by using an ESPRIT algorithm, wherein a pairing algorithm is used for carrying out the automatic pairing on two-dimensional parameters; and (5) realizing multi-target positioning according to cross points at two angles so as to obtain the positions of space targets. The multi-target positioning method provided by the invention has the advantages of low computation complexity, high computation speed, high estimation accuracy and can be used for positioning the sea-surface or low-altitude targets during tracking and guidance.

Decoding signals represented by a trellis of block length N divided into windows of length L includes a step of decoding a forward recursion from a point P1 that is before the beginning of a window up to the beginning of the window and decoding a backward recursion from a point P2 that is after the end of a window back to the end of the window to define known states at the beginning and end of the window. A next step includes decoding the window using backward recursion from the known state at the end of the window back to the beginning of the window to define a set of backward recursion state metrics. A next step includes decoding using forward recursion starting from a known state at the beginning of the window and moving forward to the end of the window to define a set of forward recursion state metrics. A next step includes calculating a soft output at each stage of the window using the forward and backward recursion state metrics, and branch metrics at each stage, and outputting the soft output for that stage.

The invention discloses a space manipulator modeling method based on different geometry. On the foundation that each level rigid body zero position of a space manipulator system is selected, initializing conditions such as a zero position joint vector, a centroid joint vector and an initial vector are calculated and given, and a matrix related to initial parameters and a coordinate position matrix are calculated. The generalized velocity of each connecting rod is calculated through forward recursion, and the generalized force of each connecting rod is calculated through reverse recursion. Each quantity is substituted into the matrixes, and a kinetic equation compact in form is written down. The model has the advantages of being uniform in modeling form and small in operating amount, improving calculation efficiency and accuracy of the space manipulator modeling, and improving the designing accuracy of the space manipulator. The space manipulator modeling method based on differential geometry can be applied to modeling of other space multi-rigid-body systems.

A wireless receiver and method employs a forward / backward recursive covariance based filter coefficient generation scheme for equalizing a received wireless signal, such as a CDMA signal or other suitable signal. A forward / backward recursive covariance based filter coefficient generating circuit (206) includes a forward recursion portion (300) that receives data representing channel estimates (18) of a communication channel, and data representing an auto-covariance matrix (22). The forward recursion portion (300) produces a set of output values (306) from each iteration stage (400a–400n) of the forward recursion portion (300). The set of output values (306) are stored and a backward recursion portion (302) obtains the set of output values (306) and also receives the data representing the auto-covariance matrix (22), and generates filter coefficients (208) based on this information.

Decoding signals represented by a trellis of block length N divided into windows of length L includes a step of decoding a backward recursion from a point P that is after the end of a window back to the end of the window. P is chosen at a sufficient distance from the end of the window such that backward recursion determines a known state metric at the end of the window. A next step includes decoding the window using backward recursion from the known state at the end of the window back to the beginning of the window to define a set of known backward recursion state metrics which are stored. A next step includes decoding using forward recursion starting from a known state at the beginning of the window and moving forward. A next step includes calculating a soft output at each stage of the forward recursion using the stored backward recursion state metrics, and branch metrics at each stage, and outputting the soft output for that stage.

The invention discloses a double-duality Turbo code encoding method and encoder based on DVB-RCS standard. (1) the double-bit system bit of the receiving code word is transmitted into an encoder to find the initial values of the forward recursion factor Alpha1 and the backward recursion factor; (2) the log likelihood ratio Lambada 1 is found; (3) the external information Le1 is found; (4) the external information Le1 is interleaved to achieve the prior information La2; (5) the double-bit system of the receiving code word is transmitted into an interleaver; (6) the interleaving result is transmitted into an pre-encoder to find the initial values of the forward factor Alpha 2 and the backward recursion factor Beta 2; (7) the log likelihood ratio Lambada2 is found; (8) the external information Le2 is worked out; (9) the external information Le2 is interleaved to find the prior information La; (10) Steps (2)-(4) and (7)-(9) are repeated until the rule of final iteration encoding is satisfied; (11) the log likelihood ratio Lambada2 is put into a de-interleaver to carry out a hard decision to get the final encoding bit. The invention is easy to realize the hardware, applicable to a plurality of optional code rates and flexible block lengths; besides, the power consumption is small and the error rate is low.

The invention provides a decoding method of a convolutional Turbo code for reducing decoding time delay and saving a memory. The decoding method comprises the following steps of: simultaneously carrying out forward recursion and backward recursion in a component decoding process; dividing the forward recursion and the backward recursion into two stages with equivalent computation quantity; and sequentially calculating and obtaining posterior likelihood ratio information at the beginning of the second stage. The time delay from the beginning of recursion operation to the end of the posterior likelihood ratio information operation is shortened once compared with the traditional decoding process. Furthermore, the traditional posterior likelihood ratio operation is serial, while the posteriorlikelihood ratio operation of the invention is carried out bidirectionally and simultaneously in parallel, the required calculation time and the recursive calculation time are overlapped, and it is unnecessary to distribute additional calculation time; in addition, a bidirectional parallel structure can ensure that the memory used for storing state metric is reduced by half. Furthermore, through the calculation of splitting branch metric, redundancy calculation is reduced, and the space for storing the branch metric is reduced by half.

A method of adaptive filtering using fast affine projection (FAP) that allows for direct solution of the projected error vector from the autocorrelation matrix using a backward and forward recursion technique or LDLT factorization of the autocorrelation matrix followed by forward substitution, scaling and backward substitution. The method results in less computational complexity in the implementation, while still having good stability and fast convergence.

The embodiment of the invention discloses a Turbo parallel decoding method, a device and a system. The method comprises the following steps that: when the current iterative number i is not equal to 0and forward recursion initial coefficients of initial symbols of No. 2 to No. N parallel processing windows are refreshed to iterative codes of the previous round, the forward recursion coefficients of the last symbols of backward neighbor windows of the windows calculate the forward recursion coefficients of each symbol of the windows according to the forward recursion initial coefficients; whenbackward recursion initial coefficients of the last symbols of No. 1 to No. P-1 parallel processing windows are refreshed to iteration of the previous round, the backward recursion coefficients of theinitial symbols of forward neighbor windows of the windows calculate the backward recursion coefficients of each symbol of the windows according to the backward recursion coefficients; and P paralleldecoders are adopted to carry out parallel decoding of the forward recursion coefficients and backward recursion coefficients of P parallel processing windows. The embodiment of the invention can reduce decoding delay and buffer units required by parallel decoding so as to improve decoding performance.

A MAP decoder, or a turbo decoder having constituted MAP decoders, is configured with bi-directional sliding windows. A sliding window architecture is applied to the forward state metric (FSM) calculation of an encoded data block having a trellis with N time-steps. The data block is divided into smaller sub-blocks and a forward recursion of two or more of the sub-blocks are performed in parallel to obtain FSMs. Each sub-block overlaps with a previous sub-block by k time-steps, where k is an integer value greater than zero. This provides a good approximation of the FSMs at time-step k + t of the sub-block. The FSMs associated with the first k time steps of each sub-block, other than the first sub-block, are discarded. A reverse recursion of each sub-block is also performed to obtain reverse state metrics (RSM) for the sub-block. Likelihood ratios are then calculated based on the FSMs and RSMs.

In a digital information processing system wherein a model of a finite state machine (FSM) receiving a plurality of FSM inputs and producing a plurality of FSM outputs is represented by a reduced-state trellis and wherein the FSM inputs are defined on a base closed set of symbols, a novel method is presented for updating soft decision information on the FSM inputs into higher confidence information whereby (1) the soft decision information is inputted in a first index set, (2) a forward recursion is processed on the input soft decision information based on the reduced-state trellis representation to produce forward state metrics, (3) a backward recursion is processed on the input soft decision information based on the reduced-state trellis representation to produce backward state metrics, wherein the backward recursion is independent of the forward recursion and (4) the forward state metrics and the backward state metrics are operated on to produce the higher confidence information.

The present invention relates to a method for decoding data, said method using windows (WID) of input data The invention is characterized in that it comprises for a current window the steps of:—Performing a forward recursion, wherein said forward recursion is initialized with a forward state metric vector (a) from the upper stake (STK) of the previous window of the same step of the previous iteration, a window (WID) comprising a lower and an upper stake (STK), and—Performing a backward recursion, wherein said backward recursion is initialized with a backward state metric vector (b) from the lower stake (STK) of the next window of the same step of the previous iteration. Use: Receiver in a communication system

The invention relates to a current harmonic analysis method based on a trigonometric function neural network. Harmonic analysis of the electric vehicle charging current is achieved by aiming at the harmonic waves generated by an electric power system when charging an electric vehicle. The electric vehicle charging current is converted into a trigonometric function expressed by weight, a trigonometric function neural network is constructed, and forward recursion is carried out through the trigonometric function neural network. The output current containing all the harmonic components is obtained; the output current is compared with the input current, and a negative gradient descent method is adopted for carrying out reverse iteration of the trigonometric function neural network on the difference value of the output current and the input current to obtain the optimal weight of the trigonometric function neural network, so that accurate estimation of the charging current harmonic parameters of the electric vehicle is obtained. The method has higher convergence characteristic and better noise tolerance.

A method is provided for performing a MAP probability decoding of a sequence R(n) including N bits of encoded data. The method includes the steps of: (a) generating a sequence rn of sot-values by processing the sequence R(n); (b) performing a forward recursion by computing alpha values αS,SG utilizing the soft-decision values; (c) performing a backward recursion by computing beta values βS,SG utilizing the soft-decision values; and (d) performing an extrinsic computation by computing probability values p′k. The alpha values αS,SG are relative log-likelihoods of an encoding process arriving at various states. The beta values βS,SG are relative log-likelihoods of the encoding process arriving at various states. The probability values p′k represent a set of probabilities indicating that each data bit of an input sequence dK had a value equal to zero or one. The sequence R(n) represents an encoded form of the input sequence dK.

A device for detecting an estimated value for a symbol at a given time, which is supplied to a phase modulation and transmitted via a transmission channel with a time-variable phase, provides a unit for determining log weighting factors in a forward recursion, a unit for determining complex coefficients in a forward recursion, a unit for determining log weighting factors in a backward recursion, a unit for determining complex coefficients in a backward recursion, a unit for determining an extrinsic information, a unit for determining the phase factor with the maximal weighting factor in a forward recursion and a unit for determining the phase factor with the maximal weighting factor in a backward recursion.

The invention discloses a continuous phase demodulation method based on reliability, which comprises the following steps: initializing a forward recursive variable and a backward recursive variable, and calculating the reliability information of each edge of the nth Tellis after correction; carryint out Forward recursion and backward recursion calculation, and translating an information vector atthe same time; And carrying out information extraction calculation and judgment calculation on the nth symbol of the CPM signal. According to the continuous phase demodulation method based on the reliability, the correlation value between the received signal and the transmitted modulation signal is used as the information measurement, the calculation complexity is greatly reduced, and the requirement for hardware resources is reduced on the premise that the demodulation performance is not sacrificed. The method solves the problem of large calculation amount of continuous phase demodulation, can be applied to various communication systems based on CPM modulation and demodulation, does not need to estimate channel noise, and has very high robustness.

The invention discloses an iteratively decoding method and system for parallel Turbo codes. The method comprises the steps of storing a Turbo code block after serially receiving codes into a random access storage unit in a set storing manner; then performing backward recursion operation on data in all the random access storage units from the tail to head according to a backward recursion formula,and performing forward recursion operation on the data in all the random access storage units from the head to tail according to a forward recursion formula; and at last, judging whether the number ofcurrent iterations is equal to a set iteration amount, if yes, computing a log-likelihood ratio according to recursion operation results and performing hard judgment to acquire a decoding result, andotherwise, computing prior information according to the recursion operation results, and returning back to the recursion step. The method and system provided by the invention has the advantage that the less redundant storage resources are used, so that the preliminary recursion process of the initial value of each sub-block in the traditional parallel Turbo iterative decoding structure is simplified and combined into a recursion process, and thus the decoding control complexity is reduced and the design timing sequence is optimized.

A method and apparatus are provided that reduce the amount of memory needed to perform forward-backward recursion to identify posterior probabilities. Under the invention, a forward recursion is performed to identify forward recursion scores. The forward recursion scores are then used directly in a backward recursion to determine posterior probabilities for each state in a set of time frames. The usefulness of this invention is especially high when there is a large number of discrete states, such as when there are more than one set of discrete states in the model.

The present invention relates to a method for decoding data, said method using windows (WID) of input data. The invention is characterized in that it comprises for a current window the steps of:—Performing a forward recursion, wherein said forward recursion is initialized with a forward state metric vector (a) from the upper stake (STK) of the previous window of the same step of the previous iteration, a window (WID) comprising a lower and an upper stake (STK), and—Performing a backward recursion, wherein said backward recursion is initialized with a backward state metric vector (b) from the lower stake (STK) of the next window of the same step of the previous iteration. Use: Receiver in a communication system.

The invention relates to a battery energy storage system scheduling method considering dynamic charge and discharge efficiency. The method comprises the following steps: A, acquiring basic informationof a battery energy storage system, and performing segmented linearization on a charge and discharge efficiency characteristic curve; B, acquiring scheduling information such as running limit, initial and final electricity quantities, a predicated time-sharing electricity price or a price multiplier and the like of the battery energy storage system; C, establishing a scheduling model; D, calculating the dynamic upper limit and lower limit of storing the electric quantity by the battery energy storage system in each time period; E, performing dynamic planning on reverse recursion to obtain anoptimization model of a core sub-problem needing to be repeatedly solved in each time period; F, using geometric features of the core sub-problem optimization model to sequentially solve an optimal path in each horizontal sub-region from bottom to top; G, solving the optimal path in the horizontal sub-regions; H, processing the optimal path jump; I, performing dynamic planning on forward recursion, and calculating the optimal charge and discharge electric quantity and power decision of each time period; and J, entering the next time period along with the schedule development, updating the scheduling information in the step B, and executing the step C. According to the method, an effective way is provided for optimized operation of the battery energy storage system, and an engineering practical value is achieved.

The invention belongs to the technical field of numerical control systems, and particularly relates to a speed curve planning method for a continuous processing path, which comprises a look-ahead section and a time-sharing processing method for a fixed section length of the look-ahead section. The method comprises the following steps: 1, carrying out small-section look-ahead on the divided look-ahead section, carrying out backward recursion in sequence, carrying out forward recursion on a small-section curve to obtain a forward recursion speed, and ensuring that the forward recursion speed is less than or equal to an inflection point speed limit; and 2, backtracking the points which do not meet the forward recursion condition to obtain a backtracking speed, and when the backtracking speed is greater than the forward pushing speed, determining that the requirement is met, and ending the look-ahead of the small section. According to the method, a simplified method combining forward recursion and backtracking is adopted for speed look-ahead planning, a method of continuing backtracking is adopted when the backtracking speed is not met, and when the backtracking speed is still not met in the first section, jerk backtracking is modified, so that the problem that the backtracking speed is not met is solved, and the speed curve is more stable.

The invention discloses a receiving signal detection method for a large compression ratio FTN system, belonging to the technical field of single-carrier communication. The present invention uses M-algorithm to carry out the forward recursive calculation of BCJR algorithm, with delay L β Calculate the LLR output, and use backtracking to traverse the state to avoid unstable LLR output. The present invention is used in a single-carrier FTN system based on the ultra-low complexity M-BCJR algorithm for the receiver to decode and detect the received signal. Compared with the existing M-BCJR algorithm, it can be realized under the same performance conditions , greatly reducing the complexity.

The invention provides a low-code-rate biorthogonal code decoder and a decoding method. The decoder comprises two branch decoders; each branch decoder comprises an extrinsic information and prior information storage unit, a backward metric calculation and storage unit and a forward metric and posterior information calculation unit; the extrinsic information and prior information storage unit is used for determining prior information based on posterior information and stored extrinsic information; the backward metric calculation and storage unit is used for performing backward recursion estimation based on a signal to be decoded and the prior information and outputting a backward metric value; the forward metric and posterior information calculation unit is used for performing forward recursive estimation based on the signal to be decoded, the prior information and the backward metric value, and outputting a forward metric value and posterior information; if the number of times of forward recursion estimation reaches the preset maximum number of iterations, a decoding result is outputted; the complexity of the structure and the implementation mode of the biorthogonal code decoder is reduced, and the low-bit-rate biorthogonal code decoder is easy to apply in engineering.

A method of adaptive filtering using fast affine projection (FAP) that allows for direct solution of the projected error vector from the autocorrelation matrix using a backward and forward recursion technique or LDLT factorization of the autocorrelation matrix followed by forward substitution, scaling and backward substitution. The method results in less computational complexity in the implementation, while still having good stability and fast convergence.

The invention discloses a modeling method for hysteresis characteristics of a small-stroke nanometer motion platform based on forward recursion, and belongs to the technical field of ultra-precision motion control. The modeling process comprises four parts of sampling data acquisition, model construction, algorithm solution and result output, the sampling data acquisition method is described in a small-stroke nanoscale motion table and thermal correlation hysteresis data measurement method, and current-force correlation data acquisition under different temperature and different displacement conditions is acquired; the model construction part comprises primary function determination and model structure and parameter determination; the algorithm solving part comprises initial condition setting, RBF network construction, RBF center number determination and weight output. According to the method, model parameters are quickly identified based on input and output data of a multi-input single-output nonlinear system, the model precision is ensured while the calculation amount is reduced, the problem that a traditional least square algorithm is unstable due to matrix morbidity is solved, and then the performance of a motion platform is improved through inverse model compensation.

The invention discloses a system state reverse estimation method for introducing a finite estimation interval. The method comprises the steps: (1) introducing the finite estimation interval [m, n], m= n-N + 1, m and n are two different moments, and N is the step length of the estimation interval; (2) calculating a state estimation value and a variance Pn of the system at the moment n based on a forward recursion Kalman filtering algorithm; and (3) calculating a state estimation value and a variance Pm of the system at the moment m based on a reverse recursion Kalman filtering algorithm by taking the initial state and Pn as an initial variance, and abandoning other state estimation values and variances in the estimation interval [m, n] to complete estimation of the system state. The reverse estimation method can be used for initializing a forward estimation algorithm, and on the other hand, the historical state of the system can be reconstructed under the condition that the current state estimation value is known.

A method and apparatus are provided that reduce the amount of memory needed to perform forward-backward recursion to identify posterior probabilities. Under the invention, a forward recursion is performed to identify forward recursion scores. The forward recursion scores are then used directly in a backward recursion to determine posterior probabilities for each state in a set of time frames. The usefulness of this invention is especially high when there is a large number of discrete states, such as when there are more than one set of discrete states in the model.