Internet of Things information updating strategy method based on non-ground network
By communicating with LEO satellites on the ground user equipment, combining LDPC encoding and channel prediction methods of long and short-term memory network layers, the problem of insufficient channel prediction accuracy in non-terrestrial networks is solved, and the uplink peak throughput and channel estimation performance of the satellite communication system is improved.
Patent Information
- Application Number
- CN202510535631.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-04-27
- Publication Date
- 2025-07-04
AI Technical Summary
Traditional channel prediction methods lack prediction accuracy and stability in complex non-terrestrial network environments, fail to effectively track channel changes, affecting the performance of satellite communication systems.
The IoT information update strategy based on non-terrestrial networks is adopted, and the data is processed through ground user equipment, LDPC channel coding, bit interleaving, M-QAM mapping and orthogonal frequency division multiplexing is used to process the data, and pilot symbols are inserted at the transmitting end, channel frequency response matrix estimation and interpolation are performed on the receiving end, and channel prediction is performed on the convolution layer configured with specific parameters and long and short-term memory network layer.
The uplink peak throughput is significantly improved, the channel estimation performance is almost unaffected, the predictor performs stable within a limited range and is scalable, and future hardware implementations are expected to achieve more accurate prediction performance.
Smart Images
Figure CN120263351A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of Internet of Things information update, and specifically to an Internet of Things information update strategy method based on non-terrestrial networks. Background Art
[0002] In a communication system, obtaining accurate channel state information (CSI) is crucial for system performance. Channel prediction, as a method to obtain CSI when the channel state cannot be actually estimated, has been widely applied in the communication field. Non-terrestrial networks (NTNs), as an important part of the 6G wireless communication standard, have typical characteristics of dynamic channel conditions, such as time-varying delay and Doppler frequency shift. The channel aging effect will lead to a decrease in the accuracy of the estimated channel information, affecting the performance of satellite communication systems. Therefore, channel prediction is particularly critical in NTNs.
[0003] Currently, most research works on channel prediction use the mean square error (MSE) as an index, but ignore the impact on key communication metrics, making it difficult to meet the performance requirements of the 6G NTN system. Summary of the Invention
[0004] The purpose of the present invention is to provide an Internet of Things information update strategy method based on non-terrestrial networks to solve the problem that traditional channel prediction methods mainly rely on some basic mathematical models and algorithms, resulting in insufficient prediction accuracy and stability when facing complex non-terrestrial network channel environments. For example, some methods do not fully consider the impact of the mobility of NTN nodes on the channel, leading to the inability to effectively track channel changes in practical applications.
[0005] To achieve the above purpose, the present invention provides the following technical solution: An Internet of Things information update strategy method based on non-terrestrial networks, including the following steps:
[0006] The terrestrial user equipment communicates with the base station equipment on the LEO satellite through the equipped 6G transceiver. The terrestrial user equipment has different degrees of mobility and communicates with the gNodeB located on the LEO satellite for uplink communication. The user equipment transmission link processes data through an LDPC channel encoder, a bit interleaver, an M-ary quadrature amplitude modulation mapper, and an orthogonal frequency division multiplexer in sequence;
[0007] Pilot symbols are inserted into some subcarriers in each time slot in the transmitting end resource grid, and all pilot symbols are removed every other time slot. The signal propagates through a frequency-selective channel. After receiving the signal, the receiving end removes the cyclic prefix and demultiplexes;
[0008] The receiving end estimates the channel frequency response matrix on the received pilot symbols using the least squares method and interpolates the estimated value within the time slot duration;
[0009] The channel estimation value is input into a channel predictor including an encoding layer, a prediction layer, and a decoding layer for processing. The channel predictor includes a convolutional layer and a long short-term memory network layer with specific parameter configurations, and the predictor outputs a channel prediction value.
[0010] The data is equalized, demapped, and decoded using the channel estimation value or prediction value.
[0011] Furthermore, the input of the channel predictor is first normalized, and the real and imaginary parts are merged into a tensor of a specific dimension as one dimension respectively, and the output has the same structure as the input tensor.
[0012] Furthermore, the convolutional layer in the channel predictor includes multiple convolutional kernels, each convolutional kernel and stride have specific configurations, and the units of the long short-term memory network layer are configured as 32.
[0013] Furthermore, the equalization process adopts a pre-equalization method.
[0014] Furthermore, the channel predictor includes an encoding layer, a prediction layer, and a decoding layer.
[0015] Furthermore, in the channel predictor, Flatten is a custom layer that flattens the tensor on the frequency axis, and each frequency serves as a dimension of the tensor.
[0016] Furthermore, in the channel predictor, TimeMirror is a custom layer that mirrors the generated tensor on the time axis and arranges it by time to ensure that the first prediction result depends on the latest sample rather than the subsequent samples.
[0017] Furthermore, the channel prediction device applying the channel prediction method of the Internet of Things information update strategy based on a non-terrestrial network includes:
[0018] A signal processing module for processing and sending data of a ground user equipment and preprocessing the received signal at the receiving end;
[0019] A channel estimation module for estimating the CFR matrix using the least squares method and interpolating;
[0020] A channel prediction module including a channel predictor with the parameter configuration as described in claim 3, for making predictions according to the channel estimation value;
[0021] A post-data processing module for equalizing, demapping, and decoding the data using the channel estimation value or prediction value.
[0022] The present invention provides an Internet of Things information update strategy method based on a non-terrestrial network, having the following beneficial effects:
[0023] 1. The channel prediction method of the present invention can remove pilots every other OFDM time slot without affecting channel estimation. In most cases, regardless of the data modulation order, it can significantly improve the uplink peak throughput while the performance is hardly affected.
[0024] 2. Within the speed limit range of the present invention, the selected neural network performs stably between the training set and the test set user equipment and has good scalability. If it is implemented in hardware in the future and extended to multiple input-output OFDM time slots, it is expected to obtain more accurate prediction performance and has the potential to further reduce pilot signals. BRIEF DESCRIPTION OF THE DRAWINGS
[0025] Figure 1 It is a schematic diagram of the overall structure of an Internet of Things information update strategy method based on a non-terrestrial network according to the present invention;
[0026] Figure 2 It is a schematic diagram of the main structure of a channel predictor of an Internet of Things information update strategy method based on a non-terrestrial network according to the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS
[0027] The following further describes in detail the embodiments of the present invention in conjunction with the drawings and embodiments. The following embodiments are used to illustrate the present invention but cannot be used to limit the scope of the present invention.
[0028] As Figure 1 and Figure 2 shown, an Internet of Things information update strategy method based on a non-terrestrial network includes the following steps:
[0029] The terrestrial user equipment performs uplink communication with the base station equipment on the LEO satellite through the equipped 6G transceiver. The user equipment transmission link processes data through an LDPC channel encoder, a bit interleaver, an M-ary quadrature amplitude modulation mapper, and an orthogonal frequency division multiplexer in sequence;
[0030] Pilot symbols are inserted into some subcarriers in each time slot of the transmitting end resource grid, and all pilot symbols are removed every other time slot. The signal propagates through a frequency-selective channel. After the receiving end receives the signal, it removes the cyclic prefix and demultiplexes;
[0031] The receiving end uses the least squares method to estimate the channel frequency response matrix on the received pilot symbols and interpolates the estimated value within the time slot duration;
[0032] The channel estimation value is input into a channel predictor including an encoding layer, a prediction layer, and a decoding layer for processing. The channel predictor includes a convolutional layer and a long short-term memory network layer with specific parameter configurations, and the predictor outputs a channel prediction value;
[0033] The data is equalized, demapped, and decoded using the channel estimation value or prediction value. The input of the channel predictor is first normalized, and the real and imaginary parts are combined into a tensor of a specific dimension as one dimension each. The output has the same structure as the input tensor. The convolutional layer in the channel predictor contains multiple convolutional kernels, and each convolutional kernel and stride have specific configurations. The units of the long short-term memory network layer are configured as 32.
[0034] Specifically, for a user equipment, N sc -dimensional subcarriers are used for uplink transmission, and the number of subcarriers is N sym , then the dimension of the entire transmission-side resource grid X is (N sc × N sym ). In a time slot, M π QAM pilot symbols are inserted into each subcarrier, and the position index is denoted as in the resource grid of OFDM, indicating that the number of pilot symbols of OFDM in each time slot is 7. Every other time slot, all pilot symbols are removed to reduce the pilot overhead, that is, half of the time slots only contain data symbols. So there is a limitation The CP-OFDM waveform carrying the resource grid is sent by the user equipment and propagates through a frequency-selective channel, which is modeled as a channel impulse response of a TDL with N taps paths;
[0035]
[0036] where ∈ D ∼ N(0, σ D ) is the residual frequency synchronization error component, which is a Gaussian distribution N with an expectation of 0 and a standard deviation of σ D . h n (t) and τ n represent the value and delay of the nth tap (where τ1 = 0), and δ(·) represents the Dirac δ function. At the receiving end, additive white Gaussian noise interferes with the waveform. So the cyclic prefix is removed from the signal, and the received symbols are demultiplexed. Let H = FFT[h(t, τ)] represent the CFR matrix, where FFT represents the fast Fourier transform. The OFDM resource grid at the receiving end can be expressed as:
[0037] Y = H ⊙ X + W
[0038] where W represents the impact of AWGN on the OFDM grid, N0 represents the noise power spectral density, and ⊙ is the Hadamard multiplication operator of matrices. For the estimated value of the CFR matrix Apply the least squares channel estimation to the received pilot symbols. The received pilot symbols can be expressed as:
[0039]
[0040] The estimated values calculated in this way are interpolated within the slot duration to cover the time span of the resource grid. For the selected system model, the interpolation should not only cover the data symbol positions in the first slot, but also all symbol positions in the second slot where there are no pilots. Here, we use channel prediction to improve the accuracy of channel estimation.
[0041] Generally speaking, first, the obtained channel estimation values are used to equalize and demap the data symbols in the first slot; then the obtained data bits are remapped onto M-QAM symbols, and these symbols are used to perform the least squares method on the entire first slot, thus forming an overall data and pilot estimation scheme.
[0042] Demapping errors will lead to inaccurate quantization of the estimated channel matrix. For example, errors on 4-QAM data symbols may cause phase shifts in the corresponding channel estimation. Therefore, the channel predictor not only needs to predict the channel, but also identify and equalize such errors. After the entire resource grid is equalized through the estimated or predicted channel coefficients, the data symbols are demapped, and then the obtained data is decoded.
[0043] As Figure 2 shown, the channel predictor includes an encoding layer, a prediction layer, and a decoding layer. Among them, Flatten is a custom layer that flattens the tensor on the frequency axis, with each frequency as a dimension of the tensor. TimeMirror is a custom layer that mirrors the generated tensor on the time axis and arranges it in time to ensure that the first prediction result depends on the latest samples rather than later samples, thereby improving the prediction accuracy. The Relu layer is an activation function, and the BatchNorm layer is a normalization operation. The units of the LSTM are configured as 32.
[0044] The channel prediction device of the applied Internet of Things information update strategy method based on non-terrestrial networks includes: a signal processing module for processing and sending data of ground user equipment and preprocessing the received signals at the receiving end; a channel estimation module for estimating the CFR matrix using the least squares method and interpolating; a channel prediction module containing a channel predictor with parameter configuration for predicting based on the channel estimation values; a post-data processing module for equalizing, demapping, and decoding the data using the channel estimation values or prediction values.
[0045] The loss function of the network uses the mean squared error function, and the optimization function uses Adam, both of which are well-known existing technologies commonly used in machine learning.
[0046] The embodiments of the present invention are provided by way of example and description, and are not exhaustive or limit the present invention to the disclosed forms. Many modifications and variations are obvious to those of ordinary skill in the art. The embodiments are chosen and described in order to better illustrate the principles of the present invention and its practical applications, and to enable those of ordinary skill in the art to understand the present invention and design various embodiments with various modifications suitable for a particular purpose.
Claims
1. An Internet of Things information update strategy method based on a non-terrestrial network, characterized in that, It includes the following steps: The ground user equipment conducts uplink communication with the base station equipment on the LEO satellite through the equipped 6G transceiver. The transmission link of the user equipment processes data through an LDPC channel encoder, a bit interleaver, an M-ary quadrature amplitude modulation mapper, and an orthogonal frequency division multiplexer in sequence; Pilot symbols are inserted into some subcarriers in each time slot of the transmitting end resource grid, and all pilot symbols are removed every other time slot. The signal propagates through a frequency-selective channel. After the receiving end receives the signal, it removes the cyclic prefix and demultiplexes; The receiving end estimates the channel frequency response matrix on the received pilot symbols using the least squares method and interpolates the estimated value within the time slot duration; The channel estimate value is input into a channel predictor including an encoding layer, a prediction layer, and a decoding layer for processing. The channel predictor includes a convolutional layer and a long short-term memory network layer with specific parameter configurations, and the predictor outputs a channel prediction value; The data is equalized, demapped, and decoded using the channel estimate value or the prediction value.
2. The method for updating the Internet of Things information based on a non-terrestrial network according to claim 1, wherein The input of the channel predictor is first normalized, and the real and imaginary parts are combined into a tensor of a specific dimension as one dimension respectively, and the output has the same structure as the input tensor.
3. The method for updating the Internet of Things information based on the non-terrestrial network according to claim 2, characterized in that, The convolutional layer in the channel predictor includes multiple convolutional kernels, and each convolutional kernel and step size have specific configurations. The units of the long short-term memory network layer are configured as 32.
4. The method for updating the Internet of Things information based on a non-terrestrial network according to claim 3, characterized in that The equalization process adopts a pre-equalization method.
5. The method for updating the Internet of Things information based on a non-terrestrial network according to claim 4, characterized in that, The channel predictor includes an encoding layer, a prediction layer, and a decoding layer.
6. The method for updating the Internet of Things information based on a non-terrestrial network according to claim 5, wherein In the channel predictor, Flatten is a custom layer that flattens the tensor on the frequency axis, and each frequency is used as a dimension of the tensor.
7. The method for updating the Internet of Things information based on the non-terrestrial network according to claim 6, characterized in that, In the channel predictor, TimeMirror is a custom layer that mirrors the generated tensor on the time axis and arranges it in time to ensure that the first prediction result depends on the latest sample rather than the subsequent samples.
8. The channel prediction device for the Internet of Things information update strategy method based on a non-terrestrial network according to claim 7, characterized in that, It includes: A signal processing module for processing and transmitting the data of the ground user equipment and preprocessing the received signal at the receiving end; A channel estimation module for estimating the CFR matrix using the least squares method and interpolating; A channel prediction module including a channel predictor with parameter configurations for making predictions based on the channel estimate value; A post-data processing module for equalizing, demapping, and decoding the data using the channel estimate value or the prediction value.