An intelligent garden irrigation method and system based on the Internet of Things

Abstract

The present application relates to the technical field of intelligent irrigation, in particular to a garden greening intelligent irrigation method and system based on the Internet of Things. The method comprises: acquiring soil humidity, environmental temperature and node power at the irrigation node side and mapping to an information bit channel, receiving channel reliability distribution parameters and setting a code rate, setting a bit value of a frozen bit channel, in a serial cancellation list decoding process of a polar code at the control terminal side, when a soil humidity change rate exceeds a range set based on a physical environment model, applying a nonlinear attenuation penalty to an intermediate path, generating an irrigation strategy, calculating a decoding confidence, generating channel reliability distribution parameters for adjusting a code rate of a next transmission period and returning to the irrigation node. That is, the scheme of the present application can eliminate incorrect decoding results caused by strong interference, improve the reliability of data decoding in a poor channel, reduce resource waste and balance energy efficiency and reliability.

Description

Technical Field

[0001] This invention relates to the field of intelligent irrigation technology. More specifically, this invention relates to an intelligent irrigation method and system for landscaping based on the Internet of Things (IoT). Background Technology

[0002] With the development of IoT technology and modern agriculture, intelligent irrigation systems for landscaping have become a key means to achieve refined water resource management and improve the efficiency of green space maintenance. Current technologies typically deploy a large number of wireless sensor nodes to monitor key environmental parameters such as soil moisture, ambient temperature, and light intensity in real time. These nodes transmit the collected data to a central controller or cloud platform via low-power wide-area network technology, such as LoRa long-range radio. The central controller receives and analyzes this data to determine the crop's water requirements and automatically executes or adjusts irrigation plans to achieve automated irrigation management. However, in practical applications, landscaping environments are often complex and variable, with factors such as building obstructions, vegetation interference, and fluctuating electromagnetic environments. These factors can cause drastic fluctuations in wireless channel quality, reducing the reliability of wireless communication and leading to signal fading and data transmission errors, severely impacting the normal operation of the intelligent irrigation system and the accuracy of data-driven decisions.

[0003] To address the challenges posed by fluctuations in wireless channel quality, existing technologies typically employ error correction coding techniques to improve data transmission reliability. Specifically, when channel conditions are favorable, the system chooses low-intensity error correction coding, which reduces redundant data and transmits more effective information within the same bandwidth. When channel conditions deteriorate, the system selects high-intensity error correction coding to ensure data transmission accuracy. In terms of channel decoding, algorithms such as serial cancellation list decoding with polar codes are used, and their decision-making process primarily relies on the statistical characteristics of the received signal to recover the original information.

[0004] However, the above solutions still have significant drawbacks in practical applications. First, existing coding methods are fixed and lack adaptability to channel changes, leading to wasted energy and bandwidth or insufficient protection. Second, soil moisture changes are typically slow and continuous; drastic jumps within a short period are illogical. Traditional decoding methods, when faced with strong noise interference, rely solely on the statistical characteristics of the received signal to output a valid result closest to the received signal in the codeword space. While this result may be statistically valid, from a physical perspective, a drastic and unreasonable jump in soil moisture value within a short period represents a completely erroneous physical state. Consequently, this statistically based decoding may lead to seemingly correct but erroneous results that contradict actual physical laws. Furthermore, existing solutions generally lack a closed-loop feedback mechanism. This means the system cannot adjust the transmitting end's coding strategy in real time based on the confidence level of the actual decoding effect. This open-loop or unidirectional detection method makes it difficult for the system to achieve an optimal balance between communication energy efficiency and reliability, resulting in persistent resource waste or insufficient reliability. Summary of the Invention

[0005] The purpose of this invention is to propose an intelligent irrigation method and system for landscaping based on the Internet of Things, in order to solve the problems of low decoding reliability, resource waste, and inability to achieve an optimal balance between energy efficiency and reliability in the prior art; to this end, this invention provides solutions in the following two aspects.

[0006] In a first aspect, the present invention provides an intelligent irrigation method for landscaping based on the Internet of Things, comprising:

[0007] At the irrigation node, soil moisture, ambient temperature, and node power are acquired; the channel reliability allocation parameters from the previous transmission cycle are received, and the code rate for the next transmission cycle is set; in the polarization sub-channel, the information bit channel and the frozen bit channel are acquired; the binary representation of the water valve switch state and irrigation duration at the previous moment of the irrigation node is XORed with a bit level, and the XOR result is used as a seed input to the hash function to generate a pseudo-random sequence, which is then used to set the bit value of the frozen bit channel; soil moisture, ambient temperature, and node power are mapped to the information bit channel, encoded, and then transmitted via the LoRa module; at the control terminal, the serial cancellation list decoding of the polarization code is received and performed. During the decoding process, when... When the rate of change of soil moisture in a portion of the sensor data sequence corresponding to any intermediate decoding path exceeds the range set based on the physical environment model, a nonlinear attenuation penalty is applied to the log-likelihood ratio path metric of the intermediate path according to the degree of exceedance; an irrigation strategy is generated based on the sensor data output from the decoding; the difference between the cumulative log-likelihood ratio path metric of the winning decoding path and the average of the cumulative log-likelihood ratio path metric of all other surviving paths in the decoding candidate list is calculated as the decoding confidence; the decoding confidence is substituted into a preset piecewise function to generate channel reliability allocation parameters for adjusting the code rate of the next transmission cycle, and the channel reliability allocation parameters and the irrigation strategy are transmitted back to the irrigation node.

[0008] Preferably, the step of performing a bit-level XOR operation on the water valve open / closed state at the previous moment of the irrigation node and the binary representation of the irrigation duration, using the XOR result as a seed to input into a hash function to generate a pseudo-random sequence, and using the pseudo-random sequence to set the bit values ​​of the frozen bit channels includes: recording the water valve open state as 1 and the closed state as 0; placing the state value in the least significant bit of an 8-bit binary number, padding the remaining high bits with 0 to form a first 8-bit binary number; converting the irrigation duration in minutes into a second 8-bit binary number; performing a bit-level XOR operation on the first 8-bit binary number and the second 8-bit binary number; using the 8-bit XOR result as a seed to input into a hash function to generate a pseudo-random binary sequence with a length consistent with the number of frozen bit channels, and sequentially assigning the bit values ​​of the sequence to each frozen bit channel.

[0009] Preferably, the construction of the physical environment model includes: collecting the average daily temperature, average light intensity, and total rainfall over the past 30 consecutive days as input features, and collecting the corresponding daily soil moisture change rate as a label; the input features and labels form a training sequence; a long short-term memory network is used to train the training sequence to obtain a physical environment model that can predict the upper and lower limits of normal soil moisture change in the next hour based on current meteorological data; the upper and lower limits constitute the range.

[0010] Preferably, the step of applying a nonlinear attenuation penalty to the log-likelihood ratio path metric of the intermediate path based on the degree of excess includes: in any bit decoding step of the serial cancellation list decoding, when the soil moisture change rate calculated in the partial sensor data sequence corresponding to the decoded intermediate path exceeds the upper limit of the range set by the physical environment model, updating the log-likelihood ratio path metric of the intermediate path; the update formula is: In the formula, This is the new log-likelihood ratio path metric. The log-likelihood ratio of the intermediate path is the path metric. To preset a positive penalty coefficient, The rate of change of soil moisture. The upper limit of the range set for the physical environment model. It is an exponential function with the natural constant e as its base.

[0011] Preferably, mapping soil moisture, ambient temperature, and node power to the information bit channel includes: mapping the range of... to Soil moisture is represented as a 10-bit unsigned integer; the range is... to The ambient temperature is represented as an 8-bit unsigned integer; the range is... to The node power is represented as a 7-bit unsigned integer; the 10-bit, 8-bit, and 7-bit data are concatenated in sequence to form a 25-bit data frame; the data frame is then filled into the information bit channel.

[0012] Preferably, the difference between the cumulative log-likelihood ratio path metric of the winning decoding path and the average of the cumulative log-likelihood ratio path metrics of all other surviving paths in the decoding candidate list is used as the decoding confidence score. This includes: after decoding, the path corresponding to the maximum cumulative log-likelihood ratio path metric in the decoding candidate list is the winning path, and the remaining paths in the decoding candidate list other than the winning path are surviving paths; the decoding confidence score is calculated using the following formula: In the formula, For decoding confidence, The cumulative log-likelihood ratio of the winning path is the path metric. The cumulative log-likelihood of the surviving paths is the average of the path metrics.

[0013] Preferably, the step of substituting the decoding confidence into a preset segmentation function to generate channel reliability allocation parameters for adjusting the code rate of the next transmission cycle includes: In the formula, The bit rate for the next transmission cycle. For the current bitrate, Decoding confidence level; code rate for the generated next transmission cycle. As a parameter for channel reliability allocation, while ensuring The value is in Within the range.

[0014] Preferably, the generated code rate for the next transmission cycle As a parameter for channel reliability allocation, while ensuring The value is in Within the range, including: when the code rate of the next transmission cycle... In the range When inside, then As a channel reliability allocation parameter; when the code rate of the next transmission cycle Beyond the range When the upper limit is 0.75, then 0.75 is used as the channel reliability allocation parameter; when the code rate of the next transmission cycle... Below the range When the lower limit is 0.25, then 0.25 is used as the channel reliability allocation parameter.

[0015] Preferably, the hash function uses the SHA-256 hash algorithm.

[0016] In the second aspect, an IoT-based intelligent irrigation system for landscaping includes:

[0017] The system includes a processor and a memory, the memory storing computer program instructions for IoT-based intelligent irrigation of landscaping, which, when executed by the processor, implement the aforementioned IoT-based intelligent irrigation method for landscaping.

[0018] The beneficial effects of this invention are as follows: First, by using a physical environment model to verify the rationality of intermediate decoding paths during the decoding process, when the calculated soil moisture change rate in a portion of the decoded sensor data sequence exceeds a reasonable range, a nonlinear attenuation penalty is applied to that intermediate path, thereby eliminating invalid decoding results caused by strong interference and improving the reliability of data decoding under harsh channel conditions. Second, by calculating the decoding confidence level, which represents the reliability of the decoding result, and combining it with a preset piecewise function to generate channel reliability allocation parameters to guide the code rate of the next transmission cycle, this invention forms a closed-loop feedback mechanism. This allows the communication code rate to be dynamically adjusted based on the actual decoding effect of the previous cycle, reducing resource waste and achieving an optimal balance between energy efficiency and reliability. Furthermore, by inputting the physical state of the irrigation system into the encoding process and specifically setting the frozen bit channel, the correlation of the encoding is enhanced. Attached Figure Description

[0019] Figure 1 The flowchart illustrating the steps of an IoT-based intelligent irrigation method for landscaping in this embodiment is shown in the illustration.

[0020] Figure 2 The diagram illustrates the structure of an IoT-based intelligent irrigation system for landscaping in this embodiment. Detailed Implementation

[0021] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.

[0022] like Figure 1 As shown, this embodiment of an IoT-based intelligent irrigation method for landscaping includes steps S1 to S4:

[0023] Step S1: On the irrigation node side, obtain soil moisture, ambient temperature and node power; receive the channel reliability allocation parameters returned from the previous transmission cycle and set the code rate for the next transmission cycle; in the polarization sub-channel, obtain the information bit channel and the frozen bit channel.

[0024] Specifically, at the irrigation node, soil moisture and ambient temperature are acquired through sensors; the real-time battery voltage is measured by the voltage detection circuit on the node, and the remaining power is calculated as the node's power. The LoRa module of the irrigation node receives a data packet containing channel reliability allocation parameters sent by the control terminal, including the channel reliability allocation parameters returned from the previous transmission cycle, i.e., the code rate. Based on the channel reliability allocation parameters returned from the previous transmission cycle, the code rate for the next transmission cycle can be set.

[0025] The preset polar code mother code length is In this embodiment The irrigation nodes are based on the preset polar code mother code length. Calculate the number of information bit channels , equal Multiply by the code rate; for example, if the code rate is 0.5, then the number of information bits in the channel. The value is 128. The Parshall parameter sorting table for the current signal-to-noise ratio estimate is consulted in the control terminal internally at the irrigation node. The 128 sub-channel indices with the smallest Parshall parameters are selected as the information bit channel set, and the remaining 128 sub-channel indices are selected as the frozen bit channel set. Thus, the information bit channel and frozen bit channel are obtained from the polarized sub-channels.

[0026] Step S2: Perform a bit-level XOR operation on the binary representation of the water valve switch status of the irrigation node at the previous moment and the irrigation duration. Use the XOR result as a seed to input the hash function to generate a pseudo-random sequence. Use the pseudo-random sequence to set the bit value of the frozen bit channel. Map the soil moisture, ambient temperature, and node power to the information bit channel. After encoding, transmit the data through the LoRa module.

[0027] In one embodiment, the step of performing a bit-level XOR operation on the water valve switch state of the irrigation node at the previous moment and the binary representation of the irrigation duration, using the XOR result as a seed input to a hash function to generate a pseudo-random sequence, and using the pseudo-random sequence to set the bit value of the frozen bit channel includes:

[0028] The water valve is marked as open (1) and closed (0). The state value is placed in the least significant bit of an 8-bit binary number, and the remaining high bits are padded with 0 to form the first 8-bit binary number. The irrigation duration in minutes is converted into a second 8-bit binary number. The first 8-bit binary number and the second 8-bit binary number are XORed. The 8-bit XOR result is used as a seed and input into a hash function to generate a pseudo-random binary sequence with a length consistent with the number of frozen bit channels. The bit values ​​of the sequence are then assigned to each frozen bit channel in sequence.

[0029] The hash function uses the SHA-256 hash algorithm.

[0030] Specifically, the irrigation node reads the previous water valve status from non-volatile memory, recording the open state as binary 1 and the closed state as binary 0; it also reads the last irrigation duration. For example, if the irrigation node's water valve was open in the previous moment and the irrigation duration was 45 minutes, the open state is recorded as 1, and this value is placed in the least significant bit of an 8-bit binary number, forming the first 8-bit binary number 00000001; the irrigation duration of 45 minutes is converted into the second 8-bit binary number 00101101. A bitwise XOR operation is performed on the two 8-bit binary numbers, i.e., 00000001 XORed with 00101101, resulting in 00101100. This 8-bit XOR result, 00101100, is used as the input seed and fed into the SHA-256 hash algorithm, which generates a 256-bit pseudo-random sequence. For the Polar code scheme determined in the above example, which requires setting 128 frozen bit channels, the first 128 bits of the 256-bit sequence are extracted as a pseudo-random binary sequence. The 128 bit values ​​of this pseudo-random binary sequence are then assigned sequentially to each frozen bit channel. For example, if the first few bits of the 256-bit sequence are 11010011, the first frozen bit channel is set to 1, the second to 1, the third to 0, the fourth to 1, and so on, until all 128 frozen bit channels have been assigned values.

[0031] In one embodiment, mapping soil moisture, ambient temperature, and node power to the information bit channel includes:

[0032] The range is to Soil moisture is represented as a 10-bit unsigned integer; the range is... to The ambient temperature is represented as an 8-bit unsigned integer; the range is... to The node power is represented as a 7-bit unsigned integer; the 10-bit, 8-bit, and 7-bit data are concatenated in sequence to form a 25-bit data frame; the data frame is then filled into the information bit channel.

[0033] Specifically, for the representation of soil moisture, the 10-bit range is 0 to 1023, which maps the soil moisture range of 0% to 100% to the integer range of 0 to 1023; for the representation of ambient temperature, the 8-bit range is 0 to 255, which maps the ambient temperature range of -20℃ to 60℃ to the integer range of 0 to 255; for the representation of node power, the 7-bit range is 0 to 127, which maps the node power range of 0% to 100% to the integer range of 0 to 127. For example, the sensor measures soil moisture of 52.8% and ambient temperature of 26.5℃, calculating the node's charge to be 90%. The soil moisture of 52.8% is converted to a decimal between 0 and 1, 0.528. Multiplying 0.528 by the maximum integer value of 1023 yields approximately 540. Converting 540 to binary gives a 10-bit binary number, represented as 1000011100. The total range of ambient temperature is calculated to be 80℃. The difference between the current ambient temperature and the minimum value of the range yields a relative position of 46.5℃. The ratio of this relative position to the total range range gives a position ratio of 0. 58. Calculate the product of the position ratio and the total span of the integer range, 255, to obtain an integer of approximately 148. Convert 148 to binary to obtain 8 bits, represented as 10010100. This allows for precise mapping between two ranges, such as ambient temperature ranges that do not start from 0, by eliminating the offset of the starting point through translation and by scaling to match the ratio of the old and new ranges. Convert 90% of the node's power to a decimal between 0 and 1, 0.9. Multiply 0.9 by the maximum value of the integer range, 127, to obtain approximately 114. Convert 114 to binary to obtain 7 bits, represented as 1110010. Concatenate the three binary data segments in the order of soil moisture, ambient temperature, and node power, i.e., 1000011100, 10010100, and 1110010 in sequence, to form a 25-bit data frame, 1000011100100101001110010. This data frame can also be called a sensor data sequence. The data frame is placed in the channel position reserved for information bits in the Polar code encoder, that is, the first 25 positions of the 128 information bit channel. The remaining information bits can be used to fill the cyclic redundancy check code or set to all zeros, and then prepared for encoding and transmission.

[0034] Step S3: On the control terminal side, the serial cancellation list of polar codes is received and decoded. During the decoding process, when the soil moisture change rate of a portion of the sensor data sequence corresponding to any intermediate decoding path exceeds the range set based on the physical environment model, a nonlinear attenuation penalty is applied to the log-likelihood ratio path metric of the intermediate path according to the degree of exceedance. An irrigation strategy is generated based on the sensor data output by the decoding.

[0035] Specifically, the server controlling the terminal executes a decoding candidate list of size [size missing]. When decoding a serial offset list, for example... =8, and 8 candidate decoding paths will be maintained in parallel. The physical environment model is based on the joint modeling of historical irrigation data and meteorological data. When decoding reaches the boundary position of sensor data, taking the first 10 bits of soil moisture corresponding to the decoding as an example, the server will check the soil moisture temporarily decoded by each candidate decoding path. If the soil moisture decoded by the 3rd path is 80%, while the soil moisture of the node in the database at the previous moment is 40%, the physical environment model determines the range of soil moisture change based on historical data and recent meteorological information without rainfall. If the current soil moisture change rate far exceeds the range set by the physical environment model, the server calculates the degree of excess and multiplies the degree of excess by a negative penalty coefficient to obtain a large positive penalty value. The obtained positive penalty value is added to the cumulative log-likelihood ratio path metric of the 3rd path.

[0036] After decoding, the path with the optimal cumulative log-likelihood ratio is selected as the decoding result. Sensor data is then extracted from this path to obtain soil moisture, ambient temperature, and node power. The irrigation decision system compares the extracted soil moisture with a preset irrigation threshold for the area. In this embodiment, the irrigation threshold is taken as an empirical value of 35%. When the extracted soil moisture is higher than the irrigation threshold, an instruction to not perform irrigation is generated, and the next data collection time is set to 2 hours later.

[0037] In an optional embodiment, the construction of the physical environment model includes:

[0038] The average daily temperature, average light intensity, and total rainfall over the past 30 consecutive days are collected as input features, and the corresponding daily soil moisture change rate is collected as a label. The input features and labels form a training sequence. A physical environment model is obtained by training the training sequence using a long short-term memory network, which can predict the upper and lower limits of normal soil moisture change in the next hour based on current meteorological data. The upper and lower limits constitute the range.

[0039] Specifically, a 30-day dataset is recorded, with each day's data items including average temperature, average light intensity, total rainfall, and soil moisture change rate. For example, day 1's data includes an average temperature of 22 degrees Celsius, average light intensity of 750 watts per square meter, total rainfall of 0 millimeters, and an average hourly decrease in soil moisture of 0.6 percentage points. The daily average temperature, average light intensity, and total rainfall serve as input features, while the daily soil moisture change rate serves as a label. These daily input features and labels form a set of training data. This recording process continues for 30 days, forming a training sequence containing 30 sets of input features and corresponding labels. This training sequence is then input into a pre-built Long Short-Term Memory (LSTM) network model for training. The LSTM model adjusts its internal parameters by learning the temporal correlation between meteorological data (average temperature, average light intensity, and total rainfall) and labels (soil moisture change rate) over the 30 days. After training, a physical environment model based on the joint modeling of historical irrigation data and meteorological data is obtained.

[0040] When forecasting is needed, the current meteorological data is input, and the physical environment model, based on its learned knowledge, predicts the most likely center of change in soil moisture over the next hour, while also providing a reasonable fluctuation range. For example, the predicted rate of change in soil moisture over the next hour is a decrease of 1.1 percentage points per hour; the physical environment model sets the range to a lower limit of 1.5 percentage points per hour and an upper limit of 0.7 percentage points per hour.

[0041] In an optional embodiment, applying a non-linear decay penalty to the log-likelihood ratio path metric of the intermediate path based on the degree of excess includes:

[0042] In any bit decoding step of the serial cancellation list decoding, when the soil moisture change rate calculated in the partial sensor data sequence corresponding to the intermediate path exceeds the upper limit of the range set by the physical environment model, the log-likelihood ratio path metric of the intermediate path is updated; the update formula is: ;

[0043] In the formula, This is the new log-likelihood ratio path metric. The log-likelihood ratio of the intermediate path is the path metric. To preset a positive penalty coefficient, The rate of change of soil moisture. The upper limit of the range set for the physical environment model. It is an exponential function with the natural constant e as its base.

[0044] Specifically, when the decoder reaches the... During the bit-by-bit decoding step, multiple possible intermediate decoding paths are maintained. The log-likelihood ratio path metric for an intermediate path is only updated when the soil moisture change rate calculated from a portion of the sensor data sequence decoded by that intermediate path exceeds the upper limit set by the physical environment model. For example, for one of the intermediate paths... Based on the sensor data parsed from the already translated partial bit sequence, the intermediate path is obtained. The soil moisture was calculated to have increased by 2.5 percentage points in the past hour, which is the rate of change in soil moisture. =2.5; Based on the current meteorological data, the physical environment model gives the upper limit of the range of soil moisture variation. The rate of increase was 0.8 percentage points per hour. At this point, the soil moisture change rate of 2.5 exceeded the upper limit of 0.8 set by the physical environment model, indicating an intermediate path. The scenario represented is unlikely to occur, requiring consideration of the intermediate path. Log-likelihood ratio to path metric Apply a penalty; the preset positive penalty coefficient Take an empirical value of 0.5; use the intermediate path Current Taking a value of 20.0 as an example, calculate the excess amount. The value is 1.7, and the new log-likelihood ratio path metric is calculated using the above formula. Approximately 8.54, intermediate path The log-likelihood ratio path metric was significantly reduced from 20.0.

[0045] Step S4: Calculate the difference between the cumulative log-likelihood ratio path metric of the winning decoding path and the average of the cumulative log-likelihood ratio path metric of all other surviving paths in the decoding candidate list, and use it as the decoding confidence. Substitute the decoding confidence into a preset piecewise function to generate channel reliability allocation parameters for adjusting the code rate of the next transmission cycle, and send the channel reliability allocation parameters and irrigation strategy back to the irrigation node.

[0046] In one embodiment, the difference between the cumulative log-likelihood ratio path metric of the winning decoding path and the average of the cumulative log-likelihood ratio path metrics of all other surviving paths in the candidate decoding list, as the decoding confidence, includes:

[0047] After decoding is completed, the path corresponding to the maximum cumulative log-likelihood ratio path metric in the decoding candidate list is the winning path, and the remaining paths in the decoding candidate list other than the winning path are the surviving paths.

[0048] The formula for calculating decoding confidence is as follows: In the formula, For decoding confidence, The cumulative log-likelihood ratio of the winning path is the path metric. The cumulative log-likelihood of the surviving paths is the average of the path metrics.

[0049] Specifically, the size of the decoding candidate list The value is 8. After the serial cancellation list decoder completes decoding a received codeword, it outputs a list containing 8 candidate paths. Each path has a cumulative log-likelihood ratio (CPR) path metric. The path with the highest CPR path metric in the candidate list is the winning path, and the remaining paths are surviving paths. For example, if the CPR path metric values ​​of the 8 paths are 30.5, 21.2, 20.8, 20.1, 19.5, 19.0, 18.8, and 18.5, and the highest CPR path metric value is 30.5, then the path corresponding to 30.5 is the winning path. The cumulative log-likelihood ratio is 30.5; calculate the average of the cumulative log-likelihood ratio path metrics for the remaining 7 surviving paths excluding the winning path, i.e. The value is 19.7; the decoding confidence level is obtained by subtracting the two values. The obtained value represents the confidence level of this decoding. The larger the value, the more prominent the advantage of the winning path in terms of reliability compared to other competing paths, and the more reliable the decoding result.

[0050] In one embodiment, substituting the decoding confidence into a preset segmentation function to generate channel reliability allocation parameters for adjusting the code rate of the next transmission cycle includes:

[0051] ;

[0052] In the formula, The bit rate for the next transmission cycle. For the current bitrate, For decoding confidence;

[0053] The bit rate of the next transmission cycle generated As a parameter for channel reliability allocation, while ensuring The value is in Within the range.

[0054] In one embodiment, the generated code rate for the next transmission cycle As a parameter for channel reliability allocation, while ensuring The value is in Within the interval, including:

[0055] When the code rate of the next transmission cycle In the range When inside, then As a channel reliability allocation parameter; when the code rate of the next transmission cycle Beyond the range When the upper limit is 0.75, then 0.75 is used as the channel reliability allocation parameter; when the code rate of the next transmission cycle... Below the range When the lower limit is 0.25, then 0.25 is used as the channel reliability allocation parameter.

[0056] Specifically, a pre-defined segmentation function is used based on experience to generate channel reliability allocation parameters for adjusting the code rate in the next transmission cycle, based on the decoding confidence level. For example, the currently used transmission code rate... The confidence level is 0.6, calculated after one decoding operation. The decoding confidence level is 15.4. Substituting the piecewise function into the judgment, if the first condition is met, it indicates that the channel quality is very good and the decoding is very reliable; according to... The generated code rate for the next transmission cycle is 0.65, increasing the code rate for the next transmission cycle; the generated code rate of 0.65 is within the range If the code rate is within the range, then 0.65 is adopted as the channel reliability allocation parameter for the next cycle. The decoding confidence level is 0.6, calculated after decoding. A value of 2.5 satisfies the third condition, indicating poor channel quality and unreliable decoding results; according to The code rate for the next transmission cycle is 0.55, so the code rate for the next transmission cycle is reduced. When the code rate... The decoding confidence score is 0.75, calculated after decoding. If the value is greater than 12.0, the generated bitrate will be 0.8, which exceeds the range. If the upper limit is 0.75, then 0.75 will be adopted as the channel reliability allocation parameter for the next cycle. When the generated code rate is lower than the interval... If the lower limit is 0.25, then 0.25 will be adopted as the channel reliability allocation parameter for the next cycle. This will determine the channel reliability allocation parameter for the next cycle. Used to guide the construction of the next Polar code, determining the ratio of information bit channel to frozen bit channel.

[0057] This invention also provides an intelligent irrigation system for landscaping based on the Internet of Things (IoT). For example... Figure 2As shown, the system includes a processor and a memory. The memory stores computer program instructions for IoT-based intelligent irrigation of landscaping. When the computer program instructions are executed by the processor, an IoT-based intelligent irrigation method for landscaping is implemented according to the present invention.

[0058] The system also includes other components well known to those skilled in the art, such as communication buses and communication interfaces, the settings and functions of which are known in the art and therefore will not be described in detail here.

[0059] In this invention, the aforementioned memory can be any tangible medium containing or storing a program that can be used or combined with an instruction execution system, apparatus, or device. For example, a computer-readable storage medium can be any suitable magnetic or magneto-optical storage medium, such as Resistive Random Access Memory (RRAM), Dynamic Random Access Memory (DRAM), Static Random Access Memory (SRAM), Enhanced Dynamic Random Access Memory (EDRAM), High-Bandwidth Memory (HBM), Hybrid Memory Cube (HMC), etc., or any other medium that can be used to store desired information and can be accessed by an application, module, or both. Any such computer storage medium can be part of a device or accessible to or connected to a device. Any application or module described in this invention can be implemented by computer-readable / executable instructions stored or otherwise maintained on such a computer-readable medium.

[0060] In the description of this specification, "multiple" means at least two, such as two, three or more, etc., unless otherwise expressly and specifically defined.

[0061] While various embodiments of the invention have been shown and described in this specification, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Many modifications, alterations, and alternatives will occur to those skilled in the art without departing from the spirit and essence of the invention.

Claims

1. An Internet of Things-based intelligent landscaping irrigation method, characterized in that, The method comprises the following steps: On the irrigation node side, the soil humidity, the environmental temperature and the node power are acquired; The channel reliability distribution parameter returned in the last transmission period is received, and the code rate of the next transmission period is set; in the polarization subchannel, the information bit channel and the frozen bit channel are acquired; The binary representation of the water valve switch state of the irrigation node at the previous moment and the irrigation duration is subjected to bit-level XOR, and the XOR result is used as a seed to input a hash function to generate a pseudo-random sequence, and the pseudo-random sequence is used to set the bit value of the frozen bit channel; the soil humidity, the environmental temperature and the node power are mapped to the information bit channel, and after encoding, the LoRa module is used for sending; On the control terminal side, the serial cancellation list decoding of the polarization code is received, and in the decoding process, when the soil humidity change rate of any intermediate path corresponding to the partial sensor data sequence exceeds the range set based on the physical environment model, a nonlinear attenuation penalty is applied to the log-likelihood ratio path metric value of the intermediate path according to the exceeding degree; An irrigation strategy is generated based on the sensor data output by the decoding; The difference between the accumulated log-likelihood ratio path metric value of the decoding winning path and the average value of the accumulated log-likelihood ratio path metric values of all the remaining surviving paths in the decoding candidate list is calculated as the decoding confidence; The decoding confidence is substituted into a preset segmented function to generate a channel reliability distribution parameter for adjusting the code rate of the next transmission period, and the channel reliability distribution parameter and the irrigation strategy are returned to the irrigation node. 2.The intelligent garden irrigation method based on the Internet of Things according to claim 1, wherein, The bit-level XOR of the binary representation of the water valve switch state of the irrigation node at the previous moment and the irrigation duration, and the XOR result as a seed input to a hash function to generate a pseudo-random sequence, and the pseudo-random sequence is used to set the bit value of the frozen bit channel, comprising: The open state of the water valve is recorded as 1, and the closed state is recorded as 0; the state value is placed in the least significant bit of an 8-bit binary number, and the remaining high bits are filled with 0 to form a first 8-bit binary number; the irrigation duration in minutes is converted into a second 8-bit binary number; the first 8-bit binary number and the second 8-bit binary number are subjected to bit-level XOR; the 8-bit XOR result is used as a seed to input into a hash function to generate a pseudo-random binary sequence with a length consistent with the number of frozen bit channels, and the bit values of the sequence are sequentially assigned to each frozen bit channel. 3.The intelligent garden irrigation method based on the Internet of Things according to claim 1, wherein, The construction of the physical environment model comprises: The average air temperature, the average light intensity and the total rainfall in the past 30 consecutive days are collected as input features, and the soil humidity change rate corresponding to each day is collected as a label; the input features and the label form a training sequence; a long short-term memory network is used to train the training sequence to obtain a physical environment model, which can predict the upper limit and the lower limit of the normal change of the soil humidity within 1 hour in the future according to the current meteorological data; the upper limit and the lower limit constitute the range. 4.The intelligent garden irrigation method based on the Internet of Things according to claim 3, wherein, The nonlinear attenuation penalty applied to the log-likelihood ratio path metric value of the intermediate path according to the exceeding degree, comprising: In any bit-decoding step of the serial cancellation list decoding, when the calculated soil moisture rate of change in the portion of the sensor data sequence corresponding to the decoding intermediate path exceeds the upper limit of the range set by the physical environment model, the log-likelihood ratio path metric value of the intermediate path is updated; the update formula is: ; wherein is a new log-likelihood ratio path metric value, is a log-likelihood ratio path metric value of an intermediate path, is a preset positive penalty coefficient, is a soil humidity change rate, is an upper limit of a range set for a physical environment model, is an exponential function with a natural constant e as a base. 5.The intelligent garden irrigation method based on the Internet of Things according to claim 1, wherein, The mapping of the soil humidity, the environmental temperature and the node power to the information bit channel, comprising: soil humidity ranging from to is represented as a 10-bit unsigned integer; ambient temperature ranging from to is represented as an 8-bit unsigned integer; node power ranging from to is represented as a 7-bit unsigned integer; the 10-bit, 8-bit and 7-bit data are concatenated in order to form a 25-bit data frame; the data frame is filled into the information bit channel. 6.The intelligent garden irrigation method based on the Internet of Things according to claim 1, wherein, The difference between the accumulated log-likelihood ratio path metric value of the decoding winning path and the average value of the accumulated log-likelihood ratio path metric values of all the remaining surviving paths in the decoding candidate list is taken as the decoding confidence, comprising: After decoding is completed, the path corresponding to the maximum accumulated log-likelihood ratio path metric value in the decoding candidate list is the winning path, and the remaining paths in the decoding candidate list except the winning path are the surviving paths; The decoding confidence is calculated by the following formula wherein is the decoding confidence, is the accumulated log-likelihood ratio path metric value of the winning path, is the average of the accumulated log-likelihood ratio path metric values of the surviving paths. 7.The intelligent garden irrigation method based on the Internet of Things according to claim 1, wherein, The decoding confidence is substituted into a preset segmentation function to generate a channel reliability allocation parameter for adjusting the code rate of the next transmission period, comprising: ； wherein the code rate for the next transmission period, the current code rate, the decoding confidence; Code rate of the next transmission period generated as a channel reliability allocation parameter, while ensuring that the value of the channel reliability allocation parameter is in the interval ​ 8.The intelligent garden irrigation method based on the Internet of Things according to claim 7, wherein, the generated code rate of the next transmission period as a channel reliability allocation parameter, while ensuring the value of the generated code rate is within the interval of the generated code rate is within the interval When the code rate of the next transmission cycle In the range When inside, then As a channel reliability allocation parameter; when the code rate of the next transmission cycle Beyond the range When the upper limit is 0.75, then 0.75 is used as the channel reliability allocation parameter; when the code rate of the next transmission cycle... Below the range When the lower limit is 0.25, then 0.25 is used as the channel reliability allocation parameter. 9.The intelligent garden irrigation method based on the Internet of Things according to claim 1, wherein, The hash function adopts a SHA-256 hash algorithm.

10. An intelligent garden irrigation system based on the Internet of Things, characterized in that, Comprise: A processor and a memory, the memory stores computer program instructions for the intelligent irrigation of landscaping based on the Internet of Things, and when the computer program instructions are executed by the processor, a method for the intelligent irrigation of landscaping based on the Internet of Things according to any one of claims 1-9 is implemented.

Images