CN114928415B - Multi-parameter networking method based on edge computing gateway link quality evaluation - Google Patents

Abstract

The invention discloses a multi-parameter networking method based on edge computing gateway link quality assessment, including: collecting sample data and preprocessing the sample data; building a link quality assessment model based on the AWOA-BP neural network, which will pass through The preprocessed sample data is input into the link quality assessment model for training, and the training is completed until the output error is reduced to the expected level or the set number of learning iterations is reached, and the trained link quality assessment model is obtained; the carrier detection multi-channel A LoRa network system based on edge computing is built by combining access CSMA and time division multiple access TDMA access; based on the trained link quality assessment model, real-time link quality assessment is performed on the LoRa network to identify changes in link quality; The link quality change adaptively updates the radio frequency parameters of the LoRa network, improving the link quality and resource utilization of the LoRa network.

Description

Multi-parameter networking method based on edge computing gateway link quality assessment

technical field

The invention belongs to the field of wireless communication, in particular to a multi-parameter networking method based on edge computing gateway link quality evaluation.

Background technique

LoRa (Long Range Radio, long-distance radio) is a kind of LPWAN (Low-Power Wide-AreaNetwork, low-power wide-area network) technology, which is mainly used in multi-node, long-distance, low-power application sites. A technology that has received much attention in the field of global communication research. LoRa is a typical emerging IoT technology that supports large-scale low-speed transmission and long-distance low power consumption. Compared with the same type of wireless communication technology, LoRa has obvious advantages in coverage, endurance and cost of use. In communication, channel resources, time slot resources, energy resources, etc. are limited, and it is an urgent problem to realize optimal utilization of resources and improve network capacity and throughput. The existing LoRa networking solutions are all designed on the basis of the LoRaWAN protocol, which lacks consideration of time application scenarios. Although reference values of radio frequency parameters and link characteristics are provided in the prior art, in practical applications, the impact of wireless link fluctuations on link characteristics is unpredictable, and link characteristics are crucial to communication performance. Influence the design and implementation of higher layer protocols. In view of the various interferences existing in the LoRa link, it has become a difficult problem to evaluate the link quality through edge computing at the gateway, identify link changes, and realize adaptive update of radio frequency parameters, and a method is urgently needed to solve it .

Contents of the invention

The purpose of the present invention is to provide a multi-parameter networking method based on edge computing gateway link quality assessment, so as to solve the above-mentioned problems in the prior art.

To achieve the above purpose, the present invention provides a multi-parameter networking method based on edge computing gateway link quality assessment, including:

Collect sample data and preprocess the sample data;

Build a link quality assessment model based on the AWOA-BP neural network, and input the preprocessed sample data into the link quality assessment model for training until the output error is reduced to the desired level or reaches the set number of learning iterations The training is completed in time, and the trained link quality evaluation model is obtained;

The LoRa network system based on edge computing is constructed by combining carrier sense multiple access CSMA and time division multiple access TDMA access;

Carry out real-time link quality assessment to the LoRa network based on the trained link quality assessment model, and identify changes in link quality;

Adaptively updating the radio frequency parameters of the LoRa network based on the change of the link quality.

Optionally, the sample data includes: received signal strength RSSI, signal-to-noise ratio SNR, and packet loss rate PLR, where the PLR is calculated based on the RSSI and the SNR.

Optionally, the process of preprocessing the sample data includes: using a Kalman filter method to filter the RSSI and the SNR.

Optionally, the process of training the link quality assessment model includes:

Adaptively adjust the size of the weight based on the current whale population distribution;

The search strategy is adaptively adjusted based on the fitness probability threshold of the current whale individual.

Optionally, in the process of building a LoRa network system based on edge computing by combining CSMA and TDMA access:

Before information transmission, the CSMA reduces communication conflicts through channel monitoring and random backoff, wherein the access mode of the CSMA is access through contention;

Time slots are divided by the TDMA, channel resources of multiple nodes are allocated based on the divided time slots, and nodes are allowed to occupy channels in their respective time slots for communication within the sampling period of the nodes.

Optionally, the process of constructing a LoRa network system based on edge computing by combining CSMA and TDMA access also includes: the LoRa network system adopts a star structure, and the star structure includes several gateways and several a node, the gateway includes a control sub-gateway, a communication sub-gateway and a gateway MCU;

Perform node allocation through the control sub-gateway, request data frame reception and response, receive data packets through the communication sub-gateway, and send them to the gateway MCU, store and forward the data packets through the gateway MCU and edge computing, the data packets include allocation request packets, synchronization request packets and periodic sampling data packets.

Optionally, during the process of adaptively updating the radio frequency parameters of the LoRa network system based on changes in the link quality, the gateway MCU performs edge calculations on the changes in the link quality to obtain radio frequency parameter updates instruction;

The radio frequency parameter update instruction is sent to the control sub-gateway, the communication sub-gateway and the node, and the corresponding radio frequency parameters are updated, and the radio frequency parameters include: frequency and spreading factor, transmission power and bandwidth.

Technical effect of the present invention is:

(1) Aiming at the mapping relationship between RSSI, SNR and PLR, an improved whale algorithm is proposed to optimize the BP neural network to build a link quality assessment model. Through adaptive weights and search strategies, the global search ability and convergence rate of the whale algorithm are improved, and the optimized weights and thresholds are assigned to the BP neural network for training to improve the accuracy of model predictions.

(2) For various interferences in the LoRa link, evaluate the link quality through edge computing at the gateway, identify link changes, realize adaptive update of radio frequency parameters, and improve the link quality and resource utilization of the LoRa network .

(3) A multi-module LoRa gateway system based on edge computing is designed to realize the management of terminal equipment and edge computing of link quality, which is not restricted by the network status and improves the stability of the LoRa network. At the same time, it is based on the IoT cloud The platform sets up online monitoring and management of terminals.

Description of drawings

The drawings constituting a part of the application are used to provide further understanding of the application, and the schematic embodiments and descriptions of the application are used to explain the application, and do not constitute an improper limitation to the application. In the attached picture:

Fig. 1 is the schematic diagram of AWOA evolution curve in the embodiment of the present invention;

Fig. 2 is a comparison chart of predicted value and real value in the embodiment of the present invention, wherein (a) is a schematic diagram of predicted value and real value, and (b) is a schematic diagram of error between predicted value and real value;

Fig. 3 is a system structure diagram in the embodiment of the present invention;

FIG. 4 is a schematic flow chart of a frequency hopping process in an embodiment of the present invention;

FIG. 5 is a schematic diagram of a transmit power update process in an embodiment of the present invention;

FIG. 6 is a schematic diagram of a bandwidth update process in an embodiment of the present invention;

Fig. 7 is a node program flow chart in the embodiment of the present invention;

FIG. 8 is a flow chart of the communication sub-gateway program in the embodiment of the present invention;

Fig. 9 is a flow chart of the gateway MCU program in the embodiment of the present invention;

FIG. 10 is a schematic structural diagram of the LoRa networking protocol in the embodiment of the present invention.

Detailed ways

It should be noted that, in the case of no conflict, the embodiments in the present application and the features in the embodiments can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and embodiments.

Aiming at the unreasonable allocation of resources and the inability to identify link changes in the LoRa network in various environments, this embodiment proposes a networking method based on edge computing, and builds a link quality evaluation model based on the AWOA-BP neural network. Link quality assessment, identify changes in link quality, perform adaptive update of radio frequency parameters, and improve communication link quality and resource utilization.

This embodiment provides a multi-parameter networking method based on edge computing gateway link quality assessment, including:

The evaluation of the quality of the LoRa link is a key indicator to judge whether the communication network is stable or not. Through the study of LoRa link characteristics, the relationship between RSSI, SNR and PLR of LoRa communication is analyzed, and an appropriate link quality evaluation model is constructed. The build process looks like this:

The LoRa communication test uses the RF transceiver of SX1276 to send data packets to the receiving end at a fixed position at the transmitting end with a transmitting power of 20dBm for data collection. The RF parameters of this test LoRa are set in Table 1.

Table 1

For each set of tests, the sending end sends 500 data packets continuously, and the receiving end records RSSI and SNR each time a data packet is received, and calculates its PLR. In order to facilitate and count the PLR, it is enlarged by 100 times during the calculation process. In order to ensure the reliability of the data, test in an open place. Constantly change the distance between the transmitter and the receiver, and count 120 sets of data. During the test, it was found that the RSSI and SNR measured by each set of data are not fixed values, and there will be jitter, so the data needs to be filtered to reduce noise and interference to be used as more accurate evaluation parameters.

In the present invention, the variance is used as a standard to evaluate the filtering effect, and the smaller the variance is, the more stable the filtered signal is and the better the filtering effect is. Through the variance comparison of the three filtering methods in Table 2, it can be seen that the variance of the Kalman filter is smaller, and the deviation between the filtered signal and the original signal is small, so the Kalman filter is used in the present invention to filter and correct the SNR and RSSI values. When a packet loss event occurs, the link parameters are relatively small, so the missing value is filled with the minimum value after filtering.

Table 2

In the test, it is found that when the SNR is less than -25dB or the RSSI is less than -125dBm, the packet loss rate is as high as 30%, and the packet loss rate changes randomly, so it is directly judged that the link quality is poor. When the SNR is greater than 25dB or the RSSI is greater than -95dB, the packet loss rate is lower than 0.4%, and it is directly judged that the link quality is good. Therefore, the present invention mainly evaluates the transition zone of the link quality.

The inertia weight is an important parameter in the whale algorithm and plays an important role in optimizing the objective function. When the weight is larger, the algorithm converges faster and the search range is wider; on the contrary, the algorithm searches more carefully and realizes searching around the optimal solution. Therefore, the choice of weights is crucial to improving the performance of the whale algorithm. The pure linear weight adjustment does not consider the actual convergence situation, which will interfere with the convergence speed of the algorithm during the iteration process. Based on this, consider adaptively adjusting the weight according to the current whale population distribution. The weight calculation formula is as follows:

w＝λ ₁ ×(P _i ^w -P _i ^b )+λ ₂ ×(ub-lb)/t (1)

In the formula: λ ₁ , λ ₂ are constants, t is the number of iterations at this time; ub, lb are the upper and lower limits of independent variables; P _i ^w , P _i ^b are the current worst and best position vectors. Adaptively adjust the weight according to the population position, and the position update formula of the improved shrinking encirclement and hunting behavior is as follows:

X(t+1)＝w×X ^* (t)-AD (2)

X(t+1)＝D'e ^bl cos(2πl)+w×X ^* (t) (3)

For the adjustment of adaptive weights, at the beginning of the algorithm iteration, if the local optimal solution is matched, and the difference between the optimal and worst position vectors is small, but the value of λ ₂ ×(ub-lb)/t does not Affected by the current population distribution, a larger weight w can also be obtained, so that the algorithm does not fall into a small-scale search in the early stage. As the number of iterations of the whale increases, the value of λ ₂ ×(ub-lb)/t gradually decreases. If the algorithm does not reach the optimal solution at this time, the value of λ ₁ ×(P _i ^w -P _i ^w ) in the weight The proportion of plays a leading role, making the algorithm optimize with a longer step size. Such a weight is composed of two parts. The latter part adjusts the algorithm to fall into local optimum, and the former part quickly seeks optimization when the number of iterations is large, so as to improve the convergence speed of the algorithm. This kind of weight self-adaptation takes into account the overall optimization of the algorithm and the enhancement of the convergence speed, and is adjusted according to the current population distribution.

In order to improve the optimization ability of the algorithm in the random search process, the search strategy is adaptively adjusted according to the fitness probability threshold Q of the current individual whale. The calculation formula of the probability threshold is:

In the formula: f _avg is the average fitness value; f _max is the maximum value of fitness value; f _min is the minimum value of fitness value. When performing random search, a random number q between [0,1] is compared with the calculated fitness probability threshold Q. If q<Q, the whale individual performs position update according to formula (5), without changing other individuals position; otherwise, the individual whale executes position update according to formula (2). The adjustment of the search strategy can generate a set of random solutions on a global scale at the beginning of the algorithm iteration, avoiding the loss of population diversity caused by the aggregation of the whale population in the early stage of the iteration, and improving the overall search ability of the algorithm.

X _rand =X _jmin +r×(X _jmax -X _jmin ) (5)

In the formula: r is a random number between [0,1], X _max and X _min are the maximum and minimum values of the variable X _rand .

In order to test and analyze the performance of the whale algorithm with the improved adaptive weight and adaptive search strategy, four basic test functions are selected, as shown in Table 3, where f ₁ and f ₂ are unimodal functions, and f ₃ and f ₄ are multimodal function. In the test, compared with the particle swarm optimization (PSO), gray wolf algorithm (GreyWolfOptimizer, GWO) and the unimproved whale algorithm, the algorithm population size is set to 30, and the maximum number of iterations is 100.

table 3

From the optimization curve of the basic test function, it can be clearly seen that the improved whale algorithm has greatly improved the convergence speed and stability, and the global search ability is stronger. In the optimization of f1 and _f4 functions, it converges to the global minimum excellent. The feasibility and high performance of the improved whale algorithm are verified. The invention takes the mean square error of the sample as the objective function, uses the improved whale algorithm to optimize the BP neural network, and assigns the optimal individual weight and threshold parameters to the BP neural network. Figure 1 shows the evolution curve of the whale optimization algorithm.

As can be seen from the figure above, when the whale algorithm iterates to 42 times, the minimum mean square error has been reached. The LoRa single node is used to collect sample data and processed by Kalman filter. Of the 120 sets of sample data obtained, 90% are used as training samples and 10% are used as test samples. The relationship between RSSI, SNR and PLR is mapped using BP neural network. In the BP neural network in the MATLAB toolkit, set RSSI and SNR as the input layer, PLR as the output layer, the maximum number of training times is 1000, the learning rate is 0.01, and the hidden layer is trained multiple times according to formula (6). The number of hidden layer nodes is determined when the mean square error of the trained model is the smallest. Among them, n represents the number of nodes in the input layer, p represents the number of nodes in the output layer, a is a constant between 1 and 10, and m represents the number of nodes in the hidden layer. The results of multiple training shows that when the hidden layer is 11, the mean square error of the training model is the smallest. In order to display the mapping model more accurately, the PLR value range of BP neural network mapping is 0-40%.

Substituting the test sample into the training model to predict the output, Figure 2 shows the comparison between the predicted value and the real value before and after the BP neural network model is optimized using the improved whale algorithm, as shown in Figure 2(a), in order to further display the predicted value before and after optimization The accuracy of , using the difference between the predicted value and the true value as the error, as shown in Figure 2(b).

From Figure 2, the absolute error between the predicted value of the basic BP neural network model and the real value can reach 3%, while the absolute error between the predicted value of the BP neural network model optimized by the whale algorithm and the real value is within 1%. MATLAB calculates the mean absolute error (mean absolute error, MAE) before optimization is 0.92968%, and the mean absolute percentage error (mean absolute percentage error, MAPE) is 17.3618%. The MAE and MAPE calculated by the BP neural network model optimized by the improved whale algorithm They are 0.28866% and 10.048% respectively. The prediction accuracy of the optimized BP neural network model is greatly improved, the evaluation effect of the link is better, and it has more research and application value.

The present invention researches and analyzes the link quality around the LoRa network. The nonlinear mapping model between RSSI, SNR and PLR is trained through the AWOA-BP neural network, and the prediction accuracy of the AWOA-BP neural network model is much higher than that of the BP neural network. Compared with the evaluation methods used in some existing technologies, the AWOA-BP neural network model has better prediction effect and accuracy, and is feasible for practical application. This model can be used to evaluate the quality of LoRa links.

In the LoRaWAN protocol, the uplink communication node uses the ALOHA access method to communicate with the gateway. The gateway acts as the access center, and each node communicates independently with the gateway. When there are multiple data packets in the channel at the same time, a collision occurs, resulting in data packet loss. At this time, the node sends data without considering whether the current communication channel is idle. After the packet is lost, the node cannot receive the response from the gateway, and the communication delay and stability are poor. The radio frequency chip SX1301 commonly used in the LoRaWAN protocol has 8 modulation channels, but it cannot demodulate multiple data packets at the same time, and cannot realize simultaneous communication of multiple channels. The concurrency is poor, and the spreading factors are not perfect orthogonality. Increased channel communication interference. Since the LoRaWAN network is a random access method, the larger the node capacity, the easier it is to cause communication conflicts. Therefore, in order to ensure the stability of the network, the network scale is reduced.

In view of the shortcomings of the LoRaWAN network, in order to improve the network capacity and resource allocation of LoRa, the customized LoRa network uses TDMA (Time division multiple access) and CSMA (Carrier Sense Multiple Access) to connect. Into a combination of ways to improve channel utilization in two different ways. The custom LoRa gateway uses 8 radio frequency modules to form 8 channels, one of which is used as a control channel, which is responsible for assigning nodes, six channels are used as communication channels to receive data uploaded by nodes periodically, and one channel is used as a backup channel to handle emergency communications and network communication. The system architecture is shown in Figure 3.

In order to realize the communication of the TDMA mode of the node, the local clock of the node must be synchronized, and time synchronization is a prerequisite for allocating fixed time slots. Commonly used time synchronization methods include satellite time service, network time service, and radio time service. Although network time service and satellite time service have high precision, in the LoRa network, considering cost, protocol and practical application adaptability, the present invention selects radio time service. When using radio time service, there will be errors in synchronization accuracy due to the transmission of synchronization messages. The LoRa network structure designed by the present invention is a star structure, and the nodes communicate with the gateway. In order to make the synchronization simple and effective, after the control channel allocates the nodes, the nodes and the sub-gateways of the communication channel perform time synchronization, and the nodes send a synchronization request to the gateway. The gateway replies to the local clock after receiving it. In order to reduce the error caused by sending synchronization messages, a two-way synchronization algorithm is used to record the local clock T ₁ of the node sending the synchronization request. After receiving the synchronization clock of the gateway, record the local clock as T ₂ , and use the difference between the two as the synchronization The compensation for message transmission over the air is

RTC＝T _rec +(T ₂ -T ₁ ) (7)

After time synchronization, the nodes can send data according to fixed time slots. However, when using a hardware clock, the crystal oscillator will drift, and the continuous accumulation of deviations will cause the failure of time synchronization and conflicts between nodes. The clock deviation formula can be expressed as follows:

Δt=νt+t _cs (8)

In the formula, v represents the crystal oscillator drift rate of the node, and t _cs represents the node synchronization deviation. With the increase of time, the deviation will become larger and larger. At this time, clock compensation correction is required. The most practical way is to perform time synchronization again after the deviation exceeds a certain accuracy range to maintain synchronization accuracy. Due to the synchronous clock skew, the division of time slots is larger than its transmission time. The communication between the node and the gateway is two-way, the node transmits sampling data uplink at a fixed period, and the gateway downlinks the control command, so the minimum value of the time slot length is set as:

τ _min =t _d +t _e +2max(Δt) (9)

The slot length τ and sampling period T determine the node capacity of each channel:

The sampling period is set according to the actual application and is generally fixed in the network, so the length of the time slot determines the capacity of the node. The minimum value of the time slot length is the uplink and downlink transmission time plus 2 times the clock deviation. Therefore, the time interval between the first synchronization node in the channel sending the synchronization request and receiving the gateway reply plus the deviation can be used as the time slot length. The synchronization request The length of the data sampled by the node is the same, and the length of the gateway's reply is the same as the length of the gateway's downlink update command. In this way, the time slot length is divided into the most, the node capacity is larger, and the throughput of the network is the highest.

The link quality of the LoRa network determines the stability and robustness of its communication. In order to achieve better communication quality of the network, the link quality must be improved. In the process of communication, due to the existence of environmental interference, the link quality is changeable. In order to ensure the reliability of communication, the radio frequency parameters can be updated to improve the link quality. In communication, channel resources, time slot resources, energy resources, etc. are limited, and it is an urgent problem to realize optimal utilization of resources and improve network capacity and throughput. The multi-channel networking method proposed by the present invention has high anti-interference ability, and can realize self-adaptive updating of radio frequency parameters through edge calculation at the gateway, thereby improving resource utilization and link quality.

After a node joins the network, it first assigns the node to an appropriate channel according to the link quality. The 8-channel gateway is designed, according to the different coverage of the receiving sensitivity of different spreading factors, the spreading factor of the control channel is set to 12, which has the largest coverage, the spreading factors of the other communication channels are set from 7 to 12, and the spare channel is set to 7. Receive sensitivity is the minimum received power that can resolve useful signals, so assign nodes to appropriate channels according to the RSSI read by the receive request allocation packet. By setting the RSSI inflection point value, as shown in Table 4 below.

Table 4

The above table is only used as an allocation strategy for nodes joining the network. In order to improve the utilization rate of channel allocation, when the number of nodes connected to the current channel reaches 1/2 of all nodes joining the network, priority is allocated to other channels. When the node is working in the communication channel, the parameters are adaptively adjusted through real-time statistics and evaluation of the link quality. In the process of communication, nodes are vulnerable to various interferences, the channel link is degraded, and the communication packet loss rate is very high. Therefore, it is necessary to improve the link quality of node communication and adaptively adjust radio frequency parameters. When testing adjacent channel interference, when the carrier frequency difference of adjacent channels is greater than 2MHz, it is not easy to suffer from adjacent channel interference. Therefore, the frequency point planning for each channel is shown in Table 5.

table 5

It can be seen from Table 5 that only one frequency point is used for the control channel, because the control channel handles burst random services and does not have high requirements for communication stability, and most of them belong to the idle state of the channel. The spreading factor of the communication channel is fixed from 7 to 12, and the carrier frequency has 21 variable frequency points, which are modifiable parameters. The design between the LoRa gateway and the node is mainly to realize the monitoring of industrial equipment, and does not require frequent uplink and downlink interactions. In order to reduce power consumption, the node only opens the receiving window when the time is synchronized, and the two exchange information, so all Edge computing is performed when nodes request synchronization. The specific update process of each radio frequency parameter is as follows:

(1) Frequency and spreading factor adaptive

In the process of LoRa network communication, it is vulnerable to various interferences or the node terminal exceeds the coverage of the sub-gateway. According to the formula (12), the spreading factor is increased, and the receiving sensitivity of the LoRa module is higher and the coverage is farther. Therefore, according to the change of link quality, link interference is identified, and the sub-gateway of the control channel obtains the free and available frequency points in Table 5 except the frequency points used in Table 5 through periodic spectrum sensing, performs frequency hopping, and adjusts the frequency and frequency of node terminals. The spreading factor is used to improve the link quality. The specific process is shown in Figure 4.

(2) Adaptive transmit power

Through the research and analysis of the transmission power, it can be known that the greater the transmission power, the better the link quality of the channel, but at the same time, the greater the energy consumption, the shorter the life cycle of the node terminal. Therefore, it is necessary to adjust the transmission power according to the link quality for the allocation of energy resources. The specific process is shown in FIG. 5 .

(3) Bandwidth adaptive

The air transit time is equal to the sum of the preamble and payload transit times. The length of the preamble is fixed. According to formula (11), the number of symbols in the payload is determined by the length of the payload, the spreading factor, and the coding rate. Therefore, when the length of the data packet, the spreading factor, and the coding rate are determined, the The definition of the symbol period T _S can be obtained, the transmission time in the air is inversely proportional to the bandwidth, and the transmission time shortens as the bandwidth increases. The shorter the data transmission time, according to the minimum slot calculation method of formula (9), the smaller t is, the larger the channel capacity is, the higher the utilization rate of the channel is, and the higher the throughput of the gateway is. Increasing the bandwidth can realize the expansion of the network, but the receiving sensitivity will decrease and the link quality will be reduced. Therefore, the expansion is performed when the link quality is good, as shown in Figure 6.

In the formula: n _pa represents the number of symbols of the payload data, PL is the length of the payload; H represents the enable header, explicit H=0, implicit H=1; when using low rate optimization, DE represents whether to enable low information Rate optimization, DE=1 if enabled, otherwise DE=0. CR represents the encoding rate, max represents the largest set element, and the ceil function returns the smallest integer not less than the expression.

The symbol period T _S is defined as:

Among them, BW represents the signal bandwidth in Hz, and SF represents the spreading factor.

S＝-174+10lg B+NF-2.5SF+10 (12)

It can be obtained from the above formula that the spreading factor and the bandwidth jointly determine the receiving sensitivity of LoRa.

The LoRa network is mainly composed of gateways and nodes. The realization of the networking is that the nodes are connected to the gateway, and the gateway performs data forwarding, storage and edge computing to manage the nodes. The gateway is composed of multiple radio frequency modules to form a multi-channel network. The program flow for gateways and nodes is as follows:

(1) Node process

The node mainly realizes the sampling function of industrial equipment information, and the present invention takes the collection of temperature information as an example. After the node starts working, it first sends a distribution request to the control sub-gateway, and after receiving the configuration command, it performs RTC clock synchronization with the communication sub-gateway, and uses TDMA to realize data transmission of multiple nodes under the same sub-gateway. If the node is the first to access the sub-gateway, it will send two synchronization requests, and calculate the time slot length according to the data packet transmission time calculated by the node. After multiple synchronization failures, the node reconnects to the control sub-gateway for distribution. After the nodes are successfully synchronized, periodic data collection and transmission are carried out. When the number of transmissions reaches the clock drift that has a great impact on the division of time slots, time synchronization is performed again. During the working process of the node, the node will receive the parameter update command from the gateway, modify the radio frequency parameters, and improve the communication quality of the node. The program flow chart of the node is shown in Figure 7.

(2) Gateway process

The multi-channel gateway is divided into a control sub-gateway and a communication sub-gateway. The control sub-gateway participates in the reception and response of the node allocation request data frame. Only the data is received and sent, and the link parameter RSSI read by the SX1276 when receiving the data packet is sent. to the MCU. The control sub-gateway is idle most of the time, and is also responsible for periodic spectrum sensing to obtain idle channels. The main function of the communication sub-gateway is to maintain a star network accessed by TDMA. The receiving window of the communication sub-gateway is always open. After receiving the data of the node, it is sent to the gateway MCU, which is stored and processed by the gateway MCU. . The gateway MCU sends the processing result to the sub-gateway, and the sub-gateway operates according to the type of the data packet. Figure 8 is a program flow chart of the communication sub-gateway.

The functions of the gateway MCU are mainly data forwarding and edge computing, and the program flow is shown in Figure 9. The gateway adopts the multi-thread concurrent method of the Linux system. The sub-thread monitors the connection status of the TCP protocol. When an abnormal connection is detected, the data is cached and the server is reconnected. When the network connection is normal, re-register the device and forward the cached data to the server, and the server can configure parameters for the gateway with downlink commands. The main thread mainly performs edge computing, processes the data forwarded by the 8-way sub-gateway, realizes the intelligentization of the gateway, and improves the reliability and robustness of LoRa network communication. After receiving the allocation request, the main thread allocates a communication channel for the node according to the physical layer parameter RSSI. When the node is newly added to the network, it needs to send a device registration frame to the server to indicate that the device has joined the network. When the synchronization request is received, the time slot length is calculated according to the first node connected to the channel to realize the maximum allocation of channel resources, and then the radio frequency parameters are self-adapted according to the link quality. According to the edge computing results, the radio frequency parameter update command is sent to the sub-gateway and the node. When the data is periodically sampled, the data is stored and sent to the server.

The networking based on edge computing refers to the characteristics of various types of communication frames in the LoRaWAN protocol, and is based on LoRa data link layer communication to meet a set of communication solutions developed for industrial automation applications. The structure of the networking protocol is shown in Figure 10.

In the networking protocol based on the parameter optimization and update of the link quality evaluation, the LoRa node realizes the collection of industrial equipment information by the physical layer sensor and sends it to the sub-gateway through the SX1276 RF chip, and the sub-gateway forwards the data frame to the gateway MCU for data processing. Store, forward and edge computing. Both the server and the gateway process and calculate data, and perform node management and link quality calculations. As we know from the above, it is an adaptive update of radio frequency parameters. The data is transmitted to the server by Ethernet, and the application layer calls and displays the data. It can be seen from the program flow of the node and the gateway in the previous article that during the communication process, there are mainly three types of uplink data packets of the nodes, which are allocation request packets, synchronization request packets, and periodic sampling data packets. For the specific data packet structure, take the periodic data packet in Table 6 as an example. The allocation request packet and the synchronization request packet have the same length, and the 4 bytes after the node bit are filled with zeros. The function of the data packet is judged by the packet type, and the CRC check bit ensures the integrity and correctness of the data transmission.

Table 6

After the gateway receives the allocation request and synchronization request from the node, it replies to the feedback command. For the periodic sampling data, the gateway only performs data storage, processing and calculation, and does not perform downlink communication. Therefore, there are mainly three types of data packets for downlink communication, namely allocation commands, synchronization commands and parameter update commands. The length of all data packets in the downlink communication is also the same, in order to maintain the accuracy of the TDMA access method and reduce the occurrence of communication conflicts. Take the allocation data packet structure Table 7 as an example, the parameter update instruction is the same, the synchronization command replaces the middle eight bytes with the RTC clock information of the sub-gateway, and the packet type distinguishes the function of the data packet.

Table 7

In addition to the information protocol for uplink and downlink communication between the gateway and the nodes, the data interaction between the gateway and the server is based on the TCP/IP protocol. The gateway processes and packages the received data and sends it to the server through the network port, and the server stores and calls the data for display. The server can send downlink instructions to configure the parameters of the gateway. According to the actual application requirements, the node capacity, sampling period, synchronization period, time slot length, etc. of the gateway can be changed. The format of the data packet is similar to the above, but the function of each bit is different. .

In this embodiment, the evaluation of the link quality is mainly realized, the link quality evaluation model based on the AWOA-BP neural network is constructed, and the accuracy and practicability of the evaluation are compared and analyzed; according to the multi-parameter adaptive update of the link quality networking method.

The above is only a preferred embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any person familiar with the technical field can easily conceive of changes or changes within the technical scope disclosed in this application Replacement should be covered within the protection scope of this application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims (3)

1. A multi-parameter networking method based on edge computing gateway link quality assessment is characterized by comprising the following steps:

collecting sample data and preprocessing the sample data;

the sample data includes: receiving signal strength RSSI, signal-to-noise ratio SNR and packet loss rate PLR, and based on the number of data packets sent by a sending end, recording the RSSI and the SNR when the data packets are received each time by a receiving end, and calculating the PLR;

the process of preprocessing the sample data comprises the following steps: filtering the RSSI and the SNR by adopting a Kalman filtering method;

constructing a link quality evaluation model based on an AWOA-BP neural network, inputting the preprocessed sample data into the link quality evaluation model for training until the output error is reduced to an expected degree or reaches a set learning iteration number, and finishing training to obtain the trained link quality evaluation model;

in the process of training the link quality assessment model, the following steps are included: self-adaptively adjusting the weight based on the current whale population distribution; adaptively adjusting a search strategy based on the fitness probability threshold of the current whale individual;

an edge computing-based LoRa network system is constructed by adopting a mode of combining Carrier Sense Multiple Access (CSMA) and Time Division Multiple Access (TDMA) access;

performing real-time link quality evaluation on the LoRa network based on the trained link quality evaluation model, and identifying the change of the link quality;

carrying out self-adaptive updating on the radio frequency parameters of the LoRa network based on the change of the link quality;

the process of constructing the LoRa network system based on edge calculation by adopting a mode of combining CSMA and TDMA access also comprises the following steps: the LoRa network system adopts a star structure, the star structure comprises a plurality of gateways and a plurality of nodes, and the gateways comprise control sub-gateways, communication sub-gateways and gateway MCU;

and performing node distribution through the control sub-gateway, requesting the receiving and responding of data frames, receiving a data packet through the communication sub-gateway, sending the data packet to the gateway MCU, and performing storage, forwarding and edge calculation on the data packet through the gateway MCU, wherein the data packet comprises a distribution request packet, a synchronization request packet and a periodic sampling data packet.

2. The multi-parameter networking method based on edge computing gateway link quality assessment according to claim 1, wherein in the process of constructing the LoRa network system based on edge computing by adopting a combination of CSMA and TDMA access:

before information transmission, the CSMA reduces communication conflict in a channel monitoring and random back-off mode, wherein the access mode of the CSMA is access in a competition mode;

and dividing time slots through the TDMA, distributing channel resources of multiple nodes based on the divided time slots, and enabling the nodes to occupy channels in respective time slots for communication in a sampling period of the nodes.

3. The multi-parameter networking method based on edge computing gateway link quality assessment according to claim 1, wherein in the process of adaptively updating the radio frequency parameters of the LoRa network system based on the change of the link quality, the change of the link quality is edge computed by the gateway MCU to obtain a radio frequency parameter updating instruction;

sending the radio frequency parameter updating instruction to the control sub-gateway, the communication sub-gateway and the node, and updating corresponding radio frequency parameters, wherein the radio frequency parameters include: frequency and spreading factor, transmit power, and bandwidth.

Images