EKF-based timestamp-free clock synchronization method and system

Abstract

The invention provides a timestamp-free clock synchronization method and system based on an EKF, and belongs to the technical field of wireless sensor networks, and the method comprises the steps: obtaining the data packet sending time of an active node and the ACK returning time of the reference node based on the data packet interaction of the active node and the reference node, and obtaining the dynamic response time and the dynamic time difference; constructing an EKF observation model based on the dynamic response time and the dynamic time difference; receiving a silent node and reference node clock skew obtained by monitoring interaction between a plurality of active nodes and corresponding reference node data packets by the silent node; constructing an independent EKF channel based on the EKF observation model and the clock skew between the silent node and the reference node; and synchronizing an active node clock, a reference node clock and a silent node clock based on the independent EKF channel. According to the invention, multi-node collaborative clock synchronization without timestamp exchange is realized, and the accuracy of clock synchronization is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of wireless sensor networks, and in particular to an EKF-based time-stamp-free clock synchronization method and system. Background Art

[0002] In wireless sensor networks, clock synchronization is the core technology for achieving node collaboration, data fusion, and event scheduling. Low-cost crystal oscillators are commonly used in nodes. However, the frequency of crystal oscillators is easily affected by factors such as ambient temperature and voltage fluctuations, causing clock deviation and offset to change dynamically over time. Therefore, how to ensure the accuracy of node dynamic tracking has become a major problem to be solved.

[0003] At present, traditional existing technology synchronization protocols mainly exchange timestamps periodically to step clock differences, but still face the problem of time-varying characteristics of crystal oscillators due to environmental changes. In terms of dynamic tracking problems, existing technologies also use timestamp-free synchronization technology to reduce communication overhead through implicit time information interaction. However, the current timestamp-free synchronization technology only supports single-parameter tracking of clock deviation or offset, cannot achieve joint optimization and lacks dynamic multi-source fusion of silent nodes, resulting in insufficient tracking accuracy of nodes. It can be seen that the technical solutions of existing technology clock synchronization that rely on timestamp exchange or clock synchronization only support single-parameter tracking of clock deviation or offset have the problem of poor clock synchronization accuracy. Summary of the Invention

[0004] In order to solve the above problems, the present invention proposes a timestamp-free clock synchronization method and system based on EKF, which realizes multi-node collaborative clock synchronization without timestamp exchange and improves the accuracy of clock synchronization.

[0005] To achieve the above-mentioned objectives, an embodiment of the present invention provides an EKF-based timestamp-free clock synchronization method, comprising: based on the data packet interaction between the active node and the reference node, obtaining the time when the active node sends the data packet and the time when the reference node returns ACK, to obtain the dynamic response time and dynamic time difference; constructing an EKF observation model based on the dynamic response time and dynamic time difference; receiving the clock deviations of the silent node and the reference node obtained by the silent node monitoring the data packet interaction between several active nodes and the corresponding reference nodes; constructing an independent EKF channel based on the EKF observation model and the clock deviations of the silent node and the reference node; and synchronizing the active node clock, the reference node clock and the silent node clock based on the independent EKF channel.

[0006] An embodiment of the present invention proposes a timestamp-free clock synchronization method based on an EKF. Active nodes and reference nodes interact through data packets. The dynamic response time and dynamic time difference are generated by the time the active node sends the data packet and the time the reference node returns an ACK. Using a timestamp-free method instead of timestamp exchange can reduce communication overhead. An EKF observation model is then constructed based on the dynamic response time and dynamic time difference to dynamically track clock deviations and offsets. A silent node then receives the clock deviations between the silent and reference nodes obtained by monitoring data packets exchanged between several active nodes and corresponding reference nodes. The silent node monitoring can indirectly extract timestamp information and establish a relative clock deviation chain relationship. Furthermore, an independent EKF channel is constructed based on the EKF observation model and the clock deviations between the silent and reference nodes. Multiple active nodes monitor and complete multi-channel observation fusion to synchronize the active, reference, and silent node clocks. This reduces communication overhead through timestamp-free synchronization technology while using the EKF to dynamically track clock deviations and offsets. This is combined with multi-channel observation fusion to synchronize the active, reference, and silent node clocks, thereby improving clock synchronization accuracy.

[0007] Furthermore, based on the data packet interaction between the active node and the reference node, the time when the active node sends the data packet and the time when the reference node returns ACK are obtained to obtain the dynamic response time and dynamic time difference, including: obtaining the time when the active node sends the data packet by recording the moment when the active node sends the data packet to the reference node; obtaining the time when the reference node returns ACK by recording the moment when the reference node responds to the data packet and returns ACK to the active node; based on the preset response parameters and the preset incremental step size, performing a modulo operation on the data packet sequence number set in the data packet to generate the dynamic response time; calculating the dynamic time difference based on the time when the active node sends the data packet, the time when the reference node returns ACK and the dynamic response time.

[0008] Through the above scheme, the time when the active node sends the data packet to the reference node and the time when the reference node responds to the data packet and returns ACK to the active node are recorded respectively, and the time when the active node sends the data packet and the time when the reference node returns ACK are obtained. According to the time when the active node sends the data packet, the time when the reference node returns ACK, and the data packet sequence number in the data packet, the dynamic response time and dynamic time difference are generated. The data packet interaction mode without timestamp is adopted to effectively reduce the communication overhead, and at the same time provide an accurate data basis for subsequent multi-channel observation fusion, thereby improving the accuracy of clock synchronization.

[0009] Furthermore, based on the time when the active node sends a data packet, the time when the reference node returns an ACK, and the dynamic response time, the dynamic time difference is calculated, including: obtaining the clock deviation between the active node and the reference node based on the time when the active node sends a data packet and the time when the reference node returns an ACK; obtaining the dynamic time difference based on the dynamic response time and the clock deviation between the active node and the reference node.

[0010] Through the above scheme, the clock deviation between the active node and the reference node is calculated by the time when the active node sends the data packet and the time when the reference node returns the ACK. Then, the dynamic time difference is calculated by combining the clock deviation between the active node and the reference node and the dynamic response time. In this way, the time difference problem is converted into a clock deviation estimation problem, which provides a basis for the subsequent silent nodes to monitor the active nodes to build a relative clock deviation chain relationship, thereby improving the accuracy of clock synchronization.

[0011] Furthermore, an EKF observation model is constructed based on the dynamic response time and the dynamic time difference, including: continuously recording the time when several active nodes send data packets and the time when several reference nodes return ACK to obtain adjacent dynamic response times and adjacent dynamic time differences; obtaining the dynamic response time difference based on the adjacent dynamic response times; obtaining the dynamic time difference change based on the adjacent dynamic time difference; and associating the dynamic response time difference and the dynamic time difference change with the clock deviation between the active node and the reference node to construct the EKF observation model.

[0012] In the above scheme, the changes in adjacent dynamic response times and adjacent dynamic time differences are extracted from the dynamic response time and dynamic time difference to construct an EKF observation model, thereby realizing dynamic tracking of clock deviations and offsets, laying the foundation for subsequent silent node monitoring and multi-node dynamic tracking, and improving the accuracy of clock synchronization.

[0013] Furthermore, the clock deviation between the silent node and the reference node obtained by the silent node monitoring the interaction of data packets between several active nodes and corresponding reference nodes includes: the first timestamp is obtained by the silent node monitoring the time when the active node sends the data packet; the second timestamp is obtained by the silent node monitoring the time when the reference node returns the ACK; the clock deviation between the silent node and the active node is obtained based on the first timestamp and the second timestamp; the clock deviation between the silent node and the reference node is obtained based on the clock deviation between the silent node and the active node and the clock deviation between the active node and the reference node.

[0014] In the above scheme, the first timestamp and the second timestamp extracted by the silent node are received, thereby indirectly extracting the timestamp information and establishing a relative clock deviation chain relationship. This takes into account the dynamic multi-source fusion of the silent node and also provides a basis for the subsequent dynamic tracking of multiple nodes, thereby improving the accuracy of clock synchronization.

[0015] Furthermore, based on the EKF observation model and the clock deviation between the silent node and the reference node, an independent EKF channel is constructed, including: constructing an observation equation based on the EKF observation model and the clock deviation between the silent node and the reference node, and obtaining observation data according to the observation equation; defining the input as the observation data, and the output as the estimated value of the clock deviation between the silent node and the reference node and the mean square error matrix, to obtain an independent EKF channel.

[0016] In the above scheme, the observation equation is constructed using the EKF observation model and the clock deviation between the silent node and the reference node to obtain the observation data, realizing EKF dynamic filtering and multi-source data fusion, improving the denoising effect of clock synchronization and the accuracy of node dynamic tracking. By constructing an independent EKF channel, a dedicated EKF channel is established for each active node to isolate the observation noise between different nodes and avoid cross contamination. At the same time, when multiple active nodes communicate with the reference node at the same time, the silent node processes multiple monitoring data in parallel, improving the stability of node monitoring and thus improving the accuracy of clock synchronization.

[0017] Furthermore, the independent EKF channel includes: initializing the estimated value and mean square error matrix of the clock deviation between the silent node and the reference node of each independent EKF channel; obtaining observation data according to the observation equation, inputting the observation data into the independent EKF channel, and obtaining the observation value and observation noise covariance of each independent EKF channel; obtaining the Kalman gain based on the observation noise covariance; and obtaining the estimated value and mean square error matrix of the clock deviation between the silent node and the reference node at the current moment based on the observation value and Kalman gain of each independent EKF channel.

[0018] Through the above scheme, when using independent EKF channels, the estimated value and mean square error matrix of the clock deviation between the silent node and the reference node of each independent EKF channel are initialized, and a dedicated EKF channel is established for each active node to isolate the observation noise between different nodes and avoid cross contamination. The observation data is input into each independent EKF channel, and the Kalman gain is calculated by combining the observation values ​​and observation noise covariance obtained from each independent EKF channel to output the estimated value and mean square error matrix of the clock deviation between the silent node and the reference node at the current moment. This provides intelligent weighting for subsequent multi-channel observation fusion to synchronize the active node clock, reference node clock and silent node clock, thereby improving the accuracy of clock synchronization.

[0019] Furthermore, based on the independent EKF channel, the active node clock, the reference node clock and the silent node clock are synchronized, including: obtaining the fusion weight based on the mean square error matrix at the current moment; obtaining the optimized estimated value based on the fusion weight and the estimated value of the deviation between the silent node and the reference node clock at the current moment; obtaining the estimated value curve based on the optimized estimated value; obtaining the true value curve based on the preset linear drift model; and synchronizing the active node clock, the reference node clock and the silent node clock based on the estimated value curve and the true value curve.

[0020] Through the above scheme, multi-channel observation data is intelligently weighted to improve the data's anti-interference ability. Then, the estimated value curve and the true value curve are constructed respectively by combining the optimized estimated value and the true value, so that the estimated value curve is close to the true value curve, that is, the real clock deviation changes, so as to synchronize the active node clock, the reference node clock and the silent node clock, thereby improving the accuracy of clock synchronization.

[0021] An embodiment of the present invention also provides an EKF-based timestamp-free clock synchronization system, including: a data acquisition module, an observation model construction module, a monitoring and receiving module, an independent EKF channel construction module and a clock synchronization module; the data acquisition module is used to obtain the time when the active node sends a data packet and the time when the reference node returns an ACK based on the data packet interaction between the active node and the reference node, so as to obtain the dynamic response time and dynamic time difference; the observation model construction module is used to construct an EKF observation model based on the dynamic response time and dynamic time difference; the monitoring and receiving module is used to receive the clock deviation between the silent node and the reference node obtained by the silent node monitoring the data packet interaction between several active nodes and the corresponding reference nodes; the independent EKF channel construction module is used to construct an independent EKF channel based on the EKF observation model and the clock deviation between the silent node and the reference node; the clock synchronization module is used to synchronize the active node clock, the reference node clock and the silent node clock based on the independent EKF channel.

[0022] The embodiment of the present invention proposes a time-stamp-free clock synchronization system based on EKF. The data acquisition module generates a dynamic response time and a dynamic time difference based on the interaction between the active node and the reference node through data packets, and uses the time when the active node sends the data packet and the time when the reference node returns the ACK. The time-stamp-free method is used instead of the time stamp exchange to reduce the communication overhead. Then the observation model construction module combines the dynamic response time and the dynamic time difference to construct the EKF observation model to dynamically track the clock deviation and offset. The monitoring receiving module then receives the clock deviation between the silent node and the reference node obtained by monitoring the interaction between the data packets of several active nodes and the corresponding reference nodes. Node monitoring can indirectly extract timestamp information and establish a relative clock deviation chain relationship. The independent EKF channel construction module further constructs an independent EKF channel based on the EKF observation model and the clock deviation between the silent node and the reference node. The clock synchronization module synchronizes the active node clock, reference node clock and silent node clock by monitoring multiple active nodes to complete multi-channel observation fusion. Therefore, while reducing the communication overhead through the timestamp-free synchronization technology, the EKF is used to dynamically track the clock deviation and offset, and combined with multi-channel observation fusion to synchronize the active node clock, reference node clock and silent node clock, thereby improving the accuracy of clock synchronization.

[0023] Furthermore, the data acquisition module is used to obtain the time when the active node sends the data packet and the time when the reference node returns ACK based on the data packet interaction between the active node and the reference node, and obtain the dynamic response time and dynamic time difference, including: a first time acquisition module, a second time acquisition module, a dynamic response time acquisition module and a dynamic time difference acquisition module; the first time acquisition module is used to obtain the time when the active node sends the data packet by recording the moment when the active node sends the data packet to the reference node; the second time acquisition module is used to obtain the time when the reference node returns ACK by recording the moment when the reference node responds to the data packet and returns ACK to the active node; the dynamic response time acquisition module is used to perform a modular operation on the data packet sequence number set in the data packet based on preset response parameters and preset incremental step size to generate a dynamic response time; the dynamic time difference acquisition module is used to calculate the dynamic time difference based on the time when the active node sends the data packet, the time when the reference node returns ACK and the dynamic response time.

[0024] Through the above scheme, the first time acquisition module and the second time acquisition module respectively record the moment when the active node sends the data packet to the reference node and the moment when the reference node responds to the data packet and returns ACK to the active node, and obtain the time when the active node sends the data packet and the time when the reference node returns ACK. The dynamic response time acquisition module and the dynamic time difference acquisition module generate the dynamic response time and dynamic time difference according to the time when the active node sends the data packet, the time when the reference node returns ACK and the data packet sequence number in the data packet, and adopt a data packet interaction mode without timestamp, which effectively reduces the communication overhead and provides an accurate data basis for subsequent multi-channel observation fusion, thereby improving the accuracy of clock synchronization. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 A schematic flow chart of a time-stamp-free clock synchronization method based on an EKF according to an embodiment of the present invention;

[0026] Figure 2 A schematic diagram of a process flow of a silent node monitoring data packets of an active node and a reference node in a clock synchronization method without timestamp based on EKF provided by one embodiment of the present invention;

[0027] Figure 3 A diagram showing the clock deviation tracking effect of an EKF-based timestamp-free clock synchronization method provided in one embodiment of the present invention;

[0028] Figure 4 A diagram showing the clock deviation tracking effect of a traditional clock synchronization method provided by an embodiment of the present invention;

[0029] Figure 5 A schematic diagram of the module structure of an EKF-based timestamp-free clock synchronization system provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0030] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0031] Example 1

[0032] See also Figure 1 , Figure 1 A flowchart of a time-stamp-free clock synchronization method based on EKF is provided in accordance with an embodiment of the present invention; Figure 1As shown, the embodiment of the present invention proposes a time stamp-free clock synchronization method based on EKF, including steps 101 to 105, each of which is specifically as follows:

[0033] Step 101: Based on the data packet interaction between the active node and the reference node, the time when the active node sends the data packet and the time when the reference node returns the ACK are obtained to obtain the dynamic response time and the dynamic time difference.

[0034] Step 102, constructing an EKF observation model based on the dynamic response time and the dynamic time difference;

[0035] Step 103: receiving a clock deviation between the silent node and the reference node obtained by monitoring the interaction of data packets between several active nodes and corresponding reference nodes by the silent node;

[0036] Step 104: construct an independent EKF channel based on the EKF observation model and the clock deviation between the silent node and the reference node;

[0037] Step 105 : Synchronize the active node clock, the reference node clock, and the silent node clock based on the independent EKF channel.

[0038] A specific implementation method is that the active node uses a timestamp-free synchronization mechanism to send data packets, wherein the active node is denoted as A, and the data packet will be sent to the reference node, wherein the reference node is denoted as R, and the time when the active node A sends the data packet and the time when the reference node R responds to the data packet and returns ACK are recorded, and the time when the active node sends the data packet and the time when the reference node returns ACK are obtained. The dynamic response time is generated according to the time when the active node sends the data packet and the time when the reference node returns ACK. The response time can usually be generated by modular operation. Then, according to the time difference observation method, the dynamic response time is calculated according to the time when the active node sends the data packet and the time when the reference node returns ACK. The observed variables response time difference and time difference change are extracted from the dynamic response time and the dynamic time difference. To construct an EKF observation model; the silent node is denoted as S, and the clock deviation between the silent node and the reference node is obtained by receiving the silent node S monitoring the interaction of the active node and the corresponding reference node data packet. Then, an independent EKF channel is constructed in combination with the EKF observation model. The input of the independent EKF channel is defined as observation data, and the output is the estimated value and mean square error matrix of the clock deviation between the silent node and the reference node. Among them, the silent node S can monitor multiple active nodes and the reference nodes corresponding to each active node. The observation data includes the original time data obtained by the silent node monitoring. Finally, multi-channel observation data are fused through the independent EKF channel to generate optimized estimated values ​​and true values ​​to construct estimated value curves and true value curves respectively, and the estimated value curves and true value curves are used to synchronize the active node clock, reference node clock and silent node clock.

[0039] An embodiment of the present invention proposes a timestamp-free clock synchronization method based on an EKF. Active nodes and reference nodes interact through data packets. The dynamic response time and dynamic time difference are generated by the time the active node sends the data packet and the time the reference node returns an ACK. Using a timestamp-free method instead of timestamp exchange can reduce communication overhead. An EKF observation model is then constructed based on the dynamic response time and dynamic time difference to dynamically track clock deviations and offsets. A silent node then receives the clock deviations between the silent and reference nodes obtained by monitoring data packets exchanged between several active nodes and corresponding reference nodes. The silent node monitoring can indirectly extract timestamp information and establish a relative clock deviation chain relationship. Furthermore, an independent EKF channel is constructed based on the EKF observation model and the clock deviations between the silent and reference nodes. Multiple active nodes monitor and complete multi-channel observation fusion to synchronize the active, reference, and silent node clocks. This reduces communication overhead through timestamp-free synchronization technology while using the EKF to dynamically track clock deviations and offsets. This is combined with multi-channel observation fusion to synchronize the active, reference, and silent node clocks, thereby improving clock synchronization accuracy.

[0040] A preferred solution, based on the data packet interaction between the active node and the reference node, obtains the time when the active node sends the data packet and the time when the reference node returns ACK, and obtains the dynamic response time and dynamic time difference, including: obtaining the time when the active node sends the data packet by recording the moment when the active node sends the data packet to the reference node; obtaining the time when the reference node returns ACK by recording the moment when the reference node responds to the data packet and returns ACK to the active node; based on the preset response parameters and the preset incremental step size, performing a modulo operation on the data packet sequence number set in the data packet to generate the dynamic response time; and calculating the dynamic time difference based on the time when the active node sends the data packet, the time when the reference node returns ACK, and the dynamic response time.

[0041] In one preferred embodiment, the active node A uses a timestamp-free synchronization mechanism to send a data packet to the reference node R, records the time when the active node A sends the data packet and the time when the reference node R responds to the data packet and returns an ACK, obtains the time when the active node sends the data packet and the time when the reference node returns an ACK, and performs a modulo operation on the data packet sequence number set in the data packet according to a preset response parameter and a preset incremental step size to generate a dynamic response time. The response time can usually be generated using a modulo operation mod method, as follows:

[0042] Δ i =a×(seq i modη)+b;

[0043] Where η is the modulus, a is the incremental step size (equivalent to the preset incremental step size, set as a constant), b is the minimum response time (equivalent to the preset response parameter, set as a constant), seq i is the data packet sequence number, Δ i It is the dynamic response time; an example is:

[0044]

[0045] Interpretation: When seq i =7, Δ7=55ms;

[0046] Then, according to the time difference observation method, the dynamic response time is generated based on the time when the active node sends the data packet and the time when the reference node returns the ACK to calculate the dynamic time difference, which is specifically expressed as follows:

[0047] S i =(1+ρ (AR) )·(Δ i +δ+ω i );

[0048] Where S i Dynamic time difference, ρ (AR) is the clock skew of the active node relative to the reference node, δ is the fixed delay, ω i is random noise, Δ i is the dynamic response time.

[0049] Then, according to the time difference observation method, the dynamic response time is generated according to the time when the active node sends the data packet and the time when the reference node returns the ACK to calculate the dynamic time difference.

[0050] Through the above scheme, the time when the active node sends the data packet to the reference node and the time when the reference node responds to the data packet and returns ACK to the active node are recorded respectively, and the time when the active node sends the data packet and the time when the reference node returns ACK are obtained. According to the time when the active node sends the data packet, the time when the reference node returns ACK, and the data packet sequence number in the data packet, the dynamic response time and dynamic time difference are generated. The data packet interaction mode without timestamp is adopted to effectively reduce the communication overhead, and at the same time provide an accurate data basis for subsequent multi-channel observation fusion, thereby improving the accuracy of clock synchronization.

[0051] A preferred solution calculates the dynamic time difference based on the time when the active node sends a data packet, the time when the reference node returns an ACK, and the dynamic response time, including: obtaining the clock deviation between the active node and the reference node based on the time when the active node sends a data packet and the time when the reference node returns an ACK; obtaining the dynamic time difference based on the dynamic response time and the clock deviation between the active node and the reference node.

[0052] A preferred solution can be implemented by using a time difference observation method to generate a dynamic response time based on the time when the active node sends a data packet and the time when the reference node returns an ACK to calculate the clock deviation between the active node and the reference node. Then, a dynamic time difference calculation equation is constructed based on the clock deviation between the active node and the reference node. The specific expression is as follows:

[0053] S i =(1+ρ (AR) )·(Δ i +δ+ω i );

[0054] Where S i Dynamic time difference, ρ (AR) is the clock skew of the active node relative to the reference node, δ is the fixed delay, ω i is random noise, Δ i is the dynamic response time.

[0055] One specific interpretation of the dynamic time difference calculation equation is: let the fixed delay δ = 5ms, the random noise ω1 ~ N (0, 0.12), ρ (AR) =50ppm; S1=(1+50×10-6)·(10+5+0.1)=15.1008ms.

[0056] Through the above scheme, the clock deviation between the active node and the reference node is calculated by the time when the active node sends the data packet and the time when the reference node returns the ACK. Then, the dynamic time difference is calculated by combining the clock deviation between the active node and the reference node and the dynamic response time. In this way, the time difference problem is converted into a clock deviation estimation problem, which provides a basis for the subsequent silent nodes to monitor the active nodes to build a relative clock deviation chain relationship, thereby improving the accuracy of clock synchronization.

[0057] A preferred solution is to construct an EKF observation model based on dynamic response time and dynamic time difference, including: continuously recording the time when several active nodes send data packets and the time when several reference nodes return ACK to obtain adjacent dynamic response times and adjacent dynamic time differences; obtaining dynamic response time differences based on adjacent dynamic response times; obtaining dynamic time difference changes based on adjacent dynamic time differences; and associating the dynamic response time differences and dynamic time difference changes with the clock deviations of the active nodes and the reference nodes to construct the EKF observation model.

[0058] A preferred embodiment of the present invention is to continuously record the time when multiple groups of active nodes send data packets and the time when the corresponding reference nodes return ACKs. The adjacent dynamic response time difference Δj′ is extracted from the adjacent dynamic response time by subtracting the dynamic response time of adjacent active nodes and reference nodes. The calculation formula is as follows:

[0059] Δj′=Δ i+1 -Δi ;

[0060] Similarly, the dynamic response time of the adjacent active node and the reference node is substituted into the dynamic time difference S i The calculation formula is used to obtain the corresponding adjacent dynamic time differences of the active nodes and the reference nodes, and then the dynamic time difference change S is extracted from the adjacent dynamic time differences. j ′, the specific calculation formula is as follows:

[0061] S j ′=S i+1 -S i ;

[0062] Then, the dynamic response time difference and the dynamic time difference variation are associated with the clock deviation between the active node and the reference node to construct the EKF observation model, which is specifically expressed as follows:

[0063]

[0064] Where Δj′ is the time difference between adjacent dynamic responses, S j ′ is the dynamic time difference variation, ω′ j is the observation noise, ρ (AR) is the clock deviation between the active node and the reference node.

[0065] In the above scheme, the changes in adjacent dynamic response times and adjacent dynamic time differences are extracted from the dynamic response time and dynamic time difference to construct an EKF observation model, thereby realizing dynamic tracking of clock deviations and offsets, laying the foundation for subsequent silent node monitoring and multi-node dynamic tracking, and improving the accuracy of clock synchronization.

[0066] A preferred solution is to receive the clock deviation between a silent node and a reference node obtained by a silent node monitoring the interaction of data packets between several active nodes and corresponding reference nodes, including: receiving the time when the silent node monitors the time when the active node sends a data packet to obtain a first timestamp; receiving the time when the silent node monitors the time when the reference node returns an ACK to obtain a second timestamp; based on the first timestamp and the second timestamp, obtain the clock deviation between the silent node and the active node; based on the clock deviation between the silent node and the active node and the clock deviation between the active node and the reference node, obtain the clock deviation between the silent node and the reference node.

[0067] A preferred embodiment of the present invention can be seen in Figure 2 , Figure 2 A schematic diagram of a process flow of a silent node monitoring active node and reference node data packets in a clock synchronization method without timestamp based on EKF provided by an embodiment of the present invention; Figure 2 As shown, the receiving silent node monitors the time when the active node sends the data packet and obtains the first timestamp, which is recorded as The receiving silent node monitors the ACK time returned by the reference node and obtains the second timestamp, which is recorded as Combined with the implicit relationship between the first timestamp and the second timestamp, the relative clock deviation between the silent node and the reference node is indirectly tracked, which is specifically expressed as:

[0068]

[0069] Where, It represents the local clock time when the first active node (A) sends a data packet during the first observation period (i=1) recorded by the silent node (S); Indicates the time when the active node sends the data packet, It is represented as the time when the reference node R responds to the data packet and returns ACK. The time when the reference node R responds to the data packet is Figure 2 In Chinese δ is a fixed delay, ω i is random noise;

[0070] Then, based on the clock deviation between the silent node and the active node and the clock deviation between the active node and the reference node, and according to the chain relationship of relative clock deviations, the clock deviation expression between the silent node and the reference node is established, which is expressed as:

[0071]

[0072] Where, ρ (SR) is the clock deviation of the silent node (S) relative to the reference node (R), ρ (AR) is the clock skew of the active node (A) relative to the reference node (R), ρ (AS) The clock skew of the silent node (S) relative to the active node (A).

[0073] In the above scheme, the first timestamp and the second timestamp extracted by the silent node are received, thereby indirectly extracting the timestamp information and establishing a relative clock deviation chain relationship. This takes into account the dynamic multi-source fusion of the silent node and also provides a basis for the subsequent dynamic tracking of multiple nodes, thereby improving the accuracy of clock synchronization.

[0074] An optimal solution is to construct an independent EKF channel based on the EKF observation model and the clock deviation between the silent node and the reference node, including: constructing an observation equation based on the EKF observation model and the clock deviation between the silent node and the reference node, and obtaining observation data according to the observation equation; defining the input as the observation data and the output as the estimated value of the clock deviation between the silent node and the reference node and the mean square error matrix to obtain the independent EKF channel.

[0075] In one preferred embodiment, in the prediction step of the EKF, the current state is estimated using the state at the previous moment, and the state transition model is defined as:

[0076] x k [n] = F·x k [n-1]+w k [n];

[0077] in, is the state vector, where θ k [n] is the clock deviation, ρ k [n] is the clock drift rate, is the state transfer matrix, where Δt is the time interval between adjacent observations, w k [n]Process noise.

[0078] In the above formula, the clock deviation θ k Evolution of [n]:

[0079] θ k [n] = θ k [n-1]+ρ k [n-1]·Δt+w θ ;

[0080] Among them, θ k [n] is the clock deviation of the kth channel at time n, ρ k [n-1] is the clock drift rate of the kth channel at time n-1, w θ is the process noise of the clock skew.

[0081] In the above formula, the clock drift rate ρ k Evolution of [n]:

[0082] ρ k [n] = ρ k [n-1]+w ρ ;

[0083] Among them, ρ k [n] The estimated clock drift rate of the kth independent EKF channel at time n, w ρ Process noise of the clock drift rate.

[0084] Combined with the EKF observation model, the observation equation is constructed using the timestamp difference and chain relationship:

[0085]

[0086] Among them, z k (n) is the observed value of the kth channel at the nth moment, S i (n) is the dynamic time difference at the nth moment, is the clock deviation of active node A relative to reference node R at time n, Δ i (n) is the dynamic response time at the nth moment.

[0087] The input of the independent EKF channel is defined as the observation value z of the kth channel at the nth time k (n), the output is the estimated value of the clock deviation between the silent node and the reference node and the mean square error matrix; where the estimated value of the clock deviation between the silent node and the reference node is expressed as follows:

[0088]

[0089] Where, is the a priori estimate (equivalent to the predicted value) of the clock drift of the silent node S relative to the reference node R at time n, The posterior estimate of the clock drift rate at time n-1, α is the Markov correlation coefficient;

[0090] The mean square error matrix is ​​expressed as follows:

[0091]

[0092] Among them, M k (n), P k (n) is the mean square error matrix, I is the identity matrix, K k (n) is the Kalman gain matrix, H k (n) is the Jacobian matrix of the observation equation, is the prior error covariance matrix.

[0093] After defining the input and output, an independent EKF channel is constructed.

[0094] In the above scheme, the observation equation is constructed using the EKF observation model and the clock deviation between the silent node and the reference node to obtain the observation data, realizing EKF dynamic filtering and multi-source data fusion, improving the denoising effect of clock synchronization and the accuracy of node dynamic tracking. By constructing an independent EKF channel, a dedicated EKF channel is established for each active node to isolate the observation noise between different nodes and avoid cross contamination. At the same time, when multiple active nodes communicate with the reference node at the same time, the silent node processes multiple monitoring data in parallel, improving the stability of node monitoring and thus improving the accuracy of clock synchronization.

[0095] A preferred solution, an independent EKF channel, includes: initializing the estimated value and mean square error matrix of the clock deviation between the silent node and the reference node of each independent EKF channel; obtaining observation data according to the observation equation, inputting the observation data into the independent EKF channel, and obtaining the observation value and observation noise covariance of each independent EKF channel; obtaining the Kalman gain based on the observation noise covariance; and obtaining the estimated value and mean square error matrix of the clock deviation between the silent node and the reference node at the current moment based on the observation value and Kalman gain of each independent EKF channel.

[0096] A preferred embodiment of the scheme is to initialize the state variables and covariance matrix of each channel, and the estimated value of the clock deviation between the silent node and the reference node of each channel is ρ (SR) , set the initial value ρ (SR) =0, initialize the covariance matrix or error matrix to in, represents the covariance matrix or error matrix of the k-th iteration or k-th time step, represents the initial variance or noise level, and I is the identity matrix.

[0097] The observation data is input into the independent EKF channel, and the observation value and observation noise covariance of each independent EKF channel are obtained through the observation equation, and the Kalman gain is further calculated, which is specifically expressed as:

[0098]

[0099] Where K k (n) is the Kalman gain, which determines the influence weight of the observation value on the state update. is the predicted covariance matrix of the kth channel at time n, H k =1, R k is the observation noise covariance, where Recursively calculated by the state transition model, R k 100 sets of z are collected in a static environment k (n), calculate the variance, H k is the observation equation z k (n) Derivative of the state variable (actually a matrix, H k =1 is a scalar special case);

[0100] K k (n) The weight of the influence of the observation value on the state update is expressed as:

[0101]

[0102] in, Is the predicted state, if the observed value z k (n) reliable, then K k (n) increases, if the predicted value Reliable, then K k (n) Decrease.

[0103] Record the output of each independent EKF channel. The output of each channel specifically includes the estimated value at the current moment. And the mean square error matrix M k (n) = Pk(n);

[0104] Through the above scheme, when using independent EKF channels, the estimated value and mean square error matrix of the clock deviation between the silent node and the reference node of each independent EKF channel are initialized, and a dedicated EKF channel is established for each active node to isolate the observation noise between different nodes and avoid cross contamination. The observation data is input into each independent EKF channel, and the Kalman gain is calculated by combining the observation values ​​and observation noise covariance obtained from each independent EKF channel to output the estimated value and mean square error matrix of the clock deviation between the silent node and the reference node at the current moment. This provides intelligent weighting for subsequent multi-channel observation fusion to synchronize the active node clock, reference node clock and silent node clock, thereby improving the accuracy of clock synchronization.

[0105] A preferred solution is based on an independent EKF channel to synchronize the active node clock, reference node clock and silent node clock, including: obtaining a fusion weight based on the mean square error matrix at the current moment; obtaining an optimized estimated value based on the fusion weight and the estimated value of the deviation between the silent node and the reference node clock at the current moment; obtaining an estimated value curve based on the optimized estimated value; obtaining a true value curve based on a preset linear drift model; and synchronizing the active node clock, reference node clock and silent node clock based on the estimated value curve and the true value curve.

[0106] A preferred embodiment of the present invention is to calculate the mean square error matrix M at the current moment output by each independent EKF channel. k (n), the fusion weight is calculated according to the scalar weighting criterion, specifically:

[0107]

[0108] Where M k (n) represents the mean square error matrix of the kth channel at time n, tr(M k (n)) represents the matrix M k The trace of (n), that is, the sum of the diagonal elements of the matrix, a k : represents the fusion weight of the kth channel, L: represents the total number of channels.

[0109] After the fusion weight is calculated, the optimized estimate is further calculated based on the fusion weight and the estimated value of the clock offset between the silent node and the reference node at the current moment. The specific calculation method is as follows:

[0110]

[0111] Where x k [n|n] represents the estimated value of the kth channel at time n, x opt [n|n] represents the optimized estimate at time n.

[0112] Finally, a linear drift model is built to obtain the true value curve, as follows:

[0113] ρ true (t) = c·t+d;

[0114] Where c is the clock drift rate, d is the initial clock deviation, and t is the time variable. By optimizing the estimated value to construct the estimated value curve, the two curves are aligned in time series to obtain a comparison curve of the true value and the estimated value to synchronize the active node clock, the reference node clock, and the silent node clock. It is worth mentioning that see Figure 3 and Figure 4 , Figure 3 A diagram showing the clock deviation tracking effect of an EKF-based timestamp-free clock synchronization method provided in one embodiment of the present invention; Figure 4 A clock deviation tracking effect diagram of a traditional clock synchronization method provided by an embodiment of the present invention; Figure 3 and Figure 4 As shown, compared with the traditional clock synchronization method, the silent node synchronization energy consumption of the time-stamp-free clock synchronization method proposed in the embodiment of the present invention is lower than that of the traditional solution. In the comparison curve, the points with the smallest error are marked. These points are considered to be the locations with the best denoising effect. By comparing and calculating the point-by-point error of the traditional clock synchronization method and the time-stamp-free clock synchronization method proposed in the embodiment of the present invention, the analysis is as follows:

[0115] e(t i ) 传统方法 = True clock deviation (t i )-estimated clock bias (t i )=1.0074ms;

[0116] e(t i ) EKF法 = True clock deviation (t i )-estimated clock bias (t i )=0.6234ms;

[0117] From the above calculation results, it can be seen that the timestamp-free clock synchronization method proposed in the embodiment of the present invention has better denoising effect and supports observation fusion of multiple active nodes.

[0118] Through the above scheme, multi-channel observation data is intelligently weighted to improve the data's anti-interference ability. Then, the estimated value curve and the true value curve are constructed respectively by combining the optimized estimated value and the true value, so that the estimated value curve is close to the true value curve, that is, the real clock deviation changes, so as to synchronize the active node clock, the reference node clock and the silent node clock, thereby improving the accuracy of clock synchronization.

[0119] Example 2

[0120] See also Figure 5 , Figure 5A schematic diagram of the module structure of a time-stamp-free clock synchronization system based on EKF is provided in one embodiment of the present invention; Figure 5 As shown, an embodiment of the present invention further provides an EKF-based timestamp-free clock synchronization system, including: a data acquisition module 201, an observation model construction module 202, a monitoring and receiving module 203, an independent EKF channel construction module 204 and a clock synchronization module 205; the data acquisition module 201 is used to obtain the time when the active node sends a data packet and the time when the reference node returns an ACK based on the data packet interaction between the active node and the reference node, so as to obtain the dynamic response time and dynamic time difference; the observation model construction module 202 is used to construct an EKF observation model based on the dynamic response time and dynamic time difference; the monitoring and receiving module 203 is used to receive the clock deviation between the silent node and the reference node obtained by the silent node monitoring the data packet interaction between several active nodes and the corresponding reference nodes; the independent EKF channel construction module 204 is used to construct an independent EKF channel based on the EKF observation model and the clock deviation between the silent node and the reference node; the clock synchronization module 205 is used to synchronize the active node clock, the reference node clock and the silent node clock based on the independent EKF channel.

[0121] A specific implementation method is as follows: the data acquisition module 201 sends a data packet using a timestamp-free synchronization mechanism according to the active node, wherein the active node is denoted as A, and the data packet will be sent to the reference node, wherein the reference node is denoted as R. The time when the active node A sends the data packet and the time when the reference node R responds to the data packet and returns ACK are recorded to obtain the time when the active node sends the data packet and the time when the reference node returns ACK. The dynamic response time is generated according to the time when the active node sends the data packet and the time when the reference node returns ACK. The response time can usually be generated by modular operation. The observation model construction module 202 then calculates the dynamic time difference according to the time difference observation method, generates the dynamic response time according to the time when the active node sends the data packet and the time when the reference node returns ACK, and extracts the observation variable response time difference and the time difference change from the dynamic response time and the dynamic time difference to construct EK F observation model; the silent node is recorded as S, the monitoring receiving module 203 receives the clock deviation of the silent node and the reference node obtained by the silent node S monitoring the interaction of the active node and the corresponding reference node data packet, and the independent EKF channel construction module 204 then combines the EKF observation model to construct an independent EKF channel. The input of the independent EKF channel is defined as observation data, and the output is the estimated value and mean square error matrix of the clock deviation between the silent node and the reference node. Among them, the silent node S can monitor multiple active nodes and the reference nodes corresponding to each active node. The observation data includes the original time data obtained by the silent node monitoring. Finally, the clock synchronization module 205 performs multi-channel observation data fusion through the independent EKF channel to generate optimized estimated values ​​and true values ​​to construct estimated value curves and true value curves respectively, and synchronize the active node clock, reference node clock and silent node clock with the estimated value curve and true value curve.

[0122] The embodiment of the present invention proposes a time-stamp-free clock synchronization system based on EKF. The data acquisition module 201 generates a dynamic response time and a dynamic time difference based on the interaction between the active node and the reference node through data packets, and uses the time when the active node sends the data packet and the time when the reference node returns the ACK. The time-stamp-free method is used instead of the time stamp exchange to reduce the communication overhead. Then the observation model construction module 202 combines the dynamic response time and the dynamic time difference to construct an EKF observation model to dynamically track the clock deviation and offset. The monitoring receiving module 203 then receives the clock deviation between the silent node and the reference node obtained by the silent node monitoring the interaction between the data packets of several active nodes and the corresponding reference nodes. Silent node monitoring can indirectly extract timestamp information and establish a relative clock deviation chain relationship, and further the independent EKF channel construction module 204 constructs an independent EKF channel according to the EKF observation model and the clock deviation between the silent node and the reference node. The clock synchronization module 205 completes multi-channel observation fusion by monitoring multiple active nodes to synchronize the active node clock, reference node clock and silent node clock. Thus, while reducing the communication overhead through the timestamp-free synchronization technology, the EKF is used to dynamically track the clock deviation and offset, and the active node clock, reference node clock and silent node clock are synchronized in combination with multi-channel observation fusion, thereby improving the accuracy of clock synchronization.

[0123] Furthermore, the data acquisition module 201 is used to obtain the time when the active node sends the data packet and the time when the reference node returns ACK based on the data packet interaction between the active node and the reference node, and obtain the dynamic response time and dynamic time difference, including: a first time acquisition module 301, a second time acquisition module 302, a dynamic response time acquisition module 303 and a dynamic time difference acquisition module 304; the first time acquisition module 301 is used to obtain the time when the active node sends the data packet by recording the moment when the active node sends the data packet to the reference node; the second time acquisition module 302 is used to obtain the time when the reference node returns ACK by recording the moment when the reference node responds to the data packet and returns ACK to the active node; the dynamic response time acquisition module 303 is used to perform a modular operation on the data packet sequence number set in the data packet based on a preset response parameter and a preset incremental step size to generate a dynamic response time; the dynamic time difference acquisition module 304 is used to calculate the dynamic time difference based on the time when the active node sends the data packet, the time when the reference node returns ACK and the dynamic response time.

[0124] Through the above scheme, the first time acquisition module 301 and the second time acquisition module 302 respectively record the moment when the active node sends the data packet to the reference node and the moment when the reference node responds to the data packet and returns ACK to the active node, and obtain the time when the active node sends the data packet and the time when the reference node returns ACK. The dynamic response time acquisition module 303 and the dynamic time difference acquisition module 304 generate dynamic response time and dynamic time difference according to the time when the active node sends the data packet, the time when the reference node returns ACK and the data packet sequence number in the data packet, and adopt a data packet interaction mode without timestamp, which effectively reduces communication overhead and provides an accurate data basis for subsequent multi-channel observation fusion, thereby improving the accuracy of clock synchronization.

[0125] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the technical principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

[0126] In the description of this specification, the reference terms "one embodiment," "some embodiments," "examples," "specific examples," or "some examples" mean that the specific features, structures, materials, or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. Moreover, the specific features, structures, materials, or characteristics described may be combined in any appropriate manner in any one or more embodiments or examples. In addition, those skilled in the art may combine and integrate different embodiments or examples described in this specification, as well as features of different embodiments or examples, unless they are mutually inconsistent.

[0127] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be understood to indicate or imply relative importance or implicitly specify the number of technical features indicated. Thus, features specified as "first" or "second" may explicitly or implicitly include at least one of such features. In the description of this application, "plurality" means two or more, unless otherwise specifically defined.

Claims

1. A time-stamp-free clock synchronization method based on EKF, characterized in that: include: Based on the data packet interaction between the active node and the reference node, the time when the active node sends the data packet and the time when the reference node returns the ACK are obtained, and the dynamic response time and dynamic time difference are obtained; Constructing an EKF observation model based on the dynamic response time and the dynamic time difference; The receiving silent node monitors the clock deviation between several active nodes and the corresponding reference node data packets, and obtains the clock deviation between the silent node and the reference node; Building an independent EKF channel based on the EKF observation model and the clock deviation between the silent node and the reference node; Based on the independent EKF channels, active node clocks, reference node clocks, and silent node clocks are synchronized.

2. The EKF-based time-stamp-free clock synchronization method according to claim 1, wherein: The method of obtaining the time when the active node sends the data packet and the time when the reference node returns the ACK based on the data packet interaction between the active node and the reference node, and obtaining the dynamic response time and dynamic time difference, includes: By recording the moment when the active node sends the data packet to the reference node, the time when the active node sends the data packet is obtained; By recording the time when the reference node responds to the data packet and returns ACK to the active node, the reference node return ACK time is obtained; Based on a preset response parameter and a preset incremental step size, performing a modulo operation on a data packet sequence number set in the data packet to generate a dynamic response time; The dynamic time difference is calculated based on the time when the active node sends the data packet, the time when the reference node returns the ACK, and the dynamic response time.

3. The EKF-based time-stamp-free clock synchronization method according to claim 2, wherein: Calculating the dynamic time difference based on the time when the active node sends the data packet, the time when the reference node returns the ACK, and the dynamic response time, including: Obtaining a clock deviation between the active node and the reference node based on the time when the active node sends a data packet and the time when the reference node returns an ACK; A dynamic time difference is obtained based on the dynamic response time and the clock deviation between the active node and the reference node.

4. The EKF-based time-stamp-free clock synchronization method according to claim 3, wherein: Constructing an EKF observation model based on the dynamic response time and the dynamic time difference includes: Continuously record the time it takes for several active nodes to send data packets and the time it takes for several reference nodes to return ACKs, and obtain the adjacent dynamic response time and adjacent dynamic time difference; Based on adjacent dynamic response times, a dynamic response time difference is obtained; Based on the adjacent dynamic time differences, a dynamic time difference variation is obtained; The dynamic response time difference and the dynamic time difference variation are associated with the clock deviation between the active node and the reference node to construct an EKF observation model.

5. The EKF-based time-stamp-free clock synchronization method according to claim 3, wherein: The receiving silent node monitors the interaction of data packets between a plurality of active nodes and corresponding reference nodes to obtain a clock deviation between the silent node and the reference node, including: The receiving silent node monitors the time when the active node sends the data packet to obtain a first timestamp; The receiving silent node monitors the ACK time returned by the reference node to obtain a second timestamp; Obtaining a clock deviation between a silent node and an active node based on the first timestamp and the second timestamp; Based on the clock deviation between the silent node and the active node and the clock deviation between the active node and the reference node, the clock deviation between the silent node and the reference node is obtained.

6. The EKF-based time-stamp-free clock synchronization method according to claim 1, wherein: Based on the EKF observation model and the clock deviation between the silent node and the reference node, an independent EKF channel is constructed, including: Constructing an observation equation based on the EKF observation model and the clock deviation between the silent node and the reference node, and acquiring observation data according to the observation equation; The input is defined as the observation data, and the output is the estimated value of the clock deviation between the silent node and the reference node and the mean square error matrix, thereby obtaining an independent EKF channel.

7. The EKF-based time-stamp-free clock synchronization method according to claim 6, wherein: The independent EKF channel includes: Initialize the estimated value and mean square error matrix of the clock deviation between the silent node and the reference node of each independent EKF channel; Obtaining observation data according to the observation equation, inputting the observation data into the independent EKF channels, and obtaining observation values ​​and observation noise covariance of each independent EKF channel; Obtaining a Kalman gain based on the observation noise covariance; Based on the observation values ​​of each independent EKF channel and the Kalman gain, an estimated value and a mean square error matrix of the clock deviation between the silent node and the reference node at the current moment are obtained.

8. The EKF-based time-stamp-free clock synchronization method according to claim 7, wherein: Synchronizing an active node clock, a reference node clock, and a silent node clock based on the independent EKF channel, including: Based on the mean square error matrix at the current moment, a fusion weight is obtained; Obtaining an optimized estimated value based on the fusion weight and an estimated value of the clock deviation between the silent node and the reference node at a current moment; Based on the optimized estimated values, obtaining an estimated value curve; Based on the preset linear drift model, the true value curve is obtained; Based on the estimated value curve and the true value curve, active node clocks, reference node clocks, and silent node clocks are synchronized.

9. A time-stamp-free clock synchronization system based on EKF, characterized in that: include: Data acquisition module, observation model construction module, monitoring reception module, independent EKF channel construction module and clock synchronization module; The data acquisition module is used to obtain the time when the active node sends the data packet and the time when the reference node returns the ACK based on the data packet interaction between the active node and the reference node, and obtain the dynamic response time and dynamic time difference; The observation model construction module is used to construct an EKF observation model based on the dynamic response time and dynamic time difference; The monitoring and receiving module is used to receive the clock deviation between the silent node and the reference node obtained by the silent node monitoring the interaction of the data packets of several active nodes and the corresponding reference nodes; The independent EKF channel construction module is used to construct an independent EKF channel based on the EKF observation model and the clock deviation between the silent node and the reference node; The clock synchronization module is used to synchronize the active node clock, the reference node clock and the silent node clock based on the independent EKF channel.

10. The EKF-based time-stamp-free clock synchronization system according to claim 9, wherein: The data acquisition module is used to acquire the time when the active node sends the data packet and the time when the reference node returns the ACK based on the data packet interaction between the active node and the reference node, and obtain the dynamic response time and dynamic time difference, including: A first time acquisition module, a second time acquisition module, a dynamic response time acquisition module and a dynamic time difference acquisition module; The first time acquisition module is used to obtain the time when the active node sends the data packet by recording the time when the active node sends the data packet to the reference node; The second time acquisition module is used to obtain the reference node return ACK time by recording the time when the reference node responds to the data packet and returns ACK to the active node; The dynamic response time acquisition module is used to perform a modulo operation on the data packet sequence number set in the data packet based on a preset response parameter and a preset incremental step size to generate a dynamic response time; The dynamic time difference acquisition module is used to calculate the dynamic time difference based on the time when the active node sends a data packet, the time when the reference node returns an ACK, and the dynamic response time.

Images