CN106912100B - Time Synchronization Method of Home Appliance Network Based on TPSN and TSync - Google Patents

Abstract

The invention discloses a home appliance network time synchronization method based on TPSN and TSync, which includes the following steps: S1. Add a GPS signal receiver to the smart home appliance, and use the smart home appliance node as the root node of time synchronization, and establish or update Root node; S2. Determine the source time: set the source time to be obtained from (T _G , T _B , T _S ). S3. Select a synchronization mode. The home appliance network time synchronization method based on TPSN and TSync described in the present invention does not select a single time source for the determination of the synchronization source time, but integrates the GPS signal receiver, the network server and the internal clock of the node for different situations. , adopting different time sources, thereby greatly improving the effectiveness of time synchronization, avoiding the time blind zone of time synchronization, is a method for adopting different wireless network node time synchronization methods for different situations.

Description

Time Synchronization Method of Home Appliance Network Based on TPSN and TSync

technical field

The invention relates to a monitoring data time synchronization method of a smart home appliance wireless network, in particular to a hybrid time synchronization method based on a TPSN method and TSync.

Background technique

At present, the smart grid is a new development direction of the power industry, and its realization depends on the real-time and accurate grasp of the information of each link of the power grid. As the "intelligent information perception end", the household electricity network provides accurate and time-synchronized data for the power operation management department, which will surely provide an important impetus for promoting the development of the smart grid.

From the end of the last century to the beginning of this century, countries around the world began to use information technology, microelectronics technology and computer technology to upgrade the traditional low-voltage electrical appliances, and a major breakthrough has been made in technology, making the electrical equipment at the user end of the power grid initially realize communication and intelligent control. and network control. Through the computer network, accurately detect, calculate and monitor the operating status of the user-side electrical equipment, so that the power distribution network can achieve the best coordination, greatly improving the economic security and reliability of the power grid operation, and providing user-side electrical equipment It has laid a technical foundation for system integration, load monitoring, fault warning, power quality management, power management, access to distributed new energy, and access to distributed energy storage devices. Two-way communication, two-way billing, and the interaction between the grid and users will also greatly improve energy efficiency. Therefore, the smart grid user-side electrical equipment plays a very important role in the energy saving, safe and reliable operation of the smart grid, the development and application of new energy, and the improvement of energy efficiency. It is an important part of a strong smart grid.

The home energy management system refers to the use of information and communication technology as a means to realize the management, monitoring and reduction of energy consumption of user electrical appliances. To realize the coordinated control of various household electrical equipment and achieve management goals, the time synchronization of each intelligent electrical equipment is essential. Being able to realize the time synchronization of each node of the home appliance network is also of great significance for scientific research related to smart home appliances. It is also of great value for the establishment of smart home appliance operation models and smart home appliance fault diagnosis.

In order to make the monitoring data of all nodes of the smart home appliance network refer to each other, and facilitate the coordinated control of the entire home appliance network and the optimization of household energy consumption, it is necessary to synchronize the time of all smart home appliance nodes. The monitoring data after time synchronization will have higher scientific research value, and can provide more accurate and reasonable data for the design and related research of the home energy management system.

With the development of the future society, there will be many home appliances included in the smart home, and some of them, including fire safety monitoring equipment, need to work continuously. Therefore, in order to improve the reliability of time synchronization, the credibility of the time synchronization source time should be increased.

Therefore, designing a time synchronization method for each node of the home appliance network has become a problem to be solved urgently.

Contents of the invention

According to the above-mentioned technical problems, a time synchronization method for home appliance network based on TPSN and TSync is provided, which is used to solve the shortcoming of lacking a reliable time synchronization method for each node of home appliance network in the prior art. The technical means adopted in the present invention are as follows:

A kind of home appliance network time synchronization method based on TPSN and TSync, comprises the following steps:

S1. Establish the root node: add a GPS signal receiver to the smart home appliance, and use the smart home appliance node as the root node for time synchronization, set N nodes in the wireless sensor network, and obtain the time of all nodes, GPS time T _G , the network server time T _S , the local root node time T _B , and establish or update the root node.

S2. Determine the source time:

Set the source time to be obtained from (T _G , T _B , T _S ); when T _G is valid, T _G has the highest priority as the source time; when T _G is invalid and T _S is valid, T _S has the highest priority Second only to T _G , as the source time; when T _G is invalid and T _S is invalid, select T _B as the source time.

S3. Select the synchronization method: use continuous synchronization and on-demand synchronization complementary methods for synchronization; continuous synchronization uses TSync algorithm for time synchronization, set to synchronize once within a specified time; on-demand synchronization uses TPSN algorithm for time synchronization, according to the user's free choice Time synchronization.

The method for establishing or updating the root node as described in the preferred step S1 specifically includes the following steps:

S11. The temporary variable is represented by T _temp , T _temp =α×T _S +β×T _G , (α+β)=1.

S12. When both T _S and T _G can be obtained accurately, then T _temp =α×T _S +β×T _G , (α+β)=1, where the weights are distributed such that α=0.3, β=0.7.

When T _S can be obtained but T _G cannot be obtained, T _G =0, α=1, then T _temp =T _S .

When T _G can be obtained but T _S cannot be obtained, T _S =0, β=1, then T _temp =T _G .

When neither T _G nor T _S can be obtained, T _G =0, T _S =0, then T _temp =0.

S13. Obtain the internal clocks {T ₁ , T ₂ , T ₃ ... T _N } of N nodes in the wireless network, and then calculate the minimum value of the difference between each internal clock and T _temp according to formula (1.1):

{abs(T ₁ -T _temp ), abs(T ₂ -T _temp ), abs(T ₃ -T _temp ),... abs(T _N -T _temp )} _min (1.1).

S14. When the value of abs(T _K −T _temp ) is the smallest, select the Kth node as the root node; when T _temp =0, do not update the root node.

As the validity judgment of GPS time T _G in the preferred step S2 specifically comprises the following steps:

—Comparing the error between T _G and T _B and T _S , take the largest value {abs(T _G -T _S ),abs(T _G -T _B )} _max , when the maximum error time is greater than 10min, then consider GPS The receiver fails, and the GPS time T _G is determined to be invalid.

—When the time server fails or the network is interrupted, judge whether the abs(T _G -T _B ) is greater than 10 minutes. When the abs(T _G -T _B ) is greater than 10 minutes, it is considered that the GPS receiver is faulty, and then determine the GPS time T _G invalid.

- Perform five difference calculations on the time value received by GPS in the last 30 minutes,

{T ₁ , T ₂ , T ₃ , T ₄ ... T _max }, where max is equivalent to the number of all time points acquired within 30 minutes, if the GPS signal does not have any time missing, then max=1800, the actual time data stored prevail,

Perform the first difference operation to obtain {T ₁₁ , T ₁₂ , T ₁₃ , T _1i ... T _1(max-1) }, where T _1i =T _i+1 -T _i , i=(1,2 ,3,4...max-1);

Perform the second difference operation, {T ₂₁ , T ₂₂ , T ₂₃ , T _2i ... T _2(max-2) }, where T _2i =T _1i+1 -T _1i , i=(1,2, 3,4...max-2);

By analogy, the k-th difference operation, {T _k1 , T _k2 , T _k3 , T _ki ... T _k(max-k) }, where T _ki =T _(k-1)i+1 -T _{( k-1)i} , i=(1,2,3,4...max-k);

The result of the fifth difference is {T ₅₁ , T ₅₂ , T ₅₃ , T _5i ... T _5(max-5) } _max If it is less than 1s, the time T _G of GPS signal acquisition is considered valid, otherwise give up GPS time T _G as time source.

As the validity judgment of the network server time T _S in the preferred step S2:

When the network is disconnected and the port network time cannot be obtained, the network server time T _S is set to 0, and T _S is invalid.

In the case of networking, the network time is obtained based on the NTP service address, the internal time of the wireless network is determined, and the discovery and establishment of the network level specifically includes the following steps:

S21. The root node sets its own level number to 1.

S22. The root node broadcasts an information packet, which includes the ID and the layer number of the root node.

S23. When a node within the communication range receives the information packet, it sets its own hierarchical level as the hierarchical number in the information packet plus 1.

S24. The receiving node broadcasts a new information packet containing its own ID and layer number, and the receiving node sets its own layer number in the same manner.

S25. Step S24 is repeated until all the nodes in the entire network have established their own hierarchical levels. When the nodes with existing hierarchical numbers receive the broadcasted information packet again, they will be ignored.

S26. Assuming that the last established layer is k layer, the number of nodes in {1,2,3...k} layer is {n ₁ ,n ₂ ,n ₃ ,...n _k }, and the total number of nodes is N, Then calculate the probability of each layer as {p ₁ ,p ₂ ,p ₃ ,...p _k } are respectively

S27. Calculate the expected value of internal node time according to different probabilities, select a node for each layer, so as to avoid the failure of nodes in the layer, and obtain their time respectively as {T _1i , T _2i , T _3i ... T _ki } , in this time subscript {1i≤n ₁ ,2i≤n ₂ ,3i≤n ₃ ,…ki≤n _k }, the calculation time expectation

S27. Calculate the difference between the internal clocks {T ₁ , T ₂ , T ₃ ... T _N } of N nodes in the wireless network and the time expectation E _T , and the time with the smallest error {abs(T ₁ -E _T ), abs (T ₂ -E _T ),abs(T ₃ -E _T ),…abs(T _N -E _T )} _min is used as the source time of the internal clock of the wireless network.

As the TSync algorithm described in the continuous synchronization in the preferred step 3 is to synchronize by HRTS and ITR two algorithms, HRTS is that the synchronization root node sends a synchronization signal to synchronize, ITR is to send a synchronization request to an adjacent node by a common node, and A synchronization method that finally passes the request to the root node. The two complement each other and run at the same time. It belongs to the complementary of active synchronization and request synchronization.

As preferably described HRTS algorithm synchronization specifically comprises the following steps:

S311. The root node broadcasts a sync_begin beacon on the control channel at time t1.

S312. A child node randomly designated by the root node jumps to a designated clock channel to reply, for example, n2.

S313. Node n2 replies to the root node its own receiving time t2 and t3 at time t3.

S314. The root node records the received timestamp t4, so that the root node has all the timestamps of t1-t4.

S315. The root node calculates the time d2, and then broadcasts t2 and d2 to all nodes of the control channel, wherein

S316. All adjacent child nodes, such as n2, n3, n4, and n5, compare the time t2 with the received timestamp t2', for example, n3 calculates the time drift d'd'=t2-t2'.

S317, the time of node n3 is corrected as T=t+d2+d', where t is the local time of node n3;

S318. Nodes n2, n3, n4, and n5 initialize sync_begin beacons to their downstream nodes, and repeat the above steps.

As preferably described ITR algorithm synchronization specifically comprises the following steps:

S321. The node n1 sends an ITR_QUERY signal in the control channel, and the clock channel for sending the synchronization request is specified in the ITR_QUERY.

S322. After the parent node n2 of node n1 receives the request, it will send an ITR_ACK signal to notify its parent node in the control channel. In this example, it is the root node BS; under normal circumstances, the upstream parent node will send the ITR_ACK signal until it reaches the root node. node.

S323. After receiving the ITR_ACK, the parent node BS of the node n2 switches to the specified clock channel, and the clock channel information is included in the ITR_ACK; all nodes along the propagation path of the ITR_ACK switch to the specified clock channel.

S324. Node n2 receives the synchronization request from n1 and sends it to node BS through a specified clock channel.

S325. The node BS initiates the same process to send the time to the node n1.

S326. The node n1 synchronizes its own time according to the time fed back by the node BS.

The on-demand synchronous TPSN algorithm described as continuous synchronization in the preferred step 3 includes a hierarchy discovery stage and a synchronization stage. In the hierarchy discovery stage, each node has its own hierarchy number, and the network structure with hierarchy is regarded as a spanning tree, specifically Include the following steps:

S331. The root node of the tree serves as the clock source node, and its level number is set to 0.

S332. The root node broadcasts an information packet, which includes the ID and the layer number of the root node.

S333. When a node within the communication range receives the information packet, it sets its own hierarchy level to the hierarchy number in the information packet plus 1.

S334. The receiving node broadcasts a new information packet containing its own ID and layer number, and the receiving node sets its own layer number in the same manner.

S335. Step S334 is repeated until all the nodes in the entire network have established their own hierarchical levels. When the nodes with existing hierarchical numbers receive the broadcast information packet again, they ignore it.

The synchronization phase specifically includes the following steps:

S341. The layer numbers of node R and node S are the kth layer and the k+1th layer respectively. During synchronization, the upper layer node R broadcasts a time synchronization request packet to notify S node to prepare for time synchronization.

S342. After waiting for _a random period of time, the node S sends _a synchronization information packet containing the time T1 to the node R at the time T1.

S343. After receiving the information packet, the node R uses the local clock to record the receiving time T ₂ , then: T ₂ =T ₁ +Δ+d, where Δ represents the time offset between nodes, and d represents the transmission time of the message delay.

S344. Node R sends a confirmation message to node S at time T ₃ in the same manner, and the message includes (T ₁ , T ₂ , T ₃ ).

S345. Node S records the time T ₄ of receiving the message in local time at T ₄ , which satisfies: T ₄ =T ₃ -Δ+d, where Δ represents the time offset between nodes, and d represents the transmission delay of the message ; Obtained by the formulas T ₂ =T ₁ +Δ+d and T ₄ =T ₃ -Δ+d:

and

S346. Modify the time of the nodes to be consistent according to the calculation result of step S345.

Compared with the prior art, the home appliance network time synchronization method based on TPSN and TSync described in the present invention, in step S1, does not select a single time source for the determination of the synchronization source time, but integrates GPS signal reception Servers, network servers, and internal clocks of nodes adopt different time sources for different situations, which greatly improves the effectiveness of time synchronization and avoids time blind spots in time synchronization.

In step S2, the five-time difference method is used to determine the validity of GPS time, and the internal time of the wireless network is adopted to first establish the network level, then assign probability values according to the number of nodes in the network level, and then calculate the expected value to obtain A more reliable time.

For the acquisition of the network server time, multiple time server addresses are selected as backups, and the time of the China National Time Service Center is used as the first. After the source time is acquired, different priorities are set, the GPS time has the highest priority, the network server time takes the second place, and the local time has the lowest priority to determine the source time according to different situations.

After determining the source time, in step S3, the user selects a different time synchronization algorithm to decide whether to perform continuous time synchronization or on-demand time synchronization. Continuous time synchronization is a time synchronization method with a shorter synchronization time period, and the time synchronization interval is set It is 0.5 hours or 1 hour. The TSync synchronization algorithm, proposed by Richard Han and HuiDai of the University of Colorado, is a flexible, self-organizing time synchronization service without fixed network topology and transmission delay.

On-demand synchronization is to determine the synchronization time according to the user's needs. There is no certain time interval, and the time synchronization method of the TPSN protocol is adopted for on-demand synchronization.

The adoption of the two time synchronization algorithms is based on the characteristics of the two algorithms. Combining the advantages of the two algorithms, the TSync synchronization algorithm is a lightweight algorithm with fast convergence and short time, suitable for continuous synchronization. The TPSN algorithm has slow speed and high reliability, and is suitable for on-demand synchronization.

The home appliance network time synchronization method based on TPSN and TSync in the present invention is a time synchronization method using different wireless network nodes for different situations.

Description of drawings

The present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments.

Fig. 1 is an overall flowchart of the time synchronization of the present invention.

Fig. 2 is a flowchart of root node establishment and update in the present invention.

Fig. 3 is a flowchart of source time determination in the present invention.

Fig. 4 is a flow chart of the method for judging the validity of the GPS clock in the present invention.

Fig. 5 is a flow chart of determining the internal source time of the network in the present invention.

FIG. 6 is a flowchart of the HRTS stage of the TSync algorithm of the present invention.

FIG. 7 is a flow chart of the ITR stage of the TSync algorithm of the present invention.

Fig. 8 is a schematic diagram of the calculation principle of d2 in the present invention.

Fig. 9 is a flow chart of the discovery phase of the TPSN algorithm level of the present invention.

Fig. 10 is a schematic diagram of the synchronization phase of the TPSN algorithm of the present invention.

Detailed ways

As shown in Figure 1, a home appliance network time synchronization method based on TPSN and TSync includes the following steps:

S1. Establish the root node: add a GPS signal receiver to the smart home appliance (such as an air conditioner), and use the smart home appliance node as the root node for time synchronization, set N nodes in the wireless sensor network, and obtain the time of all nodes , GPS time T _G , network server time T _S , local root node time T _B , and establish or update the root node.

As shown in Figure 2, the method for establishing or updating the root node described in step S1 specifically includes the following steps:

S11. The temporary variable is represented by T _temp , and this time variable is used as the benchmark for comparing the time of the internal nodes of the network. This variable is closely related to the GPS time and the network server time, so the time is set as the GPS time and the network server time. The time after the weighted calculation is added, and the weight coefficients are α and β respectively. Since the reference time must be close to the GPS time and the network server time, the sum of the weight coefficients is 1.

Temporary variable T _temp =α×T _S +β×T _G , (α+β)=1.

S12. When both T _S and T _G can be obtained accurately, then T _temp =α×T _S +β×T _G , (α+β)=1, where the weights are distributed such that α=0.3, β=0.7.

When T _S can be obtained but T _G cannot be obtained, T _G =0, α=1, then T _temp =T _S .

When T _G can be obtained but T _S cannot be obtained, T _S =0, β=1, then T _temp =T _G .

When neither T _G nor T _S can be obtained, T _G =0, T _S =0, then T _temp =0.

S13. Obtain the internal clocks {T ₁ , T ₂ , T ₃ ... T _N } of N nodes in the wireless network, and then calculate the minimum value of the difference between each internal clock and T _temp according to formula (1.1):

{abs(T ₁ -T _temp ), abs(T ₂ -T _temp ), abs(T ₃ -T _temp ),... abs(T _N -T _temp )} _min (1.1).

S14. When the value of abs(T _K −T _temp ) is the smallest, select the Kth node as the root node; when T _temp =0, do not update the root node.

In step S1, for the determination of the synchronization source time, a single time source is not selected, but a combination of the GPS signal receiver, the network server and the internal clock of the node, and different time sources are adopted for different situations, thereby making time synchronization The effectiveness is greatly improved, and the time blind zone of time synchronization is avoided.

S2. Determine the source time:

As shown in Figure 3, the set source time is obtained from (T _G , T _B , T _S );

When T _G is valid, T _G has the highest priority and is used as the source time.

As shown in Figure 4, the validity determination of the GPS time T _G in step S2 specifically includes the following steps:

—Comparing the error between T _G and T _B and T _S , take the largest value {abs(T _G -T _S ),abs(T _G -T _B )} _max , when the maximum error time is greater than 10min, then consider GPS The receiver fails, and the GPS time T _G is determined to be invalid.

—When the time server fails or the network is interrupted, judge whether the abs(T _G -T _B ) is greater than 10 minutes. When the abs(T _G -T _B ) is greater than 10 minutes, it is considered that the GPS receiver is faulty, and then determine the GPS time T _G invalid.

- Perform five difference calculations on the time value received by GPS in the last 30 minutes,

{T ₁ , T ₂ , T ₃ , T ₄ ... T _max }, where max is equivalent to the number of all time points acquired within 30 minutes, if the GPS signal does not have any time missing, then max=1800, the actual time data stored prevail,

Perform the first difference operation to obtain {T ₁₁ , T ₁₂ , T ₁₃ , T _1i ... T _1(max-1) }, where T _1i =T _i+1 -T _i , i=(1,2 ,3,4...max-1);

Perform the second difference operation, {T ₂₁ , T ₂₂ , T ₂₃ , T _2i ... T _2(max-2) }, where T _2i =T _1i+1 -T _1i , i=(1,2, 3,4...max-2);

By analogy, the k-th difference operation, {T _k1 , T _k2 , T _k3 , T _ki ... T _k(max-k) }, where T _ki =T _(k-1)i+1 -T _{( k-1)i} , i=(1,2,3,4...max-k);

The result of the fifth difference is {T ₅₁ , T ₅₂ , T ₅₃ , T _5i ... T _5(max-5) } _max If it is less than 1s, the time T _G of GPS signal acquisition is considered valid, otherwise give up GPS time T _G as time source.

When T _G is invalid and T _S is valid, the priority of T _S is second only to T _G as the source time.

Validity determination of network server time T _S in step S2:

When the network is disconnected and the port network time cannot be obtained, the network server time T _S is set to 0, and T _S is invalid.

When connected to the Internet, the network time server selects:

Usually get the network time with the following NTP service address

server ntp-sop.inria.frserver 210.72.145.44 (China National Time Service Center server IP address)

server 133.100.11.8prefer

server 210.72.145.44

server 203.117.180.36

server 131.107.1.10

server64.236.96.53

server 130.149.17.21

server66.92.68.246

serverwww.freebsd.org

server 18.145.0.30

server clock.via.net

server 137.92.140.80

server 133.100.9.2

server 128.118.46.3

server ntp.nasa.gov

server 129.7.1.66

The acquisition of network time is first obtained from the China National Time Service Center, and then obtained from other addresses.

The network time is obtained based on the NTP service address, and the internal time of the wireless network is determined, as shown in Figure 5. The discovery and establishment of the network level, as shown in Figure 9, specifically includes the following steps:

S21. The root node sets its own level number to 1.

S22. The root node broadcasts an information packet, which includes the ID and the layer number of the root node.

S23. When a node within the communication range receives the information packet, it sets its own hierarchical level as the hierarchical number in the information packet plus 1.

S24. The receiving node broadcasts a new information packet containing its own ID and layer number, and the receiving node sets its own layer number in the same manner.

S25. Step S24 is repeated until all the nodes in the entire network have established their own hierarchical levels. When the nodes with existing hierarchical numbers receive the broadcasted information packet again, they will be ignored.

S26. Assuming that the last established layer is k layer, the number of nodes in {1,2,3...k} layer is {n ₁ ,n ₂ ,n ₃ ,...n _k }, and the total number of nodes is N, Then calculate the probability of each layer and set it as {p ₁ ,p ₂ ,p ₃ ,…p _k } respectively as

S27. Calculate the expected value of internal node time according to different probabilities, select a node for each layer, so as to avoid the failure of nodes in the layer, and obtain their time respectively as {T _1i , T _2i , T _3i ... T _ki } , in this time subscript {1i≤n ₁ ,2i≤n ₂ ,3i≤n ₃ ,…ki≤n _k }, the calculation time expectation

S27. Calculate the difference between the internal clocks {T ₁ , T ₂ , T ₃ ... T _N } of N nodes in the wireless network and the time expectation E _T , and the time with the smallest error {abs(T ₁ -E _T ), abs (T ₂ -E _T ),abs(T ₃ -E _T ),…abs(T _N -E _T )} _min is used as the source time of the internal clock of the wireless network.

When T _G is invalid and T _S is invalid, T _B is selected as the source time, and T _B is the local root node time, that is, the internal time of the wireless network.

In step S2, the five-time difference method is used to determine the validity of GPS time, and the internal time of the wireless network is adopted to first establish the network level, then assign probability values according to the number of nodes in the network level, and then calculate the expected value to obtain A more reliable time.

For the acquisition of the network server time, multiple time server addresses are selected as backups, and the time of the China National Time Service Center is used as the first. After the source time is acquired, different priorities are set, the GPS time has the highest priority, the network server time takes the second place, and the local time has the lowest priority to determine the source time according to different situations.

S3. Selecting a synchronization method: performing synchronization in a complementary manner of continuous synchronization and on-demand synchronization.

Continuous synchronization uses the TSync algorithm for time synchronization. Continuous synchronization means that the synchronization time interval is relatively short. It is set to synchronize once within a specified time, usually set to synchronize once every 0.5 hours or 1 hour, and uses the TSync algorithm to synchronize the time of the entire node network.

On-demand synchronization uses the TPSN algorithm for time synchronization. On-demand synchronization is a synchronization method independent of continuous synchronization. It is a synchronization method for time synchronization according to the user's free choice, and on-demand synchronization does not affect the continuous synchronization algorithm. running.

The TSync algorithm described in the continuous synchronization in step 3 is synchronized through two algorithms, HRTS and ITR. HRTS is a synchronization root node that sends a synchronization signal for synchronization. ITR sends a synchronization request to an adjacent node through an ordinary node, and finally sends A synchronization method in which the request is passed to the root node. The two complement each other and run at the same time. It belongs to the complementary of active synchronization and request synchronization.

In some cases, the active synchronization scheme based on the HRTS algorithm cannot successfully synchronize too many nodes. Therefore, as a supplement to HRTS, ITR provides a way for each sensor node to independently obtain time and communicate with the surrounding A demand-based time synchronization method for environment synchronization.

As shown in Figure 6, the HRTS algorithm synchronization specifically includes the following steps:

S311. The root node broadcasts a sync_begin beacon on the control channel at time t1, as shown in FIG. 6-(a).

S312. A child node randomly designated by the root node jumps to a designated clock channel to reply, for example, n2.

S313. Node n2 replies to the root node its own receiving time t2 and t3 at time t3, as shown in Figure 6-(b).

S314. The root node records the received timestamp t4, so that the root node has all the timestamps of t1-t4.

S315. The root node calculates time d2, parameters involved in the calculation principle of d2, as shown in Figure 8, and then broadcasts t2, d2 to all nodes of the control channel, as shown in Figure 6-(c), wherein

S316. All adjacent child nodes, such as n2, n3, n4, and n5, compare the time t2 with the received timestamp t2', for example, n3 calculates the time drift d'd'=t2-t2'.

S317. The time of node n3 is corrected as T=t+d2+d′, where t is the local time of node n3.

S318. Nodes n2, n3, n4, and n5 initialize sync_begin beacons to their downstream nodes, and repeat the above steps.

As shown in Figure 7, the ITR algorithm synchronization specifically includes the following steps:

S321. The node n1 sends an ITR_QUERY signal in the control channel, and the clock channel for sending the synchronization request is specified in the ITR_QUERY, as shown in FIG. 7-(a).

S322. After the parent node n2 of node n1 receives the request, it will send an ITR_ACK signal to notify its parent node on the control channel. In this example, the root node BS in Figure 7; under normal circumstances, the upstream parent node will always send the ITR_ACK signal until it reaches Root node, as shown in Figure 7-(b).

S323. After receiving the ITR_ACK, the parent node BS of the node n2 switches to the specified clock channel, and the clock channel information is included in the ITR_ACK; all nodes along the propagation path of the ITR_ACK switch to the specified clock channel.

S324. Node n2 receives the synchronization request from n1 and sends it to node BS through a specified clock channel, as shown in FIG. 7-(c).

S325. The node BS initiates the same process to send the time to the node n1.

S326. The node n1 synchronizes its own time according to the time fed back by the node BS.

As shown in Figure 9, the on-demand synchronous TPSN algorithm described in continuous synchronization in step 3 includes a hierarchy discovery phase and a synchronization phase; in the hierarchy discovery phase, each node has its own hierarchy number, and the network structure with hierarchy is regarded as Generate a tree, specifically including the following steps:

S331. The root node of the tree serves as the clock source node, and its level number is set to 0.

S332. The root node broadcasts an information packet, which includes the ID and the layer number of the root node.

S333. When a node within the communication range receives the information packet, it sets its own hierarchy level to the hierarchy number in the information packet plus 1.

S334. The receiving node broadcasts a new information packet containing its own ID and layer number, and the receiving node sets its own layer number in the same manner.

S335. Step S334 is repeated until all the nodes in the entire network have established their own hierarchical levels. When the nodes with existing hierarchical numbers receive the broadcasted information packet again, they ignore the processing to realize the establishment of the TPSN topology.

As shown in Figure 10, the synchronization phase specifically includes the following steps:

S341. The layer numbers of node R and node S are the kth layer and the k+1th layer respectively. During synchronization, the upper layer node R broadcasts a time synchronization request packet to notify S node to prepare for time synchronization.

S342. After waiting for _a random period of time, the node S sends _a synchronization information packet containing the time T1 to the node R at the time T1.

S343. After receiving the information packet, the node R uses the local clock to record the receiving time T ₂ , then: T ₂ =T ₁ +Δ+d, where Δ represents the time offset between nodes, and d represents the transmission time of the message delay.

S344. Node R sends a confirmation message to node S at time T ₃ in the same manner, and the message includes (T ₁ , T ₂ , T ₃ ).

S345. Node S records the time T ₄ of receiving the message in local time at T ₄ , which satisfies: T ₄ =T ₃ -Δ+d, where Δ represents the time offset between nodes, and d represents the transmission delay of the message ; Obtained by the formulas T ₂ =T ₁ +Δ+d and T ₄ =T ₃ -Δ+d:

and

S346. Modify the time of the nodes to be consistent according to the calculation result of step S345.

The time synchronization process in the present invention should be carried out according to the flow in the specific steps, and the priority of each time synchronization method must not be changed.

After determining the source time, in step S3, the user selects a different time synchronization algorithm to decide whether to perform continuous time synchronization or on-demand time synchronization. Continuous time synchronization is a time synchronization method with a shorter synchronization time period, and the time synchronization interval is set It is 0.5 hours or 1 hour. The TSync synchronization algorithm was proposed by Richard Han and HuiDai of the University of Colorado. It is a flexible, self-organizing time synchronization service without fixed network topology and transmission delay.

On-demand synchronization is to determine the synchronization time according to the user's needs. There is no certain time interval, and the time synchronization method of the TPSN protocol is adopted for on-demand synchronization.

The adoption of the two time synchronization algorithms is based on the characteristics of the two algorithms. Combining the advantages of the two algorithms, the TSync synchronization algorithm is a lightweight algorithm with fast convergence and short time, suitable for continuous synchronization. The TPSN algorithm has slow speed and high reliability, and is suitable for on-demand synchronization.

The present invention realizes the home appliance wireless network time synchronization method based on the TPSN method and TSync hybrid time synchronization, comprehensively utilizes three time sources of external GPS clock, network time server, and smart home appliance internal clock, and adopts different time synchronization methods for different situations The method avoids the occurrence of a time synchronization blank time period, and ensures the reliability and accuracy of the monitoring data and time data of smart home appliances.

The GPS clock source of the present invention is a GPS signal receiving device that can be installed outdoors, and indoor GPS signals cannot be used as a clock signal source.

The above is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto, any person familiar with the technical field within the technical scope disclosed in the present invention, according to the technical solution of the present invention Any equivalent replacement or change of the inventive concepts thereof shall fall within the protection scope of the present invention.

Claims (6)

1. A household electrical appliance network time synchronization method based on TPSN and TSync is characterized by comprising the following steps:

s1, establishing a root node: the method comprises the steps of adding a GPS signal receiver in an intelligent household appliance, using the node of the intelligent household appliance as a root node of time synchronization, setting N nodes in a wireless sensing network, and acquiring the time of all the nodes, namely the GPS time T_GNetwork Server time T_SLocal root node time T_BAnd establishing or updating a root node;

s2, determining source time:

setting source time from (T)_G,T_B,T_S) Obtaining the three;

when T is_GEffective, then T_GThe highest priority is used as the source time;

when T is_GNull, T_SEffective, then T_SPriority next to T_GAs a source time;

when T is_GNull, T_SIf not, T is selected_BAs a source time;

s3, selecting a synchronization mode:

the method comprises the following steps of (1) synchronizing in a continuous synchronization and on-demand synchronization complementary mode;

the continuous synchronization adopts a TSync algorithm to carry out time synchronization, and is set to be synchronized once within a specified time;

the TPSN algorithm is adopted for time synchronization according to the requirement, and the time synchronization is carried out according to the free selection of a user;

GPS time T in step S2_GThe validity judgment specifically includes the following steps:

-comparing T_GAnd T_BAnd T_SThe error between them is the largest { abs (T)_G-T_S),abs(T_G-T_B)}_maxWhen the maximum error time is more than 10min, the GPS receiver is determined to be in fault, and then the GPS time T is determined_GInvalid;

-judging abs (T) when the time server fails or the network is interrupted_G-T_B) Whether it is greater than 10min, when abs (T)_G-T_B) If the time is more than 10min, the GPS receiver is considered to have a fault, and the GPS time T is further determined_GInvalid;

-five difference operations on the time value received by the GPS at the latest 30min time, { T₁,T₂,T₃,T₄……T_maxAnd obtaining { T (T) } by performing a first difference operation with reference to the stored time data, wherein max is equal to the number of all time points obtained within 30min, and is 1800 if no time is lost in the GPS signal₁₁,T₁₂,T₁₃,T_1i……T_1(max-1)Where T is_1i＝T_i+1-T_i，i＝(1,2,3,4……max-1)；

Performing a second difference operation, { T₂₁,T₂₂,T₂₃,T_2i……T_2(max-2)Where T is_2i＝T_1i+1-T_1i，i＝(1,2,3,4……max-2)；

By analogy, the k-th difference operation, { T }_k1,T_k2,T_k3,T_ki……T_k(max-k)Where T is_ki＝T_(k-1)i+1-T_(k-1)i，i＝(1,2,3,4……max-k)；

Proceed to the result of the fifth difference T₅₁,T₅₂,T₅₃,T_5i……T_5(max-5)}_maxIf the time is less than 1s, the time T of GPS signal acquisition is considered_GEffective, otherwise abandon GPS time T_GAs a source of time;

web server time T in step S2_SThe validity judgment of (1):

when the port network time can not be obtained under the condition of network disconnection, the network server time T_SIs set to 0, T_SInvalid;

in the case of networking, network time is acquired based on the NTP service address,

the determination of the time inside the wireless network,

the discovery and establishment of the network layer specifically comprises the following steps:

s21, setting the own layer number as 1 by the root node;

s22, broadcasting an information packet by the root node, wherein the information packet comprises the ID and the layer number of the root node;

s23, when the node in the communication range receives the information packet, setting the level of the node as the level number plus 1 in the information packet;

s24, broadcasting a new information packet containing the ID and the layer number of the receiving node by the receiving node, and setting the layer number of the receiving node in the same way;

s25, repeating the step S24 until the nodes in the whole network establish the hierarchy level of the nodes, wherein the nodes with the hierarchy numbers do neglect processing when receiving the broadcasted information packet again;

s26 falseLet the last established hierarchy be k-level and the number of nodes in {1,2,3 … … k } level be { n₁,n₂,n₃,……n_kAnd if the total node number is N, calculating the probability of each layer and setting the probability as p₁,p₂,p₃,……p_kAre respectively

S27, calculating expected value of internal node time according to different probabilities, selecting one node in each layer, and obtaining time of the nodes as { T }_1i,T_2i,T_3i……T_kiIn this time index {1i ≦ n₁,2i≤n₂,3i≤n₃,……ki≤n_kCalculate time expectation

S27 internal clocks { T } of N nodes in wireless network₁,T₂,T₃……T_NAnd time expectation E_TDifference calculation is performed, and the time of minimum error { abs (T) }₁-E_T),abs(T₂-E_T),abs(T₃-E_T),……abs(T_N-E_T)}_minAs the source time of the clock internal to the wireless network.

2. The home appliance network time synchronization method based on TPSN and TSync according to claim 1, wherein:

the method for establishing or updating the root node in step S1 specifically includes the following steps:

s11, T for temporary variables_tempDenotes, T_temp＝α×T_S+β×T_G,(α+β)＝1；

S12, when T_SAnd T_GCan be accurately obtained, then T_temp＝α×T_S+β×T_G(α + β) ═ 1, where the weights are assigned such that α ═ 0.3 and β ═ 0.7;

when T is_SCan obtainTo obtain, and T_GIs not obtained, then T_G0, α is 1, then T_temp＝T_S；

When T is_GCan obtain, T_SIs not obtained, then T_S0, β 1, then T_temp＝T_G；

When T is_GAnd T_SAre all unavailable, so that T_G＝0,T_SWhen the value is 0, then T_temp＝0；

S13, obtaining internal clocks { T ] of N nodes in wireless network₁,T₂,T₃……T_NThen calculate each internal clock and T according to equation (1.1)_tempMinimum of difference of (d):

{abs(T₁-T_temp),abs(T₂-T_temp),abs(T₃-T_temp),……abs(T_N-T_temp)}_min(1.1)；

s14, as abs (T)_K-T_temp) If the value is minimum, selecting the Kth node as a root node;

when T is_tempIf 0, the root node is not updated.

3. The home appliance network time synchronization method based on TPSN and TSync according to claim 1, wherein:

the TSync algorithm in step 3 is synchronized by two algorithms of HRTS and ITR, the HRTS is synchronized by a synchronization signal sent by a synchronization root node, the ITR is a synchronization mode of sending a synchronization request to an adjacent node by a common node and finally transmitting the request to the root node, the two are mutually complemented and run simultaneously, and the method belongs to active synchronization and request synchronization complementation.

4. The home appliance network time synchronization method based on TPSN and TSync according to claim 3, wherein:

the HRTS algorithm synchronization specifically comprises the following steps:

s311, the root node broadcasts a sync _ begin beacon at time t1 on the control channel;

s312, a child node randomly designated by the root node jumps to a designated clock channel for replying, for example, n 2;

s313, the node n2 replies to the root node at the time t3 with the own receiving times t2 and t 3;

s314, recording the received time stamp t4 by the root node, so that the root node has all time stamps from t1 to t 4;

s315, calculating time d2 by the root node, broadcasting t2 and d2 to all nodes of the control channel, calculating method of d2, formula

S316, all neighboring child nodes, e.g., n2, n3, n4, n5, compare time t2 and receive timestamp t2', e.g., n3 calculates time drift d'd ═ t2-t2 ';

the time of the node n3 is corrected to be T ═ T + d2+ d', wherein T is the local time of the n3 node;

s318, nodes n2, n3, n4, n5 initialize sync _ begin beacons to their downstream nodes, and the above steps are repeated.

5. The home appliance network time synchronization method based on TPSN and TSync according to claim 3 or 4, wherein:

the ITR algorithm synchronization specifically comprises the following steps:

s321, the node n1 sends an ITR _ QUERY signal in the control channel, and a clock channel for sending a synchronization request is designated in the ITR _ QUERY;

s322, after receiving the request, the father node n2 of the node n1 sends an ITR _ ACK signal in the control channel to inform the father node thereof, and the upstream father node sends the ITR _ ACK signal until reaching the root node;

s323, a father node BS of the node n2 receives the ITR _ ACK, and then converts the ITR _ ACK into a specified clock channel, wherein the clock channel information is contained in the ITR _ ACK; all nodes along the ITR _ ACK propagation path switch to the designated clock channel;

s324, the node n2 receives the synchronization request of n1 and sends the synchronization request to the node BS through a specified clock channel;

s325, the node BS initiates the same process to send time to the node n 1;

s326, node n1 synchronizes its time according to the time fed back by node BS.

6. The home appliance network time synchronization method based on TPSN and TSync according to claim 1, wherein:

the TPSN algorithm which is continuously synchronized according to needs in the step 3 comprises a level discovery stage and a synchronization stage;

in the level discovery stage, each node has its own level number, and the network structure with levels is regarded as a spanning tree, which specifically comprises the following steps:

s331, enabling a root node of the tree to serve as a clock source node, and setting a hierarchy number of the root node of the tree to be 0;

s332, broadcasting an information packet by the root node, wherein the information packet comprises the ID and the layer number of the root node;

s333, when the nodes in the communication range receive the information packet, setting the level of the nodes as the level number plus 1 in the information packet;

s334, broadcasting a new information packet containing the ID and the layer number of the receiving node by the receiving node, and setting the layer number of the receiving node in the same way;

s335, repeating the step S334 until the nodes in the whole network establish the hierarchy levels of the nodes, and when the nodes with the hierarchy numbers receive the broadcasted information packet again, performing ignoring processing;

the synchronization stage specifically comprises the following steps:

s341, the hierarchy numbers of the node R and the node S are respectively the kth layer and the (k + 1) th layer, and when the synchronization is performed, the upper node R broadcasts a time synchronization request information packet to inform the node S of preparing the time synchronization;

s342, after the node S waits for a period of random time, the node S waits for T₁Sending time to node R with time T₁The synchronization packet of (2);

s343, after the node R receives the information packet, the local clock is used for recording the receiving time T₂Then, there are:T₂＝T₁+ Δ + d, where Δ represents the time offset between nodes and d represents the transmission delay of a message;

s344, the node R at the time T in the same way₃Sending an acknowledgement message to node S, the message including (T)₁、T₂、T₃)；

S345, node S is at T₄The time of receiving the message T is recorded by local time₄And satisfies the following conditions: t is₄＝T₃- Δ + d, where Δ represents the time offset between nodes and d represents the transmission delay of the message; from the formula T₂＝T₁+ Δ + d and T₄＝T₃- Δ + d yields:

and

s346, according to the calculation result of the step S345, the time of modifying the nodes is consistent.