Age-of-information optimization scheduling method for industrial wireless network and based on multi-dimensional conflict graph

Abstract

The present invention belongs to the technical field of industrial wireless networks, and particularly relates to an age-of-information optimization scheduling method for an industrial wireless network and based on a multi-dimensional conflict graph, the method comprising: compiling statistics on age-of-information changes of each data stream in each superframe, and on the basis of an average expected age-of-information calculation formula for the data stream, modeling the problem of minimizing the overall average expected age of information into an integer programming model for time slot and channel allocation; constructing a three-dimensional conflict graph composed of a data-stream–link–time-slot three-dimensional tuple, converting a time-slot scheduling problem into a maximum weight independent set problem, and solving same to obtain a time-slot scheduling strategy; and on the basis of the time-slot scheduling strategy, adding a channel dimension to the data-stream–link–time-slot three-dimensional tuple to form a four-dimensional tuple and constructing a four-dimensional conflict graph, converting a channel allocation problem into a maximum weight independent set problem for solving, and solving same to obtain a channel allocation strategy. The optimization scheduling algorithm used in the present invention can reduce the age of information of a data stream and improve the real-time performance of a network.

Description

An industrial wireless network information age optimization scheduling method based on multi-dimensional conflict graph TECHNICAL FIELD

[0001] The application belongs to the technical field of industrial wireless networks, and particularly relates to an industrial wireless network information age optimization scheduling method based on a multi-dimensional conflict graph. BACKGROUND

[0002] In an industrial wireless sensor network, the requirements for real-time performance and reliability of data transmission are much higher than those of wired networks, which directly relates to production efficiency, quality, and industrial safety and cost. In recent years, a new indicator, information age, has been introduced, which represents the time interval from data generation to reception and is used to measure the freshness of data. By optimizing the information age, the real-time performance of industrial wireless networks in data transmission can be significantly improved, ensuring that data is timely and accurate to serve various industrial applications.

[0003] In an industrial wireless sensor network, time slot and channel scheduling are key elements to ensure real-time performance and reliability of data transmission, and optimization of information age as an indicator of data freshness has become an important research direction. However, existing time slot and channel scheduling algorithms still have some shortcomings in information age optimization:

[0004] 1) Current research on information age optimization of industrial wireless networks focuses mainly on single-hop networks, and there is a lack of research on information age optimization scheduling methods for multi-hop networks;

[0005] 2) Existing scheduling algorithms often focus on the transmission of a single data packet or node, without fully considering the overall network demand, although they can meet the local transmission demand, the information age optimization effect of the entire network is limited.

[0006] Therefore, in the application scenario of multi-hop networks, it is particularly necessary to study the optimization scheduling algorithm targeting at reducing the information age of data flow to ensure the real-time performance of data flow.

[0007] SUMMARY

[0008] In order to overcome the shortcomings of existing scheduling algorithms, and to ensure that the time slots and channels allocated to each link of each data flow in the industrial wireless sensor network are conflict-free, the application proposes an industrial wireless network information age optimization scheduling method based on a multi-dimensional conflict graph, with the goal of optimizing the average information age of the network and improving the real-time performance of the network data flow end-to-end, which specifically includes the following steps:

[0009] S1, statistics of the information age change of each data flow in each superframe, and based on the average expected information age calculation formula of the data flow, the problem of minimizing the overall average expected information age is modeled into an integer programming model of time slot and channel allocation;

[0010] S2, construct a three-dimensional conflict graph composed of data flow-link-time slot three-dimensional tuples, convert the time slot scheduling problem into a maximum weight independent set problem, and solve to obtain a time slot scheduling strategy;

[0011] S3, on the basis of the time slot scheduling strategy, add the channel dimension to the data flow-link-time slot three-dimensional tuple to form a four-dimensional tuple and construct a four-dimensional conflict graph, convert the channel allocation problem into a maximum weight independent set problem for solving, and solve to obtain a channel allocation strategy.

[0012] Further, the average expected information age calculation formula of the data flow is expressed as:

[0013] Wherein, E[A f (t k )] represents the average information age of the data flow f in the kth superframe [kT, (k+1)T]; A f (t k ) represents the average information age of the data flow f when successfully receiving the data packet in the kth superframe [kT, (k+1)T]; δ f represents the probability that the data packet of the data flow f can be successfully received by the destination node;

[0014] Α f ′(t k ) represents the average information age of the data flow f when not receiving the data packet in the kth superframe [kT, (k+1)T]; t k represents the time when the source node generates the data packet in the kth superframe; t k ′ represents the time when the data packet arrives at the destination node in the kth superframe; t k+1 represents the time when the source node generates the data packet in the k+1th superframe; a f (t k ) represents the information age function of the data flow f at tk; Δ is an intermediate parameter, which is expressed as T is the duration of a superframe.

[0015] Further, when modeling the minimum overall average expected information age problem into an integer programming model of time slot and channel allocation, the negative term and the constant term in the minimum overall average expected information age problem are removed, so that the minimum problem is converted into a maximum problem, and the objective function after conversion is expressed as:

[0016] Wherein, F is a data flow set composed of all to-be-scheduled data flows.

[0017] Further, the information age function of the data flow f is expressed as:

[0018] where a f (t) denotes the information age of data stream f at time t; denotes the generation time of the latest data packet of data stream f source node at time t.

[0019] Further, the process of constructing the three-dimensional conflict graph includes:

[0020] A three-dimensional tuple consisting of data stream-link-time slot is constructed to describe the time slot allocation of a data stream f, which is denoted as: τi(f) denotes the i-th hop link of data stream f The allocated time slot is t, F is the set of all data streams, L f is the link set of data stream f, and T is the duration of a superframe;

[0021] Each three-dimensional tuple τ is taken as a node to construct a three-dimensional conflict graph G s (V s , E s ) for time slot scheduling, where V s represents the node set in the three-dimensional conflict graph, each node is a three-dimensional tuple, E s is the edge relationship set in the three-dimensional conflict graph, and there is an edge relationship between two nodes if there is a conflict between them.

[0022] The conflict between communication devices includes: when the communication devices equipped with half-duplex radio frequency modules are allocated the same time slot and the corresponding links have a common communication device, there is a radio frequency conflict between the common communication devices; if the same link in the same data stream is repeatedly scheduled, there is a link repeated scheduling conflict between the two scheduling; if the scheduling order of the link is not the order from the source node to the destination node, there is a link order conflict between the two communication devices.

[0023] Further, the time slot scheduling problem is converted into a maximum weight independent set problem, that is, a maximum weight independent set is found in the three-dimensional conflict graph, which ensures that the maximum independent set elements of the three-dimensional conflict graph are equal to the total number of link hops in the data stream set, and the maximum weight independent set problem is denoted as:

[0024] Constraint condition:

[0025] where N v represents the number of nodes in the three-dimensional conflict graph; t k+1 denotes the time when the source node generates a data packet in the k+1th superframe; t(τ i ) is the time slot item of the i-th node τ i ; λ i is a binary variable, and if the i-th node τi In its maximum independent set, λ i = 1, otherwise λ i = 0; L f represents the link set of data flow f; F represents the data flow set of all data flows to be scheduled; |·| represents the number of elements in the set.

[0026] Further, the process of converting the time slot scheduling problem into the maximum weight independent set problem and solving it includes:

[0027] 201, initialize the iteration number i = 0, and obtain the three-dimensional conflict graph G i = G s (V s , E s );

[0028] 202, if the edge set E(G i ) of the three-dimensional conflict graph G i is empty, output the node set in the current conflict graph as the maximum independent set; otherwise, execute step 203;

[0029] 203, calculate the degree of each node in the three-dimensional conflict graph G i , and select to delete a node and the edges connected to the node according to the degree of the node and the weight value of the node, and the process of selecting the node is represented as:

[0030] Where η i represents the node selected to be deleted in the i-th iteration process; V(G i ) represents the set of all nodes in the three-dimensional conflict graph G i in the i-th iteration; W(μ) is the weight value of node μ in the three-dimensional conflict graph, which is represented as represents the degree of node μ in the three-dimensional conflict graph G i in the i-th iteration;

[0031] 204, update the edge relationship set and the node set of the three-dimensional conflict graph after deleting the node η i , and let i = i + 1, return to step 202.

[0032] Further, the process of converting the channel allocation problem into the maximum weight independent set problem to solve it includes:

[0033] 301, adopt a greedy strategy to decompose the overall channel allocation problem into a hop-by-hop scheduling sub-problem, and solve it hop by hop from the source node to the destination node;

[0034] 302, convert the hop-by-hop scheduling sub-problem into a maximum weight independent set problem, and solve the problem to obtain the channel allocation strategy in one hop.

[0035] Further, after taking the greedy strategy to decompose the whole channel allocation problem into the sub-problems of hop-by-hop scheduling, the sub-problem of the i-th hop scheduling is expressed as:

[0036] Wherein, P i represents the sub-problem of the i-th hop scheduling of channel allocation; F i represents the set of data flows whose hop number is greater than or equal to i in the set of data flows F, and is expressed as represents the time slot allocated to the i-th hop link of the data flow f; represents the link whose transmission success probability is represents the k-th hop link of the data flow f; represents the link whose transmission success probability is represents the i-th hop link of the data flow f.

[0037] Further, after converting the sub-problem of hop-by-hop scheduling into the problem of solving the maximum weight independent set, the sub-problem of the i-th hop scheduling is converted into the problem of solving the maximum weight independent set and is expressed as:

[0038] Constraint condition:

[0039] Wherein, P i ' represents the problem of solving the maximum weight independent set obtained by converting the sub-problem of the i-th hop scheduling of channel allocation; is the number of nodes in the four-dimensional conflict graph corresponding to the sub-problem of the i-th hop scheduling of channel allocation; represents the link whose transmission success probability is represents the j-th hop link of the data flow f performed on the k-th node μ k ; r(l(μ k )) represents the transmission success probability of the link l(μ k ); l(μ k ) represents the link performed on the k-th node μ k ; γ k is a binary variable, and when the k-th node μ k in the four-dimensional conflict graph is in the maximum independent set, γ k = 1, otherwise γ k = 0.

[0040] Compared with the prior art, the present application has the following beneficial effects:

[0041] 1. The present application is aimed at data flow in a multi-hop network, by counting the information age change of data flow in a superframe and calculating its average expected information age, and then converting the overall average expected information age scheduling problem of optimizing industrial wireless network into an integer programming model, so as to facilitate the optimization of the average information age of the network and the scheduling of the network data flow.

[0042] 2. The present application solves the time slot scheduling and channel allocation problem in stages, and completes scheduling by decomposing multi-dimensional conflict graph by link, so as to not only save computing resources, but also effectively improve the real-time performance of industrial wireless network in data transmission.

[0043] In summary, due to the optimization scheduling algorithm adopted by the present application, the information age of data flow can be reduced and the real-time performance of the network can be improved within a certain time. BRIEF DESCRIPTION OF DRAWINGS

[0044] Fig. 1 is a flowchart of the algorithm of the present application;

[0045] Fig. 2 is a network topology diagram of the algorithm verification case of the present application;

[0046] Fig. 3 is a simulation scheduling effect diagram of the verification case of the present application. DETAILED DESCRIPTION

[0047] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0048] The present application proposes an industrial wireless network information age optimization scheduling method based on multi-dimensional conflict graph, specifically comprising the following steps:

[0049] S1, count the information age change of each data flow in each superframe, and based on the average expected information age calculation formula of data flow, model the minimum overall average expected information age problem into an integer programming model of time slot and channel allocation;

[0050] S2, construct a three-dimensional conflict graph composed of data flow-link-time slot three-dimensional tuples, convert the time slot scheduling problem into a maximum weight independent set problem, and solve to obtain the time slot scheduling strategy;

[0051] S3, on the basis of the time slot scheduling strategy, add the channel dimension to the data flow-link-time slot three-dimensional tuple to form a four-dimensional tuple and construct a four-dimensional conflict graph, convert the channel allocation problem into a maximum weight independent set problem for solving, and solve to obtain the channel allocation strategy.

[0052] This embodiment provides a method for optimizing the scheduling of information age in industrial wireless networks based on a multi-dimensional conflict graph, as shown in Figure 1, including the following steps:

[0053] S1. Statistically analyze the information age change of each data stream within each superframe, and based on the formula for calculating the average expected information age of the data stream, model the problem of minimizing the overall average expected information age as an integer programming model of time slot and channel allocation. Specifically, this includes the following steps:

[0054] S11. Calculate the set F of data streams to be scheduled. For data stream f in F, determine its scheduling link sequence L. f for:

[0055] Among them, |L f | represents the number of hops in the data flow f. Let i represent the i-th hop link in the scheduling link of data flow f, where i∈{1,2,…,L} f At time t, when the destination node of f receives a new data packet, according to the information age function a... f (t) Update information: Age:

[0056] in, This indicates the time when the latest data packet of the source node of data stream f was generated at time t.

[0057] S12. Calculate the average information age of the data stream during the k-th superframe [kT, (k+1)T] based on whether the transmission was successful or not. This includes two cases: the first case is when the destination node of data stream f successfully receives the new data packet, in which case the average information age A is... f (t k ) is represented as:

[0058] Where x represents the independent variable in the integral, The value of the independent variable is t. k ~t k ',exist The value of the independent variable is t. k ′~t k+1 ;t k ' represents the time when the data packet arrives at the destination node in the k-th superframe, d x represents the derivative of the independent variable x; T represents the duration of a superframe.

[0059] In the second scenario, when the destination node of data stream f does not receive new data packets, the average information age A... f ′(t k ) is represented as:

[0060] Among them, in the formula The value of the independent variable is t. k ~t k+1 .

[0061] S13. Calculate and simplify the expected average information age E[A] of the data stream f in the kth superframe. f (t k )] is represented as:

[0062] in, This represents the probability that data packets in data stream f can be successfully received by the destination node. Indicates link The probability of successful transmission; Δ is an intermediate parameter, expressed as... This intermediate parameter is a constant.

[0063] S14. Establish an optimization system model for the average expected information age of each superframe intra-data stream. The objective function of the system model is expressed as follows:

[0064] Among them, t k ′≤t k+1 That is, t k ′-t k+1 If the value is less than or equal to 0, for ease of subsequent calculations, the objective function is negativeed and the constant term is removed, thus transforming it into a problem of finding the maximum value, which is equivalent to the following objective function:

[0065] S2. Construct a three-dimensional conflict graph containing three-dimensional tuples of data flow, link, and time slot, transforming the time slot scheduling problem into a maximum weighted independent set problem for solution. Specifically, this includes the following steps:

[0066] S21. Construct a three-dimensional tuple to describe the time slot allocation of the data stream f link, represented as:

[0067] Tuple τ represents the i-th hop link in data stream f. The allocated time slot t represents a possible time slot allocation scheme for a link. To obtain the required conflict-free time slot scheduling result among all time slot allocation schemes, a three-dimensional conflict graph G is constructed using each three-dimensional tuple τ as a node for time slot scheduling. s (V s E s ):

[0068] Where T = {1, 2, ..., |T|} is the time slot set, and V sA set of nodes of the three-dimensional conflict graph G s A set of edges of the three-dimensional conflict graph G s A set of edges of the three-dimensional conflict graph G s .

[0069] S22, the edges in the three-dimensional conflict graph G s (V s , E s ) represent that the node τ and τ' exist conflict, the conflict of the nodes in the conflict graph, that is, the conflict existing in the communication process of the communication device, which includes three kinds:

[0070] Radio frequency conflict, that is, when the communication device equipped with a half-duplex radio frequency module is assigned the same time slot and the corresponding link has a common communication device, radio frequency conflict exists;

[0071] Link repetition conflict, that is, if the same link in the same data flow is repeatedly scheduled, there is a link repetition scheduling conflict.

[0072] Link order conflict, that is, in order to ensure the continuity of routing forwarding, the routing order from the communication device of the routing source node to the communication device of the routing destination node needs to be scheduled, otherwise there is a link order conflict.

[0073] S23, introduce a binary variable λi, which represents whether the ith node τi in the three-dimensional conflict graph G s (V s , E s ) is in its maximum independent set, so as to ensure that all links in the set F are assigned conflict-free time slots, which is represented as:

[0074] S24, convert the time slot scheduling problem into finding a maximum weight independent set in the three-dimensional conflict graph G s (V s , E s ), and ensure that the number of elements of the maximum independent set of the three-dimensional conflict graph G s (V s , E s ) is equal to the total number of link hops in the set F, so as to guarantee the time slot resource allocation of the data flow link in the set F, and build an information age optimization scheduling model, which is represented as:

[0075] Constraint condition:

[0076] Where N v represents the number of nodes in the three-dimensional conflict graph G s (V s , E s ), which is represented as For any one node τ i ∈Vs its corresponding node weight is

[0077] S25, using a degree-based greedy strategy, gradually deleting nodes from the three-dimensional conflict graph G s (V s ,E s ) until no edges remain, and the resulting tuple set V s is the maximum independent set S0 of the three-dimensional conflict graph G s (V s ,E s ). Specifically, it includes the following steps:

[0078] S251, initialize the iteration number i = 0, and obtain the three-dimensional conflict graph G i = G s (V s ,E s );

[0079] S252, if the edge set E(G i ) of the three-dimensional conflict graph G i is empty, output the node set in the current conflict graph as the maximum independent set; otherwise, perform step S253;

[0080] S253, calculate the degree of each node in the three-dimensional conflict graph G i , and select a node and the edges connected to the node to be deleted according to the degree of the node and the weight value of the node. The selection process of the node is represented as:

[0081] where η i represents the node selected to be deleted in the i-th iteration; V(G i ) represents the set of all nodes in the three-dimensional conflict graph G i in the i-th iteration; W(μ) is the weight value of node μ in the three-dimensional conflict graph, represented as represents the degree of node μ in the three-dimensional conflict graph G i in the i-th iteration;

[0082] S254, after deleting node η i , update the edge relationship set and the node set of the three-dimensional conflict graph, and let i = i + 1, return to step S252.

[0083] S3, based on the time slot scheduling result, add a channel dimension to the three-dimensional tuple to expand it to a four-dimensional tuple and construct a four-dimensional conflict graph, convert the channel allocation problem into a maximum weight independent set problem for solving, and output the final scheduling result. Specifically, it includes the following steps:

[0084] S31, on the basis of completing time slot scheduling, allocating channels for the links of data streams in set F. Adding the channel dimension to the elements in set S0, and constructing a four-dimensional conflict graph G for channel scheduling c (V c ,E c ) is expressed as:

[0085] wherein V c and E c represent the node set and edge set of four-dimensional conflict graph G c (V c ,E c ) respectively, u and u' are nodes in four-dimensional conflict graph G c (V c ,E c ), c represents the allocated channel, and C represents the channel set. A four-dimensional tuple, i.e. a node in the four-dimensional conflict graph, represents the i-th hop link of data stream f allocated time slot t transmits data using channel c. After completing time slot scheduling, the radio frequency conflict and the link order conflict have been resolved, the edges of the four-dimensional conflict graph represent the link repetition conflict and the channel interference conflict, wherein the channel interference conflict refers to the existence of channel interference conflict when the nodes of two different links are in the interference range of each other and transmit data in the same time slot and channel.

[0086] S32, the overall channel allocation problem is decomposed into the per-hop scheduling sub-problems P i , which is expressed as:

[0087] In the above formula, P i represents the channel allocation sub-problem of the i-th hop scheduling, i={1, 2, …, h max}, h max is the hop number of the data stream with the largest hop number in data stream set F, which is expressed as represents the allocated time slot of the i-th hop link of data stream f. Since the data streams in set F have completed time slot scheduling, i.e. is a certain value.

[0088] S33, the h max sub-problems are solved in the order of , and for each sub-problem P i , a corresponding four-dimensional conflict graph G is constructed, which is expressed as:

[0089] wherein the node set V and the edge set E All come from the four-dimensional conflict graph G c (V c ,E c )。

[0090] S34, solving the sub-problem P i is equivalent to finding an optimal maximum independent set problem in the four-dimensional conflict graph . The kth node μ k in the four-dimensional conflict graph is represented in the maximum independent set by γ k , and is represented as:

[0091] The sub-problem P i is converted into a maximum weight independent set problem P i ', and is represented as:

[0092] The constraint condition is:

[0093] wherein, is the number of nodes in the four-dimensional conflict graph corresponding to the channel allocation sub-problem of the ith hop scheduling, and is represented as The constraint condition ensures that the ith hop link of the data flow in the set F i can be allocated to a channel; represents the probability of successful transmission of the link ; r(l(μ k )) represents the probability of successful transmission of the link l(μ k ).

[0094] S35, solving the maximum weight independent set problem P i ', wherein the node μ has a weight value of i from 1 to h max , and outputting the allocation scheme of all data flow link time slots and channels in the set F.

[0095] The embodiment also provides a pseudo code for obtaining a channel allocation strategy, and specifically includes:

[0096] Figures 2-3 show the verification results of a specific example. Figure 2 shows a simulation network containing 60 nodes randomly distributed in a 200x200 square meter area. In this network, the communication distance between any two nodes is set to 90 meters, and the electromagnetic interference distance between any two nodes is set to 50 meters. To simulate the transmission of data in the network, source nodes and destination nodes are randomly selected from the 60 nodes to generate data streams, and it is ensured that there are no identical source node and destination node pairs. To determine the optimal path from the source node to the destination node for each data stream, the Dijkstra algorithm is used to calculate and select the scheduling path.

[0097] Figure 3 shows the information age change curve of a random data stream in detail. The curve is based on a scenario containing 6 randomly generated data streams and 4 channels. As can be observed from the figure, as the information age of a data stream gradually increases, the subsequent scheduling strategy tends to allocate higher quality channel resources to the data stream, thereby effectively reducing its information age and ensuring the timeliness and transmission efficiency of the data stream.

[0098] Although embodiments of the present application have been shown and described, it is to be understood that various modifications, substitutions, replacements and changes can be made to these embodiments without departing from the principles and spirit of the present application, and the scope of the present application is defined by the appended claims and their equivalents.

Claims

1. A method for information age optimization scheduling of industrial wireless networks based on multi-dimensional conflict graph, characterized in that, Specifically comprising the following steps: S1, statistics of each data flow in each super-frame information age change, and based on the average expected information age calculation formula, the minimum overall average expected information age problem is modeled into an integer programming model of time slot and channel allocation; S2, a three-dimensional conflict graph composed of data flow-link-time slot three-dimensional tuples is constructed, the time slot scheduling problem is converted into a maximum weight independent set problem, and a time slot scheduling strategy is obtained by solving; S3, on the basis of the time slot scheduling strategy, the channel dimension is added to the data flow-link-time slot three-dimensional tuple to form a four-dimensional tuple and construct a four-dimensional conflict graph, the channel allocation problem is converted into a maximum weight independent set problem for solving, and a channel allocation strategy is obtained by solving.

2. The method of claim 1, wherein, The average expected information age of a data flow is expressed by the formula: where E[A f (t k )] denotes the average information age of data flow f in the kth super-frame [kT, (k+1)T]; A f (t k ) denotes the average information age of data flow f when a data packet is successfully received in the kth super-frame [kT, (k+1)T]; δ f denotes the probability that a data packet of data flow f can be successfully received by the destination node; A f '(t k ) denotes the average information age of data flow f when no data packet is received in the kth super-frame [kT, (k+1)T]; t k denotes the time when a data packet is generated by the source node in the kth super-frame; t k ' denotes the time when a data packet arrives at the destination node in the kth super-frame; t k+1 denotes the time when a data packet is generated by the source node in the (k+1)th super-frame; a f (t k ) denotes the information age function of data flow f at t k ; Δ is an intermediate parameter, and is expressed as T is the duration of a super-frame.

3. The method of claim 1 or 2, wherein, When modeling the minimization of the overall average expected information age problem as an integer programming model of time slot and channel allocation, the negative term and the constant term in the minimization of the overall average expected information age problem are removed, so as to convert the minimization problem into a maximization problem, and the objective function after conversion is expressed as: Where f represents a data stream; F is the set of all data streams to be scheduled; t k ' represents the time when the data packet arrives at the destination node in the k-th superframe; t k+1 This represents the time when the source node generates the data packet during the (k+1)th superframe; a f (t k The t represents the time t when the source node generates data packets in the k-th superframe of data stream f. k The information age function of time; T is the duration of a superframe; δ f This represents the probability that a data packet in data stream f can be successfully received by the destination node.

4. The method of claim 2, wherein, The information age function of the data stream f is expressed as: wherein a f (t) denotes the information age of the data stream f at time t; Indicates the generation time of the latest data packet of the data flow f source node at t time.

5. The method of claim 1, wherein, The process of constructing a three-dimensional conflict graph includes: A three-dimensional tuple consisting of data flow-link-time slot is constructed to describe the time slot allocation of a data flow f, which is expressed as: T represents the ith hop link in data stream f The allocated time slots are t, F is the set of all data streams, L f is the set of links for data stream f, and T is the duration of one superframe. A three-dimensional conflict graph G for time slot scheduling is constructed with each three-dimensional tuple τ as a node s (V s ,E s ), wherein V s represents a node set in the three-dimensional conflict graph, each node being a three-dimensional tuple, E s represents an edge relationship set in the three-dimensional conflict graph, an edge relationship existing between two nodes if a conflict exists between the two nodes The conflict between nodes, that is, the conflict existing between two communication devices in the communication process, includes: when the communication device equipped with a half-duplex radio frequency module is allocated the same time slot and the corresponding link has a common communication device, there is a radio frequency conflict between the common communication devices; if the same link in the same data flow is repeatedly scheduled, there is a link repeated scheduling conflict between the two scheduling; if the scheduling order of the link is not the order from the source node to the destination node, there is a link order conflict between the two communication devices.

6. The method of claim 1 or 5, wherein, The time slot scheduling problem is converted into a maximum weight independent set problem, i.e. finding a maximum weight independent set in a three-dimensional conflict graph, ensuring that the maximum independent set elements of the three-dimensional conflict graph are equal to the total number of link hops in the data flow set, and the maximum weight independent set problem is expressed as: Constraints: where N v denotes the number of nodes in the three-dimensional conflict graph; t k+1 denotes the time when the source node generates a data packet at the k+1th superframe; t(τ i ) is the time slot term of the ith node τ i ; a f (t k ) denotes the information age function of the data flow f when the source node generates a data packet at time t k at the kth superframe; T is the duration of a superframe; λ i is a binary variable, λ i = 1 if the ith node τ i is in its maximum independent set, otherwise λ i = 0; L f denotes the link set of the data flow f; F is the data flow set composed of all data flows to be scheduled; |·| denotes the number of elements in a set.

7. The method of claim 6, wherein, The process of converting the time slot scheduling problem into a maximum weight independent set problem and solving it includes: 201、initialize iteration number i = 0, obtain three-dimensional conflict graph G i = G s (V s , E s ); 202、if the edge set E(G i ) of the three-dimensional conflict graph G i is empty, output the node set in the current conflict graph as the maximum independent set; otherwise, perform step 203; 203、computing a three-dimensional conflict graph G i the degree of each node, and selecting a node and edges connected to the node according to the degree of the node and the weight value of the node, the process of selecting a node is represented as: wherein η i represents the node selected for deletion in the i-th iteration; V(G i ) represents the set of all nodes in the three-dimensional conflict graph G i in the i-th iteration; W(μ) is the weight value of node μ in the three-dimensional conflict graph, represented as denotes the degree of node μ in the three-dimensional conflict graph G at the i-th iteration i .

204. delete node η i The edge relation set and the node set of the post-updated three-dimensional conflict graph are updated, and i = i + 1, returning to step 202.

8. The method of claim 1, wherein, The process of converting the channel allocation problem into a maximum weight independent set problem for solving includes: 301, the overall channel allocation problem is decomposed into a hop-by-hop scheduling sub-problem by adopting a greedy strategy, and the hop-by-hop scheduling sub-problem is solved from the source node to the destination node; 302, the hop-by-hop scheduling sub-problem is converted into a maximum weight independent set problem, and the channel allocation strategy in one hop is obtained by solving the problem.

9. The method of claim 8, wherein, After the overall channel allocation problem is decomposed into the sub-problems of hop-by-hop scheduling by taking the greedy strategy, the sub-problem of the i-th hop scheduling is expressed as: where P i represents the i-th hop scheduled channel allocation sub-problem; F i is the set of data flows whose hop number is greater than or equal to i, and represents t k+1 denotes the time at which the source node generates the data packet at the (k+1)th superframe; denotes the time slot allocated to the i-th hop link of data flow f; T is the duration of one superframe; a f (t k ) denotes the information age function of data flow f at tk; representing a link probability of transmission success, represents the kth hop link of the data flow f; representing a link probability of transmission success, Indicates the i-th hop link of the data flow f.

10. The method of claim 8 or 9, wherein, Converting the sub-problem of hop-by-hop scheduling into solving the maximum weight independent set problem, the sub-problem of the i-th hop scheduling is converted into solving the maximum weight independent set problem and is expressed as: Constraints: where P i represents the i-th hop scheduled channel assignment sub-problem converted to a maximum weighted independent set problem; The number of nodes in the four-dimensional conflict graph corresponding to the channel allocation subproblem for the ith hop; t k+1 denotes the time at which the source node generates a data packet at the k+1th superframe; denotes the time slot allocated to the i-th hop link of data flow f; T is the duration of one superframe; a f (t k ) denotes the information age function of data flow f at tk; representing a link probability of transmission success, denotes the data flow f at the k-th node μ k the j-th hop link on the path from the source node to the k-th node; r(l(μ k )) denotes the probability of successful transmission over the link l(μ k ), l(μ k ), k denotes the link performed at the kth node μ k kth node μ k is a binary variable, γ k = 1 when the kth node μ k in the four-dimensional conflict graph is in a maximal independent set, otherwise γ k = 0; | · | denotes the number of elements in a set.

Images