Disaster area communication network multi-layer information propagation model

Abstract

The application provides a multi-layer information propagation model of a disaster area communication network, which comprises the following steps: designing a multi-layer information propagation model of a disaster area communication network, obtaining a balanced state of information propagation in the model, and proving the influence of parameters such as node density and node communication range on the balanced state of information propagation. An interference optimization problem based on the multi-layer information propagation model of the disaster area communication network is proposed, a convex optimization technique is used to minimize the network deployment cost, the optimal node density and node communication range are obtained, and network interference is effectively avoided. The effectiveness of the proposed model is verified by experiments under different environments, and simulation results show that the multi-layer information propagation model of the disaster area communication network can describe the information propagation process between nodes of the disaster area communication network, the interference of the environment on the network is considered, and the cost of rescue node deployment and communication is considered, so that the parameters of the disaster area communication network are optimized.

Description

Technical Field

[0001] This invention relates to the field of information dissemination, and in particular to a multi-layered information dissemination model for communication networks in disaster areas. Background Technology

[0002] Disaster area communication networks can transmit critical disaster information in real time, overcoming the communication interruptions that occur in traditional networks when core nodes are damaged, thus providing support for the dissemination of disaster information. Equipment in disaster area communication networks (such as drones) can overcome the limitations of complex terrain in disaster areas; their flexible networking and high mobility make them suitable for tasks such as environmental monitoring and disaster investigation. Furthermore, many devices do not need to establish direct connections through base stations and feature high speed and low power consumption, solving the problem of damaged traditional communication network infrastructure while ensuring the endurance of various devices in disaster areas, playing a crucial role in disaster relief. Therefore, researching information propagation models and interference optimization methods for disaster area communication networks has significant practical implications.

[0003] Currently, research on information propagation in networks is typically based on propagation dynamics models. However, single-layer information propagation dynamics models do not consider the propagation range of information within the network, nor do they layer the network according to the actual environment and analyze multi-layer networks. Multi-layer information propagation dynamics models fail to consider the impact of interference on information propagation results in the special environment of disaster area communication networks in practical applications, and do not analyze the correlation between the spatial distribution, connectivity, interference, and information propagation of nodes in the network. This invention proposes a multi-layer information propagation model for disaster area communication networks to improve upon the above shortcomings. Addressing the problem that disaster area communication networks are susceptible to interference leading to information propagation interruptions, it demonstrates the coupling relationship between parameters such as node density, node communication range, and interference. Furthermore, by minimizing network deployment costs, it obtains the optimal node density and node communication range, effectively avoiding network interference. Finally, experimental simulations are designed to verify the effectiveness of the model. Summary of the Invention

[0004] This invention addresses the problem of information propagation interruption in disaster area communication networks due to interference, proposing a multi-layered information propagation model for such networks. This model utilizes an improved propagation dynamics method to analyze the information propagation process between nodes in the disaster area communication network, obtaining the equilibrium state of information propagation and demonstrating the influence of parameters such as node density, node communication range, and interference on this equilibrium state. Furthermore, based on the multi-layered information propagation model, an interference optimization problem is proposed, and convex optimization techniques are used to minimize network deployment costs, obtaining optimal node density and node communication range, effectively avoiding network interference. Simulation results show that the proposed multi-layered information propagation model for disaster area communication networks can describe the information propagation process and provide parameter optimization for disaster area communication networks in response to network interference.

[0005] The aforementioned multi-layered information propagation model for disaster area communication networks includes the following steps:

[0006] 1) Design a multi-layer information propagation model for disaster area communication networks based on propagation dynamics;

[0007] 2) The equilibrium state of information propagation in the disaster area communication network is obtained by using dynamic equations, revealing the influence of parameters such as node density, node communication range and interference on the equilibrium state of information propagation.

[0008] 3) Based on the multi-layer information propagation model of the communication network in the disaster area, an interference optimization problem and related constraints are proposed. Through derivation, the interference optimization problem is transformed into a convex optimization problem.

[0009] 4) Conduct experiments under different environments to verify the effectiveness of the proposed model.

[0010] The method for designing a multi-layer information propagation model for disaster area communication networks based on propagation dynamics in step 1 above is as follows:

[0011] Communication networks in disaster areas, such as Figure 1 As shown, the network includes network nodes such as drones, rescue vehicles, and rescue personnel. Nodes communicate using a D2D (Device-to-Device) communication method. Based on the differences in communication capabilities, nodes are divided into three types: Type I nodes, Type II nodes, and Type III nodes. Type I nodes are drones deployed in the air, Type II nodes are rescue vehicles, and Type III nodes are rescue personnel carrying communication modules. To analyze information propagation in the disaster area communication network, this invention abstracts the disaster area communication network into a multi-layered information propagation model, such as... Figure 2As shown, the entire disaster area communication network is divided into three layers based on the information dissemination range. The first layer transmits the location information of the drones and consists of Type I nodes. The second layer transmits terrain and disaster information collected by the drones and consists of Type I and Type II nodes. The third layer transmits specific rescue plans and other information and consists of Type II and Type III nodes.

[0012] This invention defines the degree of a node as the number of its neighboring nodes, and uses the degree of a node to describe the connectivity of the network. The degree distribution of nodes in each layer of the network is as follows:

[0013] 1) Degree distribution of nodes in the first layer network: The first layer network contains only type I nodes. Nodes in this layer primarily communicate via type A antennas. Therefore, the degree distribution of nodes in this layer depends on the connection relationships between type A antennas, as shown in the equation:

[0014]

[0015] Where K1 represents the degree of a type I node in the first layer network, the average degree of the nodes in the first layer network is obtained as follows:

[0016]

[0017] 2) Degree Distribution of Second-Layer Network Nodes: The average degree of nodes in the second-layer network is denoted by E(K²). In this network layer, Type I nodes communicate with each other via Class A antennas, and Type II nodes communicate with Type I and Type II nodes within their communication range via Class B antennas. Therefore, the degree distribution of the second-layer network nodes is related to the connections between Class A antennas and the connections between Class B antennas. The average degree of the second-layer network nodes is derived as shown in the following equation:

[0018]

[0019] 3) Third-layer network connections: Similar to the second-layer network, the average degree E(K3) of a node in the third-layer network can be expressed as:

[0020]

[0021] The method for obtaining the equilibrium state of information propagation in the disaster area communication network in step 2 above is as follows:

[0022] Interference is the primary factor affecting the average probability δ of successful information transmission between two adjacent nodes in the disaster area communication network; therefore, it is used as a parameter to quantify the degree of interference in the disaster area communication network. Information propagation in the disaster area communication network is divided into single-message propagation and multi-message propagation. Within the same network layer, the propagation of different messages is independent; therefore, the message propagation process within the same network layer can be decomposed into single-message propagation within that layer. In reality, multiple messages propagate simultaneously across different network layers; therefore, further analysis of multi-message propagation is needed to describe the information propagation process of the entire disaster area communication network. The specific definitions of the two information propagation processes are as follows:

[0023] 1) Single message propagation: A node with a degree of k can be divided into three states: unknown state (U k ), informed status (I) k ) and broadcast status (B k The state transition model of a node is as follows: Figure 3 As shown, the dynamic equation for information propagation in the system is thus obtained:

[0024]

[0025]

[0026]

[0027] Among them, U k (t), I k (t) and B k (t) represents the state U of a node with degree k at time t. k I k and B k The proportions. Let them be respectively. This allows us to determine the equilibrium state of information propagation in the disaster area's communication network, revealing the impact of parameters such as node density, node communication range, and interference on the information propagation equilibrium state.

[0028] 2) Multiple Message Propagation: Assume a three-layer disaster zone communication network where one message is propagated simultaneously at each layer. Message 1 propagates in the first layer, message 2 in the second layer, and message 3 in the third layer. A type I node has a degree of in the first layer and a degree of in the second layer. This node could be in one of the following nine states: B k B l B k I l B k U l I k B l U k B l I k Il I k U l U k I l and U k U l The dynamic equations for multi-message propagation are as follows:

[0029]

[0030]

[0031]

[0032]

[0033]

[0034]

[0035]

[0036]

[0037]

[0038] Using the stationarity condition, we obtain the following system of equations:

[0039] ∑S k S l (t)=1, S∈{U, I, B} (17)

[0040]

[0041] Solving the above equations yields the equilibrium state of multi-message propagation.

[0042] The method for proposing the interference optimization problem based on the multi-layer information propagation model of the disaster area communication network in step 3 above is as follows:

[0043] To ensure that the communication network in the disaster area can meet the information dissemination needs despite interference, based on the network information dissemination balance obtained above, and under the premise that the network can overcome interference and the information dissemination capacity meets the rescue requirements, the minimum deployment cost is used, and its cost function is as follows:

[0044]

[0045] Where c0 represents the unit power consumption cost of a node, c1, c2, and c3 represent the costs of deploying a single Type I node, Type II node, and Type III node, respectively, and η represents the path loss exponent. To ensure the widespread dissemination of critical information in the network, the proportion of nodes in state B and state I in the network needs to be as large as possible. Taking U(t) as a constraint, the interference optimization problem of the entire network is as follows:

[0046]

[0047]

[0048]

[0049]

[0050] U k∨l (t)≤p4 (24)

[0051] U l∨m (t)≤p5 (25)

[0052]

[0053] Where p1 represents U in the first layer network k (t) represents the maximum allowed value. Similarly, p2 and p3 represent U in the second and third layer networks, respectively. l (t), U m (t) represents the maximum allowed proportion, where p4 and p5 represent the U in multi-message propagation, respectively. k∨l 9t) and U l∨m (t) represents the maximum allowed proportion, while equation (26) represents the hardware constraints of the node.

[0054] The cost function is a function of λ1, λ2, λ3, r1, r2, and r3. Solving it requires considering the constraints and the correlation between the parameters to determine the domain of the function. The equilibrium state U of single-message propagation in the first-layer network. k The constraint of (t) is U k If (t)≤p1, substituting the equilibrium solution into the constraint conditions yields:

[0055]

[0056] From this equation, we can derive an inequality concerning E(K1), which provides a method for calculating E(K1) using only λ1 and r1. Thus, for U... k The constraints on (t) are then transformed into constraints on parameters λ1 and r1. Similarly, the constraints on E(K2) and E(K3) can be transformed into constraints on λ2, r2, λ3, and r3, respectively:

[0057]

[0058]

[0059] For constraints, U can be... k∨l (t) decomposes into U k (t)+U l (t)-U k U l The form (t) for U k∨l The constraint on (t) is then transformed into constraints on λ1, r1, λ2, and r2. Similarly, for U l∨m The constraint on (t) is then transformed into constraints on λ2, r2, λ3, and r3. Correspondingly, the constraint condition represented by ~ becomes:

[0060]

[0061]

[0062]

[0063]

[0064]

[0065]

[0066]

[0067] The optimization problem proposed in the equation has a convex objective function and constraints. Therefore, convex optimization techniques can be used to solve this problem.

[0068] The method for verifying the effectiveness of the proposed model in different environments in step 4 above is as follows:

[0069] For a multi-layered information propagation model of a disaster area communication network, the proportions of nodes in three states are set to 0.9, 0, and 0.1, respectively. The proportion of nodes in state U gradually decreases to a stable value of 0.28 over time. The proportion of nodes in state I gradually increases in the early stages of information propagation and then decreases to a stable value of 0.38. The proportion of nodes in state B gradually increases over time and finally stabilizes at 0.34. The stable proportions of each state node are consistent with the mathematical derivation results, indicating that the model proposed in this invention can describe the information propagation process of a disaster area communication network.

[0070] To address the interference optimization problem of the multi-layer information propagation model in disaster area communication networks, interference optimization experiments were conducted under three different propagation thresholds of 0.1, 0.3, and 0.5. Under all three environments, as δ increases, r1, r2, and r3 gradually decrease, while λ1, λ2, and λ3 show a decreasing trend with increasing δ. The network deployment cost also decreases with increasing δ, indicating that the model proposed in this invention can provide parameter optimization for disaster area communication networks in response to network interference. Attached Figure Description

[0071] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will now be described in further detail with reference to the accompanying drawings, wherein:

[0072] Figure 1 For the communication network in the disaster area;

[0073] Figure 2 A multi-layered information propagation model for communication networks in disaster areas.

[0074] Figure 3 Node state transition model

[0075] Figure 4 Information dissemination process of communication networks in disaster areas

[0076] Figure 5 Optimal node density under different propagation thresholds

[0077] Figure 6 Optimal communication range under different propagation thresholds

[0078] Figure 7 Network deployment costs under different propagation thresholds Detailed Implementation

[0079] The specific embodiments of the present invention will be described in further detail below with reference to the accompanying drawings.

[0080] Figure 1 This is a communication network for the disaster area. It includes network nodes such as drones, rescue vehicles, and rescue personnel. Nodes communicate using a D2D (Device-to-Device) communication method. Each node in the disaster area communication network can be considered a Poisson point. Based on the differences in communication capabilities, nodes are divided into three types: Type I nodes, Type II nodes, and Type III nodes. Type I nodes are drones deployed in the air, with a node density of λ1 (nodes / km). 2 The drone is equipped with both Type A and Type B antennas, with communication ranges of r1 (meters) and r2 (meters) respectively. Type II nodes are for rescue vehicles, with a node density of λ2 (nodes / km). 2Rescue vehicles are equipped with both Class B and Class C antennas, with the Class C antenna having a communication range of r³ meters. Type III nodes are rescue personnel, with a node density of λ³ nodes per km², who carry communication modules equipped with Class C antennas. Only nodes with the same type of antenna and within the communication range can communicate with each other. If a node can communicate with another node, then the two nodes are adjacent.

[0081] The information transmitted through the disaster area's communication network includes drone location information, terrain and disaster information, and rescue plan information. Drone location information is transmitted between drones to avoid collisions; terrain and disaster information, collected by drones, is used by rescue vehicles to formulate rescue plans, and this information is transmitted between drones and rescue vehicles. Rescue plan information requires coordinated processing by rescue vehicles and personnel, therefore it is transmitted between rescue vehicles and personnel.

[0082] Figure 2 This is a multi-layered information propagation model of the disaster area communication network, abstracted from the actual network. Based on the scope of information propagation, the entire disaster area communication network is divided into three layers. The first layer propagates the location information of drones; therefore, this layer consists of Type I nodes. The second layer propagates terrain and disaster information collected by drones; this layer consists of Type I and Type II nodes. The third layer propagates specific rescue plans and other information, and consists of Type II and Type III nodes.

[0083] Figure 3 This is a node state transition model. The state transition process is as follows: Assume a node of degree k is in state U. k It will be in α k The probability of state I k Transition, due to state U k After receiving a message from a neighboring node, the node changes to state I. k Therefore, α k In reality, it is the average probability that a node of degree k receives messages from its neighboring nodes, expressed as α. k = kθ1δ, where θ1 is the average probability that neighboring nodes are aware. For state I k A node that receives a message containing a rescue mission arrangement for that node will transition to state B with probability γ. k Otherwise, it transitions to state U with probability β. k Where β + γ = 1. When a node transitions to state B k Afterwards, it will perform the rescue mission outlined in the message and broadcast the received message to neighboring nodes. Since it needs to wait for new messages, it will transition to state U. k .

[0084] Figure 4 This refers to the information dissemination process of communication networks in disaster areas. Figure 4 The three curves in the diagram represent the proportions of nodes in states U, I, and B, respectively. Initially, the proportions of nodes in the three states are set to 0.9, 0, and 0.1, respectively. The proportion of nodes in state U gradually decreases to a stable value of 0.28 over time. The proportion of nodes in state I gradually increases in the early stages of information propagation and then decreases to a stable value of 0.38. The proportion of nodes in state B gradually increases over time and finally stabilizes at 0.34. The stable proportions of each state are consistent with the mathematical derivation, indicating that the multi-layer information propagation model for disaster area communication networks proposed in this invention can describe the information propagation process of disaster area communication networks.

[0085] Figure 5 The optimal node density is shown for different propagation thresholds. Subgraphs (a), (b), and (c) illustrate the relationship between δ and the optimal node density when the propagation threshold is 0.1, 0.3, and 0.5, respectively. The three curves in each subgraph represent the optimized λ1, λ2, and λ3, respectively. In subgraph (a), when the propagation threshold is 0.1, λ1 decreases monotonically with δ, while the curve for λ2 is relatively flat. λ3 decreases significantly between δ = 0.2 and 0.3, and remains relatively flat at other δ values. Subgraph (b) shows that when the propagation threshold is 0.3, both λ1 and λ2 decrease monotonically with δ, while λ3 remains at its minimum value. Subgraph (c) shows that when the propagation threshold is 0.5, both λ1 and λ2 decrease monotonically with δ, while λ3 remains at its minimum value. When the propagation threshold is low and interference is high, a large number of nodes are needed to meet communication requirements. Therefore, when the propagation threshold is 0.1 and δ is small, λ1, λ2, and λ3 reach their maximum values. Increasing δ means reducing interference, and fewer nodes are needed to meet communication requirements, leading to a gradual decrease in node density. When communication requirements are low, i.e., when the propagation threshold is 0.3 and 0.5, only a small number of Type I and Type II nodes are needed to complete the information propagation task. Therefore, λ1 and λ2 only have larger values ​​when δ is small, and gradually decrease as δ increases, while λ3 only needs to reach its minimum value.

[0086] Figure 6The optimal communication range is shown for different propagation thresholds. Subfigure (a) shows the curves of r1, r2, and r3 as a function of δ obtained through optimization when the propagation threshold is 0.1. All three curves gradually decrease as δ increases. Subfigure (b) shows the curves of r1, r2, and r3 as a function of δ when the propagation threshold is 0.3. Among them, r1 and r2 gradually decrease to the minimum value as δ increases, while r3 rebounds to some extent as δ increases, but the overall trend is downward. Subfigure (c) shows the curves of r1, r2, and r3 as a function of δ when the propagation threshold is 0.5. At this point, r1 remains at the minimum value, r2 gradually decreases to the minimum value as δ increases, and r3 generally shows a downward trend as δ increases. The above results occur because as δ increases, a lower communication distance is sufficient to meet the needs of information transmission. At this point, it is necessary to minimize the communication distance between nodes to save costs. Therefore, r1, r2, and r3 decrease with the increase of δ. r3 rebounds in subgraphs (b) and (c) because the communication capability decreases due to the significant reduction in node density to reduce costs. At the same time, increasing r3 has the lowest cost, so slightly increasing r3 enhances the communication capability.

[0087] Figure 7 The figure shows the network deployment cost under different propagation thresholds. The curves in the figure represent the network deployment cost. It can be seen that the deployment cost curves in subgraphs (a), (b), and (c) all decrease as δ increases. This is because an increase in δ indicates a reduction in interference, which reduces the difficulty of information propagation and saves the overall deployment cost of the communication network in the disaster area.

[0088] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Obviously, those skilled in the art can make various modifications and variations to the present invention without departing from the spirit and scope of the present invention. Therefore, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention also intends to include these modifications and variations.

Claims

1. A multi-layer information propagation model for disaster area communication networks, characterized by the following steps: 1) A multi-layer information propagation model for disaster area communication networks is designed based on propagation dynamics. The method is as follows: The disaster area communication network comprises network nodes of drones, rescue vehicles, and rescue personnel, communicating via D2D. Based on their communication capabilities, nodes are categorized into three types: Type I, Type II, and Type III. Type I nodes are drones deployed in the air, Type II nodes are rescue vehicles, and Type III nodes are rescue personnel carrying communication modules. To analyze information propagation within the disaster area communication network, it is abstracted into a multi-layered information propagation model. Based on the propagation range, the entire network is divided into three layers: the first layer propagates drone location information and consists of Type I nodes; the second layer propagates terrain and disaster information collected by drones and consists of Type I and Type II nodes; the third layer propagates specific rescue plan information and consists of Type II and Type III nodes. This yields the average degree of each network layer's nodes. Where λ1 is the density of type I nodes, r1 is the communication range of type I nodes, λ2 is the density of type II nodes, r2 is the communication range of type II nodes, λ3 is the density of type III nodes, and r3 is the communication range of type III nodes. 2) The equilibrium state of information propagation in the disaster area communication network is obtained by using dynamic equations, revealing the influence of node density, node communication range and interference parameters on the equilibrium state of information propagation. 3) Based on the multi-layer information propagation model of the communication network in the disaster area, an interference optimization problem and related constraints are proposed. Through derivation, the interference optimization problem is transformed into a convex optimization problem. To ensure that the communication network in the disaster area can meet the information dissemination needs despite interference, based on the network information dissemination equilibrium state obtained in step 2), and under the premise that the network can overcome interference and the information dissemination capability meets the rescue requirements, the minimum deployment cost is used, and its cost function is as follows: Where c0 represents the unit power consumption cost of a node, c1, c2, and c3 represent the costs of deploying a single Type I node, Type II node, and Type III node, respectively, and η represents the path loss exponent. To ensure the widespread dissemination of critical information in the network, the proportion of nodes in state B and state I in the network needs to be as large as possible. Taking U(t) as a constraint, the interference optimization problem of the entire network is as follows: U k∨l (t)≤p4 (9) U l∨m (t)≤p5 (10) Where p1 represents U in the first layer network k (t) represents the maximum allowed value. Similarly, p2 and p3 represent U in the second and third layer networks, respectively. l (t), U m (t) represents the maximum allowed proportion, where p4 and p5 represent the U in multi-message propagation, respectively. k∨l (t) and U l∨m (t) represents the maximum allowed proportion, and equation (11) represents the hardware constraints of the node. Constraints (6) to (10) are infinite-dimensional constraints, and in practice, an infinite number of results cannot be obtained. Therefore, we take representative k = E(K1), l = E(K2), and m = E(K3) to represent the entire constraint. Then, we can substitute the equilibrium solution into constraint (6) to obtain: E(K1) can be represented by λ1 and r1, so constraint (6) is transformed into constraints on λ1 and r1; similarly, constraints (7) to (10) are transformed into constraints on λ2, r2, λ3 and r3 respectively: The optimization problem is transformed into: At this point, for the proposed optimization problem, both the objective function and the constraints are convex, and convex optimization is used to solve the problem. 4) Conduct experiments under different environments to verify the effectiveness of the proposed model.

2. The multi-layer information propagation model for a disaster area communication network as described in claim 1, wherein the method for obtaining the equilibrium state of information propagation in the disaster area communication network in step 2) is as follows: A node with a degree of k is divided into three states: unknown state (U k ), informed status (I) k ) and broadcast status (B k Based on the node state transition model, the dynamic equation for information propagation in the system is obtained: in, U k (t), I k (t) and B k (t) represents the state U of a node with degree k at time t. k I k and B k The proportions; respectively let This allows us to determine the balance state of information propagation in the disaster area's communication network, revealing the impact of node density, node communication range, and interference parameters on the balance state of information propagation. In a multi-message scenario, assuming a three-layer disaster zone communication network, each network layer has one message propagating simultaneously: message 1 propagates in the first layer, message 2 in the second layer, and message 3 in the third layer. A type I node has a degree of in the first layer and a degree of in the second layer. This node is in one of the following nine states: B k B l B k I l B k U l I k B l U k B l I k I l I k U l U k I l and U k U l The dynamic equations for multi-message propagation are as follows: Using the stationarity condition, we obtain the following system of equations: ∑S k S l (t)=1,S∈{U,I,B} (34) Solving the above equations yields the equilibrium state of multi-message propagation.

3. The multi-layer information propagation model for disaster area communication networks according to claim 1, wherein in step 4), the method for verifying the effectiveness of the proposed model through experiments under different environments is as follows: For the multi-layer information propagation model of the communication network in the disaster area, the proportions of three state nodes in the communication network in the disaster area are set to 0.9, 0 and 0.1 respectively. The changes in the proportions of each state node are observed, and the proportions after the state nodes stabilize are compared with the mathematical derivation results to determine whether the model proposed in the invention can describe the information propagation process of the communication network in the disaster area. To address the problem of multi-layer information propagation interference optimization in disaster-stricken communication networks, the average probability of successful node information transmission is set as δ, the value of which is determined by interference. Interference optimization experiments are conducted for different propagation thresholds. The trends of network parameters r1, r2, r3, λ1, λ2, λ3, and network deployment costs with the increase of δ are observed under different environments, illustrating the optimization effect on network interference.

Images