Low-overhead ad hoc network positioning method

Abstract

The invention relates to the technical field of wireless ad hoc network positioning, and discloses a low-overhead ad hoc network positioning method. The method comprises the following steps that: (1) an anchor node periodically broadcasts the ID, the position information and the Cramer-Rao lower bound value of the node, and a common node monitors and collects the information of positioned neighbor nodes; (2) dynamically determining the number of optimal reference neighbors and the number of times of distance measurement through a genetic algorithm, and meeting the global positioning precision constraint by taking the minimum resource consumption as a target function; (3) according to an optimization result of the step (2), TOA distance measurement is carried out on selected neighbor nodes, and a current position estimation value is calculated; and (4) checking whether the Cramer-Rao lower bound value of the node meets the global minimum error threshold, if so, updating the node state and broadcasting the position information, and otherwise, returning to the step (1). The ad hoc network positioning method realized by the invention has the characteristics of high resource utilization rate, strong error accumulation inhibition capability and excellent dynamic adaptability, and provides possibility for efficient positioning of a large-scale ad hoc network.

Description

Technical Field

[0001] This invention relates to the field of wireless ad hoc network positioning technology, specifically to a low-overhead ad hoc network positioning method based on the Cramerlow lower bound (CRLB) error propagation model and genetic algorithm optimization, aiming to solve the problems of resource waste, serious error accumulation and insufficient dynamic adaptability in traditional cascaded positioning systems. Background Technology

[0002] In wireless ad hoc networks, accurate localization of distributed nodes is fundamental for network collaboration, path planning, and service optimization. Existing technologies primarily employ a cascaded localization model: anchor nodes transmit location information via multiple hops, while ordinary nodes complete their own localization based on ranging data from their located neighbors. However, traditional methods have the following drawbacks:

[0003] (1) Low resource utilization: Fixed parameter allocation strategy (such as fixed reference neighbor number N=3, ranging number T=10) leads to low-level nodes consuming excessive resources, while high-level nodes may not be able to meet accuracy requirements due to insufficient resources.

[0004] (2) Accumulated error cascade: The positioning error of the preceding node will propagate to the subsequent levels through the ranging process, causing the positioning accuracy of the higher-level nodes to drop sharply.

[0005] (3) Lack of theoretical support: Existing dynamic resource allocation schemes mostly rely on empirical rules (such as linearly increasing the number of distance measurements) and have not established a strict error propagation model, making it difficult to achieve global optimization of accuracy and resources. Summary of the Invention

[0006] To address the aforementioned shortcomings, this invention proposes a low-overhead ad hoc network positioning method. By dynamically adjusting the number of reference neighbors and the number of ranging operations, this method significantly reduces system resource consumption and improves network scalability while meeting global positioning accuracy constraints.

[0007] The technical solution adopted in this invention is as follows:

[0008] A low-overhead ad hoc network positioning method includes the following steps:

[0009] (1) Anchor nodes periodically broadcast known location information, ordinary nodes initialize and set initial parameters; after the ordinary nodes are initialized, they start listening to neighbor node information and collect information of the located neighbor nodes, including neighbor node ID, location information and Cramerlow lower bound value. After the listening period ends, it is determined whether the positioning condition is met, that is, the number of located neighbor nodes is greater than or equal to three. If so, proceed to step (2), otherwise continue to listen; among them, the located nodes are anchor nodes.

[0010] (2) Define the resource optimization objective of a normal node as: minimizing the resource usage of the node while satisfying the given positioning error. The constraint is: the lower bound of the Cramerlow value at this node is less than the global maximum permissible error threshold. ; and through iterative optimization using a genetic algorithm, the optimal neighbor combination for this node and the corresponding distance measurement count allocation scheme are obtained; among which, This represents the number of neighboring nodes that have been located. Let be the number of distance measurements required for the i-th neighbor node;

[0011] (3) Based on the optimization results of step (2), the node performs TOA distance measurement with the corresponding neighboring nodes to calculate the node's position;

[0012] (4) Update information and location broadcast: Based on the location information of this node and the corresponding neighbor node information, recalculate the lower bound of the Cramer-Rao value of this node, and verify whether the lower bound of the Cramer-Rao value of this node is less than the global maximum allowable error threshold. If the requirements are met, update the status and broadcast it; otherwise, return to step (1).

[0013] Furthermore, the formula for calculating the lower bound value of the local node in steps (2) and (4) is as follows:

[0014] ;

[0015] In the formula, Let Cramerlow be the lower bound of this node. This is the location of this node. The location of the neighboring node, and the distance information are: , , , The number of distance measurements taken for each neighboring node. The variance of a single distance measurement. and The first The neighbor node and the first The lower bound of the Cramer-Rao term for each neighboring node; wherein, when calculating in step (2) The value is N. To estimate the location, the calculation in step (4) is performed... To determine the optimal number of neighboring nodes, , The position is calculated in step (3).

[0016] Furthermore, updating the state and broadcasting in step (4) includes: (4.1) Compare the lower bound of the current node's Cramerlow with the global maximum permissible error threshold. If the lower bound of the current node's Cramer-Rao threshold is less than the global maximum permissible error threshold... If so, then mark it as a located node;

[0017] (4.2) Encapsulate the node ID, location information, and Cramerlow lower bound of this node into a broadcast message;

[0018] (4.3) After receiving the broadcast message, other nodes will add the nodes in the broadcast message to the neighbor list and participate in their own positioning optimization.

[0019] The advantages of this invention compared to the prior art are:

[0020] The self-organizing network positioning method implemented by this invention has the advantages of high resource utilization, strong error accumulation suppression capability, and excellent dynamic adaptability, providing the possibility for efficient positioning of large-scale self-organizing networks.

[0021] This invention significantly reduces system resource consumption and improves network scalability by dynamically adjusting the number of reference neighbors and the number of ranging operations, while meeting global positioning accuracy constraints. Attached Figure Description

[0022] Figure 1 This is a flowchart of the method described in this invention, illustrating the complete iterative process from neighbor information collection to location broadcasting;

[0023] Figure 2 It is a randomly generated scene graph with node positioning, where red nodes are anchor nodes and black nodes are ordinary nodes; Figure 3 The simulation results comparison chart illustrates the differences between the present invention and the fixed resource allocation strategy in terms of resource usage at various node levels. Detailed Implementation

[0024] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0025] This invention provides a low-overhead ad hoc network location method, such as... Figure 1 As shown, the main steps include:

[0026] (1) Initialization and Neighbor Information Collection

[0027] Specifically, it includes the following steps:

[0028] (1.1) Anchor nodes obtain absolute positions via GPS and broadcast beacon messages containing position coordinates and Cramerlow lower bounds every 5 seconds; among them, the nodes that have been located are anchor nodes;

[0029] (1.2) After initialization, ordinary nodes enter listening mode, set the initial Cramerlow lower bound to infinity, and record the neighbor node ID, location, and Cramerlow lower bound of the beacon message received;

[0030] (1.3) If the number of located neighbor nodes detected is ≥3 (e.g.) Figure 2 Middle node When location is required, the already located neighbor is , and If the condition is met, the resource optimization process in step (2) will be triggered; otherwise, continue listening.

[0031] (2) Resource optimization and neighbor selection

[0032] In this embodiment, we assume a global maximum permissible error threshold. The distance is 10m, and the standard deviation of a single distance measurement is σ = 10m. Based on the neighbor information obtained in step (1.3), calculate the lower bound of Cramérault:

[0033] ;

[0034] In the formula, Let be the lower bound of the Cramérault value for this node, and let be the estimated location of this node. The location information of neighboring nodes is , N is the number of neighboring nodes. Distance information is defined as: , , ,definition and The first The neighbor node and the first The lower bound of the Clamerlow neighbor nodes. For the number of neighbors used, To measure the distance to each neighbor, This represents the variance of a single distance measurement.

[0035] The resource optimization objective for a regular node is defined as: minimizing the resource usage of the node while satisfying a given positioning error. The constraint is: the lower bound of the Cramerlow value at this node is less than the global maximum permissible error threshold. The objective function (minimizing the total number of distance measurements) is optimized using a genetic algorithm, and the optimal solution is output after 100 generations of population generation. For example, nodes... At level 2, it may be assigned to select 2 neighbors ( , Each of the three neighbors was measured 4 times, while the three neighbors were measured 10 times each, which is not a fixed strategy.

[0036] (3) Perform distance measurement and position calculation

[0037] Based on the optimization results of step (2), the node performs TOA distance measurement with its corresponding neighboring nodes to calculate the node's position;

[0038] In this embodiment, based on the result of step (2), the node and , Four TOA distance measurements were performed to construct the FIM matrix and calculate the position estimate.

[0039] (4) Update information and location broadcast

[0040] According to the node Location information and corresponding neighbor node information are used to recalculate the node. The lower bound of Cramer-Rao is calculated using the following formula:

[0041] ;

[0042] In the formula, The recalculated lower bound value for this node. The position of this node is calculated in step (3). The location of the neighboring node, and the distance information are: , , , The number of distance measurements taken for each neighboring node. The variance of a single distance measurement. and The first The neighbor node and the first The Cramerlow bound of the neighboring nodes; where... , This represents the optimal number of neighboring nodes.

[0043] Verify whether the lower bound of the current node's Cramer-Rao threshold is less than the global maximum permissible error threshold. If the lower bound value of Cramer-Rao is calculated =8.5m≤10m, then update the status and broadcast; otherwise, return to step (1).

[0044] The specific process of updating the status and broadcasting includes:

[0045] (4.1) Compare the lower bound of the current node's Cramerlow with the global maximum permissible error threshold. If the lower bound of the current node's Cramer-Rao threshold is less than the global maximum permissible error threshold... If so, update its own Cramerlow lower bound to 8.5m and mark it as located;

[0046] (4.2) Encapsulate the node's location and Craméror's lower bound into a broadcast message; the broadcast message carries the following information: the node ID, location information, and Craméror's lower bound of the node;

[0047] (4.3) Other nodes (such as After receiving a broadcast message, the node in the broadcast message will be added to the neighbor list and participate in its own positioning optimization.

[0048] Randomly generated node location scene diagram as follows Figure 2 As shown in the figure, the simulation results are compared. Figure 3 As shown.

Claims

1. A low-overhead self-organizing network positioning method, characterized in that, Includes the following steps: (1) Anchor nodes periodically broadcast known location information, and ordinary nodes initialize and set initial parameters; After the initialization of the ordinary node is completed, it starts to listen to the information of neighboring nodes and collects the information of the located neighboring nodes, including the neighboring node ID, location information and the lower bound of the Cramerlow. After the listening period ends, it is determined whether the positioning condition is met, that is, the number of located neighboring nodes is greater than or equal to three. If so, it proceeds to step (2), otherwise it continues to listen. Among them, the located nodes are anchor nodes. (2) Define the resource optimization objective of a normal node as: minimizing the resource usage of the node while satisfying the given positioning error. The constraint is: the lower bound of the Cramerlow value at this node is less than the global maximum permissible error threshold. ; and through iterative optimization using a genetic algorithm, the optimal neighbor combination for this node and the corresponding distance measurement count allocation scheme are obtained; among which, This represents the number of neighboring nodes that have been located. Let be the number of distance measurements required for the i-th neighbor node; (3) Based on the optimization results of step (2), the node performs TOA distance measurement with the corresponding neighboring nodes to calculate the node's position; (4) Update information and location broadcast: Based on the location information of this node and the corresponding neighbor node information, recalculate the lower bound of the Cramer-Rao value of this node, and verify whether the lower bound of the Cramer-Rao value of this node is less than the global maximum allowable error threshold. If the requirements are met, update the status and broadcast it; otherwise, return to step (1).

2. The low-overhead ad hoc network positioning method according to claim 1, characterized in that, The formulas for calculating the lower bound value of the local node in steps (2) and (4) are as follows: ； In the formula, Let Cramerlow be the lower bound of this node. This is the location of this node. The location of the neighboring node, and the distance information are: , , , The number of distance measurements taken for each neighboring node. The variance of a single distance measurement. and The first The neighbor node and the first The lower bound of the Cramer-Rao term for each neighboring node; wherein, when calculating in step (2) The value is N. To estimate the location, the calculation in step (4) is performed... To determine the optimal number of neighboring nodes, , The position is calculated in step (3).

3. The low-overhead ad hoc network positioning method according to claim 1, characterized in that, The step (4) of updating the status and broadcasting includes: (4.1) Compare the lower bound of the current node's Cramerlow with the global maximum permissible error threshold. If the lower bound of the current node's Cramer-Rao threshold is less than the global maximum permissible error threshold... If so, then mark it as a located node; (4.2) Encapsulate the node ID, location information, and Cramerlow lower bound of this node into a broadcast message; (4.3) After receiving the broadcast message, other nodes will add the nodes in the broadcast message to the neighbor list and participate in their own positioning optimization.

Images