Search

CN-122002215-A - Wireless network positioning method, server and system

CN122002215ACN 122002215 ACN122002215 ACN 122002215ACN-122002215-A

Abstract

The invention discloses a wireless network positioning method, a server and a system, and relates to the technical field of wireless communication and positioning navigation. The method comprises the steps of receiving signals sent by a target node to be positioned through a plurality of positioning reference nodes, extracting an original positioning parameter sequence, then carrying out improved Z-score abnormal elimination based on median absolute deviation and adaptive attenuation coefficient smoothing based on average inter-frame jump quantity on the sequence, outputting a stable estimated value and a measurement variance, and finally carrying out position calculation according to the CHAN algorithm or the weighted least square method of the self-adaptive switching of the number of the effective positioning reference nodes. The invention adopts the improved Z-score filtering and dynamic weight fusion technology, overcomes the defects of the prior art, solves the problem of poor robustness of a fixed algorithm when positioning the topology change of a reference node through an algorithm self-adaptive switching mechanism, realizes the remarkable improvement of precision and adaptability under a complex scene, and reduces the deployment and maintenance cost.

Inventors

  • YANG SHAOSHI
  • ZHAO HAOWEI
  • LUO YUSONG
  • Zhai Houyu

Assignees

  • 北京邮电大学

Dates

Publication Date
20260508
Application Date
20260209

Claims (10)

  1. 1. A wireless network positioning method, comprising: receiving wireless signals from a target node to be positioned through a plurality of positioning reference nodes, and extracting an original positioning parameter sequence corresponding to the target node to be positioned; sequentially performing improved Z-score anomaly rejection based on median absolute deviation and adaptive attenuation coefficient smoothing based on average inter-frame jump quantity on the original positioning parameter sequence, and outputting stable parameter estimation value and measurement variance; And according to the stable parameter estimation value and the measurement variance, based on the number of positioning reference nodes according to which the stable parameter estimation value is generated, adaptively switching different geometric calculation algorithms, and calculating to obtain the space coordinates of the target node to be positioned.
  2. 2. The wireless network positioning method according to claim 1, wherein the improved Z-score anomaly rejection based on median and median absolute deviation specifically comprises: calculating the median and the median absolute deviation of the original positioning parameter sequence; calculating an improved Z-score value corresponding to each positioning parameter based on the median and the median absolute deviation; And comparing the improved Z-score value with a preset threshold value, and eliminating positioning parameters exceeding the preset threshold value.
  3. 3. The method of claim 2, wherein the modified Z-score is obtained by calculating an absolute value of a difference between a positioning parameter and a median and comparing the absolute value with a median adjusted by a normalization factor.
  4. 4. The wireless network positioning method according to claim 2, wherein the adaptive attenuation coefficient smoothing based on the average inter-frame hopping amount specifically comprises: calculating the average inter-frame jump quantity of the cleaned parameter sequence obtained after abnormal elimination and quantifying the fluctuation degree of the parameter; dynamically determining an adaptive attenuation coefficient according to the relation between the average inter-frame jump quantity and a preset reference fluctuation threshold value; and calculating the time attenuation weight of each parameter sample based on the self-adaptive attenuation coefficient, and carrying out weighted fusion on the cleaned parameter sequence.
  5. 5. The method according to claim 4, wherein the adaptive attenuation coefficient is defined between a preset minimum value and a maximum value, and is obtained by performing a dynamic mapping or clipping process according to a ratio of the average inter-frame hopping amount to a reference fluctuation threshold.
  6. 6. The wireless network positioning method according to claim 1, wherein the adaptively switching between different geometric calculation algorithms is specifically: When the number is smaller than a preset threshold, performing position calculation by adopting a Chan's analysis algorithm; And when the number is greater than or equal to the preset threshold value, adopting a weighted least square method to carry out position calculation.
  7. 7. The method of claim 6, wherein when the weighted least square method is adopted, an observation weight matrix is constructed, wherein the observation weight matrix is a diagonal matrix, each element on the diagonal is in one-to-one correspondence with a stable parameter estimation value participating in the calculation, and each element value is the inverse of the sum of a corresponding measurement variance and a measurement variance of a certain reference node, and is used for representing the confidence distribution of different reference node signals in the current environment.
  8. 8. A server device for wireless network location comprising a processor, a memory, and a communication interface, wherein: The processor and the memory are connected through a system bus; the communication interface is connected to the system bus through an input/output interface; the memory comprises an internal memory and a nonvolatile storage medium, wherein the nonvolatile storage medium is used for storing an operating system, a computer program and a database; The processor is configured to implement the wireless network positioning method according to any of claims 1-7 by executing the computer program and invoking the internal memory, receiving the original positioning parameter sequence via the communication interface.
  9. 9. A wireless network location system, comprising: A target node to be positioned for transmitting or receiving wireless signals; the positioning reference nodes are used for carrying out wireless communication with the target node to be positioned and extracting positioning parameters; the server apparatus of claim 8, configured to obtain positioning parameters from the positioning reference node, and perform positioning resolution to output position coordinates of the target node to be positioned.
  10. 10. A computer readable storage medium having stored thereon a computer program, which when executed by a processor implements a wireless network positioning method according to any of claims 1-7.

Description

Wireless network positioning method, server and system Technical Field The invention relates to the technical field of wireless communication and positioning navigation, in particular to a wireless network positioning method, a server and a system. Background Wireless location technology is used as a core support for modern information society, and the position of a target is determined by using geometric measurement or statistical deduction methods through extracting location related parameters (such as arrival time, signal strength and the like) from the radio wave propagation process. The technology is widely applied to the fields of public security, internet of things, intelligent transportation and the like, and the basic flow comprises three stages of signal feature extraction, positioning parameter estimation and space coordinate calculation. The positioning technology based on the communication network multiplexes the existing network infrastructure (such as base station and access point), has coverage advantage in indoor or urban canyon environment where satellite signals are blocked, and becomes a key component of a 'perception-communication' integrated architecture in a 5G/6G communication system. For example, early Cell positioning (Cell-ID) uses base station identification to roughly estimate position, while modern 4G/5G high-precision positioning uses multi-base station cooperative measurement to improve accuracy. However, as application scenarios become complicated (e.g., dynamic environments, multi-system networks coexist), the prior art exposes many limitations. Ext>ext> inext>ext> theext>ext> aspectext>ext> ofext>ext> aext>ext> specificext>ext> schemeext>ext>,ext>ext> theext>ext> Chineseext>ext> patentext>ext> applicationext>ext> withext>ext> theext>ext> applicationext>ext> numberext>ext> ofext>ext> CNext>ext> 202511223911.5ext>ext> disclosesext>ext> aext>ext> fusionext>ext> positioningext>ext> methodext>ext> basedext>ext> onext>ext> aext>ext> 5ext>ext> Gext>ext> -ext>ext> Aext>ext> baseext>ext> stationext>ext> andext>ext> aext>ext> UWBext>ext> baseext>ext> stationext>ext>,ext>ext> andext>ext> theext>ext> deploymentext>ext> costext>ext> isext>ext> reducedext>ext> byext>ext> optimizingext>ext> theext>ext> deploymentext>ext> positionext>ext> andext>ext> theext>ext> numberext>ext> ofext>ext> theext>ext> baseext>ext> stationsext>ext> andext>ext> utilizingext>ext> theext>ext> wideext>ext> coverageext>ext> ofext>ext> theext>ext> 5ext>ext> Gext>ext> -ext>ext> Aext>ext> toext>ext> complementext>ext> theext>ext> highext>ext> -ext>ext> precisionext>ext> characteristicext>ext> ofext>ext> UWBext>ext>.ext>ext> According to the scheme, a base station network is planned according to environment information, ranging signals are screened based on signal intensity, and algorithms such as Kalman filtering are introduced to conduct track prediction. The method has the core problems that the filtering algorithm (such as extended Kalman filtering) has high computational complexity, strictly depends on a preset motion model (such as a constant speed model), is easy to cause filtering divergence when a terminal makes a sharp turn or stops suddenly, has strict requirements on physical layout of a base station, needs fine planning in the initial stage of construction, and is difficult to adapt to a deployed heterogeneous network. Another related scheme, chinese patent application number CN201810076225.3, discloses an indoor positioning method based on ranging error correction, which builds an error compensation function model by measuring environmental data in advance, and performs linear correction on the ranging value to suppress the deviation caused by non-line-of-sight and multipath effects. However, the scheme needs a large amount of manual sampling to construct a static prior model, the model fails when the environment is changed (such as furniture is moved), the maintenance cost is high, and the sudden impulse interference cannot be identified by the fixed function model, so that corrected data still contains outlier errors. The deeper problems are that the prior art generally lacks environment self-adaption capability, namely 1) an algorithm level, a single resolving strategy (such as trilateration method) cannot dynamically respond to the quantity change of base stations (such as base station failure caused by shielding), and resolving stability is poor when the base stations are sparse, and 2) a data level depends on static modeling or simple linear processing, so that non-Gaussian noise under a complex channel is difficult to deal with. Together, these problems result in a positioning system that is not robust in dynamic scenarios and has poor deployment flexibility. Therefore, a wireless network positioning scheme capable of sensing environmental fluctuation in real time and adaptively adjusting a processing strategy is needed, so that dependence on hardware deployment and a static model in the prior ar