CN112800983A - Non-line-of-sight signal identification method based on random forest - Google Patents

Non-line-of-sight signal identification method based on random forest Download PDF

Info

Publication number
CN112800983A
CN112800983A CN202110138933.7A CN202110138933A CN112800983A CN 112800983 A CN112800983 A CN 112800983A CN 202110138933 A CN202110138933 A CN 202110138933A CN 112800983 A CN112800983 A CN 112800983A
Authority
CN
China
Prior art keywords
sight
line
training
distance
decision tree
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN202110138933.7A
Other languages
Chinese (zh)
Other versions
CN112800983B (en
Inventor
杨小凤
韦艳芳
王强
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Yulin Normal University
Original Assignee
Yulin Normal University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Yulin Normal University filed Critical Yulin Normal University
Priority to CN202110138933.7A priority Critical patent/CN112800983B/en
Publication of CN112800983A publication Critical patent/CN112800983A/en
Application granted granted Critical
Publication of CN112800983B publication Critical patent/CN112800983B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
    • G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G06F18/243—Classification techniques relating to the number of classes
    • G06F18/24323—Tree-organised classifiers
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N20/00—Machine learning
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
    • G06F2218/12—Classification; Matching
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
    • Y02D30/00—Reducing energy consumption in communication networks
    • Y02D30/70—Reducing energy consumption in communication networks in wireless communication networks

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • General Engineering & Computer Science (AREA)
  • Artificial Intelligence (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Evolutionary Computation (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Software Systems (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Evolutionary Biology (AREA)
  • Computing Systems (AREA)
  • Mathematical Physics (AREA)
  • Medical Informatics (AREA)
  • Radar Systems Or Details Thereof (AREA)
  • Position Fixing By Use Of Radio Waves (AREA)
  • Mobile Radio Communication Systems (AREA)

Abstract

The invention discloses a non-line-of-sight signal identification method based on random forest, which relates to the technical field of wireless positioning and solves the technical problems of poor real-time performance and low precision of the existing positioning method, and the method comprises the following steps: constructing a random forest model consisting of a plurality of decision trees, and sequentially measuring received signals r from target nodes to each base station*(t) adding r*And (t) inputting the characteristic parameters into a random forest model to obtain the sight distance/non-sight distance recognition results of the target node and the signals received by each base station, and removing the signals recognized as the non-sight distances when positioning is carried out by utilizing the received signals.

Description

Non-line-of-sight signal identification method based on random forest
Technical Field
The invention relates to the technical field of wireless positioning, in particular to a non-line-of-sight signal identification method based on random forests.
Background
Wireless Localization (Wireless Localization) is widely applied to the fields of military affairs, logistics, safety, medical treatment, search, rescue and the like. The method improves the positioning accuracy of a positioning system in a complex multipath and non-line-of-sight (NLOS) environment, which is a research hotspot of wireless positioning based on Time-of-Arrival (TOA) at present, and one of the key problems is non-line-of-sight signal identification.
The non-line-of-sight signal identification means that when the distance measurement value is more, the non-line-of-sight measurement value is identified and removed, and positioning is carried out by using only the line-of-sight measurement value. Currently, there are three main types of methods:
1) the method based on distance measurement compares the variance of a plurality of distance measurement values with a preset threshold according to the fact that the variance of the distance measurement values under the non-line-of-sight environment is larger than the variance of the distance measurement values under the line-of-sight environment, the distance measurement values are judged to be non-line-of-sight signals when the variance of the distance measurement values is larger than the threshold, and the distance measurement values are judged to be line-of-sight signals when the variance of the distance measurement values is smaller than the threshold;
2) the method based on the channel statistical characteristics identifies non-line-of-sight signals by analyzing the cumulative distribution function of the amplitude of the channel impulse response, or identifies non-line-of-sight signals by comparing the joint likelihood function value of kurtosis (kurtosis), mean time delay (mean excess delay) and root mean square delay (root mean square delay) of the channel with a threshold value, but the definition of the decision threshold is fuzzy;
3) the channels are identified by the geographical geometrical information of the environment, and the non-line-of-sight signals are identified by using a ray tracing algorithm, so that the layout of the environment needs to be known in advance.
The above algorithms all belong to statistical methods, and have the common disadvantages that: firstly, the prior distribution of samples is generally required to be known in advance, and enough sample data is required to be collected, and the requirements are often difficult to achieve in practical application and the algorithm instantaneity is not high; the feature joint probability distribution required by the algorithm (II) is sometimes difficult to determine and has poor stability.
Disclosure of Invention
The invention aims to solve the technical problem of the prior art, and aims to provide a non-line-of-sight signal identification method based on random forests, which is high in real-time performance and good in stability.
The technical scheme of the invention is as follows: a non-line-of-sight signal identification method based on random forests comprises the following steps:
s1, randomly selecting N training positions in a test area containing 1 target node and A base stations, sequentially placing training communication nodes at each training position, and measuring received signals r of the training communication nodes from each training position N to the Kth base stationn(t)，K＝1，2，…，A；
Respectively calculate to obtain rn6 characteristic parameters of (t), including:
the energy parameter e ═ jj rn(t)|2dt、
Maximum amplitude parameter rmax＝maxr|rn(t)|、
Rise time parameter trise＝mint{t：|rn(t)|≥0.6rmax}-mint{t：|rn(t)|≥6σn}、
Average delay parameter
Figure BDA0002927886280000021
Root mean square delay parameter
Figure BDA0002927886280000022
Kurtosis parameter
Figure BDA0002927886280000023
Forming the 6 characteristic parameters into a characteristic set F ═ { e, r ═ emax，trise，τm，τr，κsR obtained from N positionsn(t) forming a training input matrix
Figure BDA0002927886280000024
Building a training output matrix
Figure BDA0002927886280000025
Wherein y isnIs rn(t) a sight-distance or non-sight-distance identification mark, and if it is a sight distance, ynWhen the non-line-of-sight distance is recognized as 1, ynGet the complete training set as 0
Figure BDA0002927886280000026
S2, measuring a received signal r from the target node to the base station*(t) and calculating r*6 characteristic parameters of (t);
s3, constructing a random forest model consisting of a plurality of decision trees, and for each decision tree: sampling N times from a complete training set D in a mode of putting back to sample to form a training set D' of the decision tree, randomly selecting M characteristic parameters from a characteristic set F as the characteristics to be selected of the decision tree, wherein M is less than 6, calculating the Gini index of the characteristics to be selected, and sequentially using the characteristics to be selected as the splitting characteristics of a root node, a middle node and a leaf node of the decision tree according to the sequence of the Gini index from small to large,
Figure BDA0002927886280000031
Figure BDA0002927886280000032
wherein Gini (D ', f) identifies and divides the training set D' into D according to the feature f to be selected1(stadia) and D2Gini index of two classes (non-line of sight), Gini (D)i) Is of class DiGini index, | D, of (i ═ 1, 2)1|、|D2Respectively is set D1、D2The number of samples in D', piIs of class DiIdentifying the probability of classification correctness and errors;
s4, receiving a signal r by a target node*(t) inputting the characteristic parameters of the random forest model, comparing the characteristic parameters with the splitting characteristic value interval range of each node by each decision tree, outputting the line-of-sight/non-line-of-sight identification type of each decision tree, and judging results of all decision treesMost of them are used as the final sight distance/non-sight distance identification result of the base station;
s5, repeating the steps S1-S4 for the rest base stations in the test area to obtain the sight distance/non-sight distance identification result of the target node and the signals received by each base station, and removing the signals identified as the non-sight distance when positioning is carried out by utilizing the received signals.
As a further improvement, A is more than or equal to 3.
Advantageous effects
Compared with the prior art, the invention has the advantages that:
the method of the invention takes the non-line-of-sight signal identification as a line-of-sight/non-line-of-sight two-class classification problem to be processed, adopts a machine learning method-random forest for identification and classification, the random forest is a high-precision classifier comprising a plurality of decision trees, the output class of the random forest is determined by most of the class output by each decision tree, compared with the existing statistical method, the method of the invention belongs to the machine learning method, does not depend on the distribution form of the population to which the sample belongs, only depends on the property of a small amount of data observation values which is irrelevant to the population distribution for inspection and estimation, can effectively reduce the inference deviation, improve the non-line-of-sight positioning accuracy, has high real-time and good stability, and is very suitable for a wireless positioning system based on the arrival time.
Drawings
FIG. 1 is a flow chart of the present invention.
Detailed Description
The invention will be further described with reference to specific embodiments shown in the drawings.
Referring to fig. 1, a non-line-of-sight signal identification method based on a random forest includes the following steps:
s1, randomly selecting N training positions in a test area containing 1 target node and A base stations, sequentially placing training communication nodes at each training position, and measuring received signals r from each training position N to the Kth base station of the training communication nodesn(t)，A≥3，K＝1，2，…，A；
Respectively calculate to obtain rn(t) 6 characteristics ofA number, comprising:
the energy parameter e ═ jj rn(t)|2dt、
Maximum amplitude parameter rmax＝maxr|rn(t)|、
Rise time parameter trise＝mint{t：|rn(t)|≥0.6rmax}-mint{t：|rn(t)|≥6σn}、
Average delay parameter
Figure BDA0002927886280000041
Root mean square delay parameter
Figure BDA0002927886280000042
Kurtosis parameter
Figure BDA0002927886280000043
The 6 characteristic parameters are combined into a characteristic set F ═ { e, r ═ emax，trise，τm，τr，κsR obtained from N positionsn(t) forming a training input matrix
Figure BDA0002927886280000044
Building a training output matrix
Figure BDA0002927886280000045
Wherein y isnIs rn(t) a sight-distance or non-sight-distance identification mark, and if it is a sight distance, ynWhen the non-line-of-sight distance is recognized as 1, ynGet the complete training set as 0
Figure BDA0002927886280000051
S2, measuring a received signal r from a target node to the base station*(t) and calculating r*6 characteristic parameters of (t);
s3, constructing a random forest model consisting of a plurality of decision trees, and for each decision tree: sampling N times from a complete training set D in a mode of putting back to sample to form a training set D' of the decision tree, randomly selecting M characteristic parameters from a characteristic set F as the characteristics to be selected of the decision tree, wherein M is less than 6, calculating the Gini index of the characteristics to be selected, sequentially using the characteristics to be selected as the splitting characteristics of a root node, a middle node and a leaf node of the decision tree according to the sequence of the Gini index from small to large,
Figure BDA0002927886280000052
Figure BDA0002927886280000053
wherein Gini (D ', f) identifies and divides the training set D' into D according to the feature f to be selected1(stadia) and D2Gini index of two classes (non-line of sight), Gini (D)i) Is of class DiGini index, | D, of (i ═ 1, 2)1|、|D2Respectively is set D1、D2The number of samples in D', piIs of class DiIdentifying the probability of classification correctness and errors;
s4, receiving a signal r by a target node*(t) inputting the characteristic parameters of the decision tree into a random forest model, comparing the characteristic parameters with the range of the splitting characteristic value interval of each node by each decision tree, and outputting the line-of-sight/non-line-of-sight identification type of each decision tree, wherein most of judgment results of all decision trees are used as the final line-of-sight/non-line-of-sight identification result of the base station;
s5, repeating the steps S1-S4 for the rest base stations in the test area to obtain the sight distance/non-sight distance identification result of the target node and the signals received by each base station, and removing the signals identified as the non-sight distance when positioning is carried out by utilizing the received signals.
The method of the invention takes the non-line-of-sight signal identification as a line-of-sight/non-line-of-sight two-class classification problem to be processed, adopts a machine learning method-random forest for identification and classification, the random forest is a high-precision classifier comprising a plurality of decision trees, the output class of the random forest is determined by most of the class output by each decision tree, compared with the existing statistical method, the method of the invention belongs to the machine learning method, does not depend on the distribution form of the population to which the sample belongs, only depends on the property of a small amount of data observation values which is irrelevant to the population distribution for inspection and estimation, can effectively reduce the inference deviation, improve the non-line-of-sight positioning accuracy, has high real-time and good stability, and is very suitable for a wireless positioning system based on the arrival time.
The above is only a preferred embodiment of the present invention, and it should be noted that it is obvious to those skilled in the art that several variations and modifications can be made without departing from the structure of the present invention, which will not affect the effect of the implementation of the present invention and the utility of the patent.

Claims (2)

1. A non-line-of-sight signal identification method based on random forests is characterized by comprising the following steps:
s1, randomly selecting N training positions in a test area containing 1 target node and A base stations, sequentially placing training communication nodes at each training position, and measuring received signals r of the training communication nodes from each training position N to the Kth base stationn(t)，K＝1，2，…，A；
Respectively calculate to obtain rn6 characteristic parameters of (t), including:
the energy parameter e ═ jj rn(t)|2dt、
Maximum amplitude parameter rmax＝maxr|rn(t)|、
Rise time parameter trise＝mint{t：|rn(t)|≥0.6rmax}-mint{t：|rn(t)|≥6σn}、
Average delay parameter
Figure FDA0002927886270000011
Root mean square delay parameter
Figure FDA0002927886270000012
Kurtosis parameter
Figure FDA0002927886270000013
Forming the 6 characteristic parameters into a characteristic set F ═ { e, r ═ emax，trise，τm，τr，κsR obtained from N positionsn(t) forming a training input matrix
Figure FDA0002927886270000014
Building a training output matrix
Figure FDA0002927886270000015
Wherein y isnIs rn(t) a sight-distance or non-sight-distance identification mark, and if it is a sight distance, ynWhen the non-line-of-sight distance is recognized as 1, ynGet the complete training set as 0
Figure FDA0002927886270000016
S2, measuring a received signal r from the target node to the base station*(t) and calculating r*6 characteristic parameters of (t);
s3, constructing a random forest model consisting of a plurality of decision trees, and for each decision tree: sampling N times from a complete training set D in a mode of putting back to sample to form a training set D' of the decision tree, randomly selecting M characteristic parameters from a characteristic set F as the characteristics to be selected of the decision tree, wherein M is less than 6, calculating the Gini index of the characteristics to be selected, and sequentially using the characteristics to be selected as the splitting characteristics of a root node, a middle node and a leaf node of the decision tree according to the sequence of the Gini index from small to large,
Figure FDA0002927886270000021
Figure FDA0002927886270000022
wherein Gini (D ', f) identifies and divides the training set D' into D according to the feature f to be selected1(stadia) and D2Gini index of two classes (non-line of sight), Gini (D)i) Is of class DiGini index, | D, of (i ═ 1, 2)1|、|D2Respectively is set D1、D2The number of samples in D', piIs of class DiIdentifying the probability of classification correctness and errors;
s4, receiving a signal r by a target node*(t) inputting the characteristic parameters of the random forest model, comparing the characteristic parameters with the splitting characteristic value interval range of each node by each decision tree, and outputting the line-of-sight/non-line-of-sight identification type of each decision tree, wherein most judgment results of all decision trees are used as the final line-of-sight/non-line-of-sight identification result of the base station;
s5, repeating the steps S1-S4 for the rest base stations in the test area to obtain the sight distance/non-sight distance identification result of the target node and the signals received by each base station, and removing the signals identified as the non-sight distance when positioning is carried out by utilizing the received signals.
2. A non-line-of-sight signal identification method based on random forests as claimed in claim 1 wherein A ≧ 3.
CN202110138933.7A 2021-02-01 2021-02-01 Random forest-based non-line-of-sight signal identification method Active CN112800983B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110138933.7A CN112800983B (en) 2021-02-01 2021-02-01 Random forest-based non-line-of-sight signal identification method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110138933.7A CN112800983B (en) 2021-02-01 2021-02-01 Random forest-based non-line-of-sight signal identification method

Publications (2)

Publication Number Publication Date
CN112800983A true CN112800983A (en) 2021-05-14
CN112800983B CN112800983B (en) 2024-03-08

Family

ID=75813467

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110138933.7A Active CN112800983B (en) 2021-02-01 2021-02-01 Random forest-based non-line-of-sight signal identification method

Country Status (1)

Country Link
CN (1) CN112800983B (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113382354A (en) * 2021-06-08 2021-09-10 东南大学 Wireless positioning non-line-of-sight signal discrimination method based on factor graph
CN113625319A (en) * 2021-06-22 2021-11-09 北京邮电大学 Non-line-of-sight signal detection method and device based on ensemble learning

Citations (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20090281981A1 (en) * 2008-05-06 2009-11-12 Chen Barry Y Discriminant Forest Classification Method and System
CN107124762A (en) * 2017-04-26 2017-09-01 玉林师范学院 A kind of wireless location method of efficient abatement non-market value
CN107273920A (en) * 2017-05-27 2017-10-20 西安交通大学 A kind of non-intrusion type household electrical appliance recognition methods based on random forest
CN107563437A (en) * 2017-08-31 2018-01-09 广州中海达定位技术有限公司 Ultra wide band non line of sight discrimination method based on random forest
CN107820206A (en) * 2017-11-15 2018-03-20 玉林师范学院 Non line of sight localization method based on signal intensity
US20180260531A1 (en) * 2017-03-10 2018-09-13 Microsoft Technology Licensing, Llc Training random decision trees for sensor data processing
KR20190110276A (en) * 2018-03-20 2019-09-30 한양대학교 산학협력단 Multi-sensor based noncontact sleep monitoring method and apparatus using ensemble of deep neural network and random forest
CN111257827A (en) * 2020-01-16 2020-06-09 玉林师范学院 High-precision non-line-of-sight tracking and positioning method
CN111832417A (en) * 2020-06-16 2020-10-27 杭州电子科技大学 Signal modulation pattern recognition method based on CNN-LSTM model and transfer learning
CN111916066A (en) * 2020-08-13 2020-11-10 山东大学 Random forest based voice tone recognition method and system
CN112245728A (en) * 2020-06-03 2021-01-22 北京化工大学 Respirator false positive alarm signal identification method and system based on integrated tree

Patent Citations (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20090281981A1 (en) * 2008-05-06 2009-11-12 Chen Barry Y Discriminant Forest Classification Method and System
US20180260531A1 (en) * 2017-03-10 2018-09-13 Microsoft Technology Licensing, Llc Training random decision trees for sensor data processing
CN107124762A (en) * 2017-04-26 2017-09-01 玉林师范学院 A kind of wireless location method of efficient abatement non-market value
CN107273920A (en) * 2017-05-27 2017-10-20 西安交通大学 A kind of non-intrusion type household electrical appliance recognition methods based on random forest
CN107563437A (en) * 2017-08-31 2018-01-09 广州中海达定位技术有限公司 Ultra wide band non line of sight discrimination method based on random forest
CN107820206A (en) * 2017-11-15 2018-03-20 玉林师范学院 Non line of sight localization method based on signal intensity
KR20190110276A (en) * 2018-03-20 2019-09-30 한양대학교 산학협력단 Multi-sensor based noncontact sleep monitoring method and apparatus using ensemble of deep neural network and random forest
CN111257827A (en) * 2020-01-16 2020-06-09 玉林师范学院 High-precision non-line-of-sight tracking and positioning method
CN112245728A (en) * 2020-06-03 2021-01-22 北京化工大学 Respirator false positive alarm signal identification method and system based on integrated tree
CN111832417A (en) * 2020-06-16 2020-10-27 杭州电子科技大学 Signal modulation pattern recognition method based on CNN-LSTM model and transfer learning
CN111916066A (en) * 2020-08-13 2020-11-10 山东大学 Random forest based voice tone recognition method and system

Non-Patent Citations (6)

* Cited by examiner, † Cited by third party
Title
周绮凤;杨小青;周青青;雷家艳: "基于随机森林的建筑结构损伤识别方法", 振动、测试与诊断, vol. 32, no. 2, 31 December 2012 (2012-12-31) *
杨小凤;陈铁军;刘峰: "高精度低复杂度的无线定位新方法", 计算机应用, vol. 34, no. 002, 31 December 2014 (2014-12-31) *
杨小凤;陈铁军;陈宇宁;: "基于TOA估计的超宽带快速定位方法", 现代雷达, no. 03, 15 March 2016 (2016-03-15) *
杨小凤;陈铁军;黄志文;李琼: "基于超宽带的TOA-DOA联合定位方法", 重庆邮电大学学报(自然科学版), no. 002, 31 December 2016 (2016-12-31) *
王鑫;汪晋宽;刘志刚;胡曦: "基于随机森林的认知网络主用户信号调制类型识别算法", 东北大学学报. 自然科学版, vol. 35, no. 12, 31 December 2014 (2014-12-31) *
谭正骄: "基于随机森林的通信信号调制识别算法研究", CNKI, 31 December 2018 (2018-12-31) *

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113382354A (en) * 2021-06-08 2021-09-10 东南大学 Wireless positioning non-line-of-sight signal discrimination method based on factor graph
CN113382354B (en) * 2021-06-08 2022-04-22 东南大学 Wireless positioning non-line-of-sight signal discrimination method based on factor graph
CN113625319A (en) * 2021-06-22 2021-11-09 北京邮电大学 Non-line-of-sight signal detection method and device based on ensemble learning
CN113625319B (en) * 2021-06-22 2023-12-05 北京邮电大学 Non-line-of-sight signal detection method and device based on ensemble learning

Also Published As

Publication number Publication date
CN112800983B (en) 2024-03-08

Similar Documents

Publication Publication Date Title
CN108696932B (en) Outdoor fingerprint positioning method using CSI multipath and machine learning
CN111832417B (en) Signal modulation pattern recognition method based on CNN-LSTM model and transfer learning
CN103945533B (en) Wireless real time position localization methods based on big data
CN112800983B (en) Random forest-based non-line-of-sight signal identification method
CN112135248B (en) WIFI fingerprint positioning method based on K-means optimal estimation
CN106851571A (en) WiFi localization methods in a kind of quick KNN rooms based on decision tree
CN112748397A (en) UWB positioning method based on self-adaptive BP neural network under non-line-of-sight condition
CN108566253B (en) It is a kind of based on the signal recognition method extracted to power spectrum signal fit characteristic
CN106941718A (en) A kind of mixing indoor orientation method based on signal subspace fingerprint base
CN107124762B (en) Wireless positioning method for efficiently eliminating non-line-of-sight errors
CN113759323B (en) Signal sorting method and device based on improved K-Means joint convolution self-encoder
CN110222776A (en) A kind of indoor Passive Location based on CSI finger print information
CN110376290A (en) Acoustic emission source locating method based on multidimensional Density Estimator
CN113269825A (en) Forest breast diameter value extraction method based on foundation laser radar technology
CN107396309A (en) A kind of wireless sensor network forest localization method
CN109141420B (en) Indoor positioning method based on crowdsourcing data
CN108989988B (en) Indoor positioning method based on machine learning
CN112801268B (en) Positioning method based on graph convolution and multilayer perceptron hybrid network
CN112990125A (en) Method for judging whether radiation source radar belongs to target platform
Cheng et al. A fingerprint localization method based on weighted KNN algorithm
CN115442887B (en) Indoor positioning method based on RSSI of cellular network
CN116466314A (en) Radar individual sorting method, device, storage medium and equipment
CN115017958A (en) Non-line-of-sight signal identification method based on channel characteristic weighting model
CN115567871A (en) WiFi fingerprint indoor floor identification and position estimation method
CN114584230A (en) Predictive channel modeling method based on countermeasure network and long-short term memory network

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant