CN105738866A - 60GHz Non-Line-of-Sight identification and wireless fingerprint positioning method based on energy detection - Google Patents

60GHz Non-Line-of-Sight identification and wireless fingerprint positioning method based on energy detection Download PDF

Info

Publication number
CN105738866A
CN105738866A CN201610065460.1A CN201610065460A CN105738866A CN 105738866 A CN105738866 A CN 105738866A CN 201610065460 A CN201610065460 A CN 201610065460A CN 105738866 A CN105738866 A CN 105738866A
Authority
CN
China
Prior art keywords
nlos
thresholding
signal
environment
toa
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
CN201610065460.1A
Other languages
Chinese (zh)
Other versions
CN105738866B (en
Inventor
梁晓林
张�浩
刘兴
吕婷婷
魏兆强
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.)
Ocean University of China
Original Assignee
Ocean University of China
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 Ocean University of China filed Critical Ocean University of China
Priority to CN201610065460.1A priority Critical patent/CN105738866B/en
Publication of CN105738866A publication Critical patent/CN105738866A/en
Application granted granted Critical
Publication of CN105738866B publication Critical patent/CN105738866B/en
Expired - Fee Related legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
    • G01S5/00—Position-fixing by co-ordinating two or more direction or position line determinations; Position-fixing by co-ordinating two or more distance determinations
    • G01S5/02—Position-fixing by co-ordinating two or more direction or position line determinations; Position-fixing by co-ordinating two or more distance determinations using radio waves
    • G01S5/06—Position of source determined by co-ordinating a plurality of position lines defined by path-difference measurements

Landscapes

  • Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Remote Sensing (AREA)
  • Position Fixing By Use Of Radio Waves (AREA)
  • Mobile Radio Communication Systems (AREA)

Abstract

The invention discloses a 60GHz Non-Line-of-Sight identification and wireless fingerprint positioning method based on energy detection. The method provided by the invention is characterized in that 1) a combined parameter J constituted by skewness, leptokurtosis, maximum curl, and maximum slope of a signal, an optimized normalized threshold X, and a parameter D constituted by maximum curl and maximum slope can be acquired; 2) the fingerprint database between the J and the X can be established; 3) the optimized threshold can be estimated according to the J by adopting the fingerprint database; 4) the Non-Line-of-Sight identification can be carried out by adopting the D; 5) the distance can be calculated by adopting the TOA estimation; 6) the 60GHz wireless positioning can be carried out: the wireless positioning based on the 60GHz signal can be carried out by adopting the conventional positioning algorithm according to the Non-Line-of-Sight identification result and the TOA estimation value. The result shows that 1) the NLOS identification success rate is higher than the Non-Line-of-Sight identification algorithm of the same type; 2) in the Line-of-Sight environment and the Non-Line-of-Sight environment, and in the great signal to noise ratio range, the method provided by the invention has the higher precision and the better robustness by comparing with other methods based on the energy detection.

Description

A kind of 60GHz non line of sight identification based on energy measuring and wireless fingerprint localization method
Technical field
The invention belongs to wireless location technology field, specifically a kind of 60GHz non line of sight identification based on energy measuring and wireless fingerprint localization method.
Background technology
Pulse 60GHz wireless communication technology is a kind of without carrier wave, adopts a kind of wireless communication technology that the discrete pulse of hundred psecs or shorter duration communicates.It is high that 60GHz wireless communication technology has frequency spectrum reusability compared with current existing communication system, and capacity of resisting disturbance is strong, available frequency spectrum width, it is allowed to transmitting power is big, and power system capacity is big, temporal resolution and multi-path resolved rate advantages of higher.In recent years one of 60GHz topmost reason of technology extensive concern is because huge licensing free frequency band bandwidth.Compared with the same super-broadband tech using licensing free frequency band, the frequency band of 60GHz technology is continuous, and Power Limitation is less.Owing to radio ultra wide band system is symbiotic system, to be therefore subject to strict restriction and the constraint of different regulations.Bandwidth huge for 60GHz is to be about to maximum one piece of licensing free frequency band of distribution.Huge bandwidth means potential capacity and motility, so that 60GHz technology is particularly suitable for gigabit wireless application.Impulse radio communications technology near 60GHz frequency range is owing to having higher temporal resolution, thus at receiving terminal, can more effectively separate multipath signal, thus having higher multi-path resolved rate, it is possible to achieve centimetre even grade precision distance measurement and location.This Indoor Robot precision navigation location and some special producing industries (need not people maybe can not someone participate) etc. need the pinpoint field of centimetre rank there is important using value.
In order to realize the wireless location of 60GHz, relevant hardware device is mainly made up of mobile terminal undetermined, locating base station and location-server.
Mobile terminal to be positioned is mobile in region, location, it is necessary to the terminal of location, is usually the discharger that power is low.
Locating base station is by the locating base station being distributed in region, location, the 60GHz signal that terminal to be positioned sends can be received, and carry out degree of bias S, kurtosis K, maximum curl C and the isoparametric calculating of greatest gradient MS, utilize the fingerprint database of design in advance, the propagation delay of signal calculated, finally can be sent to location-server by value of calculation.Generally by the locating base station of more than three.
Location-server is usually a computer, it is possible to receives and comes from the propagation delay that locating base station sends, and it is carried out data process, performs location algorithm.
Location technology the most frequently used at present is all based on greatly finding range and carries out, this is because, the general positioning precision of location technology being not based on distance is poor, and needs the cooperation of substantial amounts of base station (terminal that position is known).The most frequently used localization method can be divided into based on receiving TOA (TimeofArrival) and TDOA (TimeDifferenceofArrival), the RSS (ReceivedSignalStrength) based on received signal strength estimation and the AOA (AngleofArrival) based on angle of arrival estimation that time of arrival (toa) is estimated.Pulse 60GHz signal has high bandwidth, and the persistent period reaches hundred psecs or shorter, thus has very strong time resolution.So this characteristic strong in order to make full use of pulse 60GHz time resolution, the location technology that TOA, TDOA estimate is used to be best suited for pulse 60GHz's.Both approaches affects the impact of measurement that the principal element of measurement error is propagation delay time and non line of sight (NLOS) environment.
TOA TDOA method of estimation the most frequently used at present generally can be divided into correlation reception (as matched filtering detects) and irrelevant reception (such as energy receiver).Coherent detection based on matched filtering, it is believed that be the best mode for signal detection being currently known, but, it needs about the prior information (such as, modulation format, impulse waveform, phase place etc.) launching characteristics of signals.But in practice, such information can not always accurately be predicted by receiver often, this results in the correlation receiver based on matching detection is infeasible in many cases.Different from correlation reception, based on the energy detection priori completely without signal, and there is the complexity of relatively low calculating and enforcement, the hardware requirement of docking point is low, being suitably applied in the node of simple in construction, based on the plurality of advantages of energy receiver, energy detector has been widely used as the cognitive radio of spectrum-sensing, impulse radio ultra-wideband system, sensor network and Terrestrial trunked radio.Energy receiver mainly includes amplifier, squarer, integrator, a decision device.Owing to the frequency spectrum of pulse 60GHz is in higher frequency range, so matched filter detector is proposed higher requirement on hardware realizes, in actual applications, it is relatively difficult to realize.Therefore in the present invention, to the detection of signal will first-selected complexity lower, realize requiring less energy-detection receiver to hardware.The TOA of energy-detection receiver (as shown in Figure 1) estimates mainly the output of integrator to be compared with suitable threshold value, selects to exceed at first threshold value and obtains the value of energy block TOA is estimated.In NLOS identification, current NLOS recognizer majority is based on channel characteristic estimation and utilizes the mode of correlation reception that signal is processed.As mentioned above, at 60GHz wireless communication field, lot of challenges is there is in correlation receiver on hardware realizes, cannot realize smoothly, so the NLOS recognizer based on correlation reception cannot effectively be used on energy receiver, and the NLOS recognizer being currently based on energy acceptance cannot be effectively guaranteed in accurately identifying NLOS.
Traditional TOA the basic step following (as shown in Figure 2) of TDOA location algorithm:
(1), whole alignment system is initialized: the soft and hardware mainly including each base station and location-server is installed；
(2), terminal transmission 60GHz pulse train to be positioned；
(3), locating base station receives the propagation delay of signal signal calculated；
(4), propagation delay result of calculation is sent to location-server by locating base station；
(5), location-server receives the propagation delay of each base station；
(6), location-server calculates the range measurement of each base station；
(7), terminal to be positioned is positioned by location-server application TOA TDOA based on the location algorithm of distance.
In view of the huge difference between correlation reception and irrelevant reception, particularly complexity energy receiver low, low sampling rate can be widely applied in numerous environment, so the simple and practical energy receiver that hardware requirement is low will be adopted in (3) to identify to calculate propagation delay and NLOS, estimated result and the NLOS recognition result of propagation delay will be transferred to location-server, and terminal to be positioned is positioned by the information comprehensively utilizing these two aspects at location-server end.In energy acceptance, the method estimating propagation delay conventional at present can be divided into two kinds.
Maximum energy method: select maximum energy block correspondence position to estimate TOA, it is common that select central authorities' estimated value as TOA of energy block.But, the position at the frequent also non line of sight place, position that ceiling capacity block is corresponding, particularly in a nlos environment.The through energy block through place is before ceiling capacity block of being everlasting on average.
Threshold method: namely based on the TOA algorithm for estimating of thresholding, the energy block and the suitable thresholding that receive signal compare, and the moment that first energy block exceeding this thresholding is corresponding is TOA estimated value.But, to directly determine that a threshold value is relatively difficult, thus through frequently be normalized thresholding.After obtaining normalized thresholding, at receiving terminal, according to formula α=αnorm(max (z [n])-min (z [n]))+min (z [n]) calculates final threshold value.Problem just becomes how that the fingerprint characteristic according to signal sets suitable normalization thresholding, simplest in threshold method is fixed threshold method, and its thresholding is a fixing value, but in actual applications, varying environment Lower Threshold all the time be change, so cannot on a large scale in be applied.Next is the normalization threshold method based on kurtosis, although this algorithm complex reduces, but proposition based on gradient in these algorithms and the present invention, standard deviation is compared with the associating TOA fingerprint algorithm for estimating of the degree of bias, no matter in precision or in stability particularly multipath, have very big gap under nlos environment.And the NLOS recognizer based on energy measuring that we are proposed in invention all improves based on the NLOS recognizer of energy measuring with other in precision or in stability.Especially, NLOS correct recognition rata reaches the discrimination based on channel characteristic estimation algorithm in some cases.
Summary of the invention
For existing technological deficiency, the present invention proposes a kind of 60GHz non line of sight identification based on energy measuring and wireless fingerprint localization method, to overcome the deficiencies in the prior art.
A kind of 60GHz non line of sight identification based on energy measuring and wireless fingerprint localization method, comprise the following steps:
(1), alignment system is set up, involved alignment system includes multiple multiple locating base station being able to receive that signal that terminal to be positioned sends, and receive the location-server of location information that locating base station sends, and whole alignment system is initialized: include setting sample frequency and the integration period T of each locating base station；
(2), terminal transmission 60GHz pulse sequence signal to be positioned；
(3), locating base station receives propagation delay and the NLOS identification of above-mentioned signal signal calculated；
(4), propagation delay result of calculation and NLOS recognition result are sent to location-server by locating base station；
(5), location-server receives propagation delay and the NLOS recognition result of each base station；
(6), location-server calculates the range measurement of each base station；
(7), terminal to be positioned is positioned by location-server application TOA TDOA based on the location algorithm of distance；
It is characterized in that described step (3) is that locating base station receives above-mentioned signal, this signal is integrated computing and obtains integral energy block, and then obtain combined parameters value, calculate optimum gate limit value further according to combined parameters value, choose the propagation delay that the moment corresponding to the center of the energy block exceeding this threshold value at first is signal；Including tri-steps of following A-C:
A. the signal of step (2) is integrated computing and obtains integral energy block by locating base station, calculate the degree of bias S of this energy block, kurtosis K, maximum curl C and greatest gradient MS, and each variable above-mentioned is normalized, by each variable after normalization and then obtain combined parameters J and maximum curl/greatest gradient product D, try to achieve the fingerprint database setting up combined parameters meansigma methods J2P, TOA estimation difference, optimum normalization tri-parameters of thresholding X；
B. fingerprint database is carried out curve fitting, set up the corresponding relation F of the average combined parameters J2P corresponding to minimum TOA estimation difference and optimum normalization thresholding X；
C. according to the step A average combined parameters J2P obtained and maximum curl/greatest gradient product D, D is utilized to compare with the NLOS threshold value set in advance, judge that signal comes from LOS environment or NLOS environment, recognition result is preserved, utilize corresponding relation F, calculating obtains optimum normalization thresholding X, obtains propagation delay (i.e. TOA estimated value) according to this thresholding；
Specifically, step A is refined as following calculation procedure:
1), first setup parameter value, a signal to noise ratio snr is selected within the scope of 4-32dB, then under a selected SNR, determine different channel circumstances and multiple different integration period, described different channels environment is sighting distance and two kinds of varying environments of non line of sight, described multiple different integration periods are to select two or more value as integration period within the scope of 0.1ns 4ns, the quantity of selected different integration periods is designated as P, P and is greater than being equal to the natural number of 2；2P different environment and integration period combination then can be obtained at same SNR；
2) energy block, according to integral operation obtained, calculates 2P different environment and the degree of bias S of energy block of integration period combination, kurtosis K, maximum curl C and greatest gradient MS respectively；Calculate the ratio of degree of bias S and kurtosis K, be denoted as KS=K/S；Calculate the product of maximum curl C and greatest gradient MS, be denoted as D=C*MS；
A new combined parameters J=N*norm (KS)-M*norm (D) is measured according to KS and C two, wherein norm represents the normalized to parameter, N, M are that arithmetic number and N are be more than or equal to 6M, obtain 2P combined parameters J, average and be designated as average combined parameters J2P；
3) under same SNR, then calculate the optimum normalization thresholding under different channel circumstances and multiple different integration period:
First TOA estimation difference and best normalization thresholding are calculated:
Using the interval of (0:0.1:1) or less as normalization thresholding, calculate the integral energy block 1000 TOA errors under each thresholding respectively, and average as TOA estimation difference, thus obtaining the multiple TOA estimation differences corresponding with normalization threshold number, choose the minimum normalization thresholding corresponding to TOA error as best normalization thresholding；
Then can obtain 2P best normalization thresholding under different channels environment (sighting distance and non line of sight), different integration period, using the meansigma methods of 2P the best normalization thresholding as optimization thresholding X；
4) step 1 is returned) select next signal to noise ratio, and recalculate corresponding to the average combined parameters J2P under this signal to noise ratio, TOA estimation difference and optimization thresholding X, until all signal to noise ratios gone through within the scope of 4-32dB；
5) by step 4) value of 29 groups of average combined parameters J2P, TOA estimation difference and optimization thresholding X that obtains, as the fingerprint database being made up of three parameters；
Step B, fingerprint database is carried out curve fitting, utilize neutral net that above-mentioned fingerprint database is trained, finally set up the corresponding relation F of average combined parameters J2P and optimum normalization thresholding X, namely owing to average combined parameters J2P and SNR is relevant, and optimum normalization thresholding is calculated under certain specific SNR, the corresponding relation of J and optimum normalization thresholding therefore can be set up；
Step C, when signal transmission delay is carried out Practical Calculation, actual average combined parameters J2P and parameter D is obtained according to the degree of bias S of actual signal gathered, kurtosis K, maximum curl C and greatest gradient MS, D is utilized to compare with the NLOS threshold value set in advance, judge that signal comes from LOS environment or NLOS environment, recognition result is preserved；Utilize corresponding relation F, calculate the normalization thresholding obtained corresponding to this actual average combined parameters J2P, obtain TOA estimated value according to this normalization thresholding: be input to step B by the actual average combined parameters J2P of gained) own trained good neutral net, corresponding normalization thresholding can be obtained according to corresponding relation F, utilizing normalization thresholding to identify the energy block exceeding this thresholding at first, the moment corresponding using the centre position of this energy block is as TOA estimated value.
In step A) formula J=N*norm (KS)-M*norm (D) in, in order to make the numerical value of J in coordinate diagram picture, there is stable change in each per channel is than scope, can by selecting suitable coefficient N, M to be achieved, wherein above-mentioned N value is less than 20.
For the sake of simplicity, step A) formula J=N*norm (KS)-M*norm (D) in, N value 12, M value 2.
In above-mentioned steps C, during according to the D of gained to Practical Calculation, terminal transmission 60GHz pulse sequence signal to be positioned carries out the step of NLOS identification, specific as follows:
1) thresholding is set according to formula (1),
In IEEE802.15.3c60GHz channel model, 1000 channel impulse responses of each generation under LOS and NLOS environment respectively, corresponding 60GHz signal all can be obtained for each channel impulse response, then each 60GHz signal is all calculated and obtain corresponding maximum curl C and greatest gradient MS, and then obtain maximum curl/greatest gradient product D=C*MS, namely under LOS and NLOS environment, obtain the sample value of 1000 gradients/standard deviation product D respectively, be denoted as D respectivelyLOSWith DNLOS, for getting rid of the extremum of the minority of the existence impact on determining thresholding in sample value, respectively by DLOSWith DNLOSArrange by ascending order, choose DLOSWith DNLOSThe sample value of front 90% is denoted as respectivelyWithCalculate respectivelyWithMeansigma methods be denoted asWithChooseWithIn less value identify required thresholding α as final NLOSD；
2) utilize calculated parameter D to compare with the thresholding set in advance and judge that terminal transmission 60GHz pulse sequence signal to be positioned comes from LOS environment or NLOS environment；NLOS recognizer is expressed as:
TX=360 ° (2)
TX=360 ° (3)
TX < 360 ° (4)
TX < 360 ° (5)
When transmitting antenna is omnidirectional antenna (TX=360 °), as maximum curl D≤αDThen think that signal comes from LOS environment, otherwise, as maximum curl D > αDThen think that signal comes from NLOS environment；When transmitting antenna is non-omnidirectional antenna (TX < 360 °), as maximum curl D≤αDThen think that signal comes from NLOS environment, otherwise, as maximum curl D > αDThen think that signal comes from LOS environment.
3) in above-mentioned steps (7), utilize above-mentioned steps 2) LOS/NLOS recognition result, location-server will preferentially utilize the TOA estimated value under LOS environment, and in conjunction with the TOA estimated value under nlos environment, terminal to be positioned is positioned by application TOA TDOA based on the location algorithm of distance, thus obtaining more accurate positioning result.
Invention advantage
In the present invention, use energy receiver that the communication environments (sighting distance and non line of sight) of signal is estimated with propagation delay, the utilization of energy receiver solves the necessary prior information to transmitting signal of correlation receiver (such as modulation format, impulse waveform, phase place etc.) shortcoming that cannot accurately estimate.The proposed NLOS recognizer based on energy measuring, makes the NLOS accuracy rate identified bring up to 80% far above the accurate rate of the current NLOS recognizer 60% based on energy measuring under major part channel.Proposed combined parameters is simultaneously independent of integration period and channel circumstance (sighting distance and non line of sight).Overcome traditional signal transmission delay algorithm for estimating based on energy measuring and have to differentiate between this shortcoming of integration period, use artificial neural network to solve nonlinear problem simultaneously, make the non-linear relation between optimum normalization thresholding and combined parameters more accurate, overcome traditional curve matching and cannot accurately estimate this shortcoming of non-linear relation between input variable and output variable.
Accompanying drawing explanation
Fig. 1 energy receiver schematic diagram.
Localization method flow chart traditional for Fig. 2.
Fig. 3 normalized parameter situation of change.
Fig. 4 combined parameters situation of change to signal to noise ratio.
The situation of change of Fig. 5 optimum normalization thresholding.
The flow chart of steps of Fig. 6 present invention.
The overview flow chart of Fig. 7 present invention
Detailed description of the invention
The method of the present invention, mainly in step (3), adopts the mode of energy acceptance to carry out the estimation of TOA and NLOS state recognition mainly includes the following steps that (such as Fig. 6):
A. acquired integrated energy block, calculates degree of bias S, kurtosis K, maximum curl C and greatest gradient MS, calculates 1 respectively >, maximum curl C and greatest gradient MS asks product be denoted as D carry out NLOS identification；2 >, degree of bias S, kurtosis K, maximum curl C and greatest gradient MS being normalized, comprehensive degree of bias S, kurtosis K, maximum curl C and greatest gradient MS obtain combined parameters J, finally give average combined parameters J2P；Set up the fingerprint database of average combined parameters J2P, TOA estimation difference and optimum normalization tri-parameters of thresholding X；
B. carry out curve fitting, set up the corresponding relation F of the average combined parameters J2P corresponding to minimum TOA estimation difference and optimum normalization thresholding X；
C. average combined parameters J2P is obtained according to the degree of bias S of live signal gathered, kurtosis K, maximum curl C and greatest gradient MS, utilize corresponding relation F, calculating obtains optimum normalization thresholding, according to optimum normalization thresholding, obtain TOA estimated value, send TOA estimated value and NLOS recognition result to location-server.
Specifically, A. acquired integrated energy block, calculate degree of bias S, kurtosis K, maximum curl C and greatest gradient MS, calculate 1 respectively >, maximum curl C and greatest gradient MS asks product be denoted as D carry out NLOS identification；2 >, degree of bias S, kurtosis K, maximum curl C and greatest gradient MS being normalized, comprehensive degree of bias S, kurtosis K, maximum curl C and greatest gradient MS obtain combined parameters J, finally give average combined parameters J2P；Set up the fingerprint database of average combined parameters J2P, TOA estimation difference and optimum normalization tri-parameters of thresholding X；Can specifically be refined as following calculation procedure:
1. the energy block, according to energy measuring collected, calculates the degree of bias S of energy block, kurtosis K, maximum curl C and greatest gradient MS respectively.Maximum curl C and greatest gradient MS is utilized to carry out NLOS state recognition, find in very big SNR range when IEEE802.15.3c channel emulates, when transmitting antenna is omnidirectional antenna, the maximum curl maximum under LOS environment is still just also little than the maximum curl minima under NLOS environment.Equally, when the non-omnidirectional antenna of transmitting antenna, the maximum curl minima under LOS environment is still just also big than the maximum curl maximum under NLOS environment.So, NLOS can be identified accurately completely as long as setting suitable threshold value.NLOS recognizer is expressed as:
TX=360 ° (6)
TX=360 ° (7)
TX < 360 ° (8)
TX < 360 ° (9)
When transmitting antenna is omnidirectional antenna (TX=360 °), as maximum curl D≤αDThen think that signal comes from LOS environment, otherwise, as maximum curl D > αDThen think that signal comes from NLOS environment；When transmitting antenna is non-omnidirectional antenna (TX < 360 °), as maximum curl D≤αDThen think that signal comes from NLOS environment, otherwise, as maximum curl D > αDThen think that signal comes from LOS environment.The calculation of threshold value is:
In IEEE802.15.3c60GHz channel model, 1000 channel impulse responses of each generation under LOS and NLOS environment respectively, corresponding 60GHz signal all can be obtained for each channel impulse response, then each 60GHz signal is all calculated and obtain corresponding maximum curl C and greatest gradient MS, and then obtain maximum curl/greatest gradient product D=C*MS, namely under LOS and NLOS environment, obtain the sample value of 1000 gradients/standard deviation product D respectively, be denoted as D respectivelyLOSWith DNLOS, for getting rid of the extremum of the minority of the existence impact on determining thresholding in sample value, respectively by DLOSWith DNLOSArrange by ascending order, choose DLOSWith DNLOSThe sample value of front 90% is denoted as respectivelyWithCalculate respectivelyWithMeansigma methods be denoted asWithChooseWithIn less value identify required thresholding α as final NLOSD；
2., the degree of bias S of sample, kurtosis K, maximum curl C and greatest gradient MS being normalized, obtained result is in figure 3.Result shows no matter be under LOS and NLOS environment, and result display KS=K/S, S and K increase along with the increase of SNR but KS changes faster, and same D=C*MS, C, MD=MS*SD, MS and SD reduce along with SNR reduction but D changes faster.Because KS and D changes faster, SNR information more can be reflected, so being more suitable for for selecting threshold value.Find as SNR simultaneously > 9dB time KS change faster, but during less than 9dB, KS change is relatively slow, but now D change is very fast；It is contrary that as SNR, < during 9dB, D changes faster, but during more than 9dB, K change is relatively slow, but now KS change is very fast.So relying solely on single variable cannot react the situation of change of SNR under any SNR accurately.Therefore, obtaining a new combined parameters J=N*norm (KS)-M*norm (D) according to KS, D, and then cut to obtain the meansigma methods J2P of combined parameters, simulation result shows, works as NLOS≠NNLOSWith MLOS≠MNLOSTime, average combined parameters J2P, independent of channel model, it is limited only by the impact of integration period, but in actual applications, under various circumstances, integration period can set at random, now this algorithm certainly will cannot be widely used in various environment well；Work as NLOS=NNLOSAnd MLOS=MNLOSTime, average combined parameters J2P, simultaneously independent of channel model and integration period, now will need not consider change such as Fig. 4 of integration period.It is a monotonically increasing function relative to SNR that Fig. 4 is shown in the interior average combined parameters J2P of all of SNR range, and therefore it is more sensitive to SNR than any single parameter.Calculating respectively under identical SNR environment, the estimation difference of the different TOA corresponding to normalization thresholding (such as [0:0.1:1]), choosing the normalization thresholding corresponding to minimum TOA error is best normalization thresholding.Owing to channel model and integration step are little on average combined parameters J2P impact, so taking the meansigma methods of different channels difference integration step when setting up corresponding relation as optimization thresholding such as Fig. 5.
Specifically step B " fingerprint database is carried out curve fitting, set up the corresponding relation F of the average combined parameters J2P corresponding to minimum TOA estimation difference and optimum normalization thresholding X " can be detailed be expressed as:
In recent years, artificial neural network is used widely in signal processing field, owing to inevitably there are NLOS, multipath, reflection, intersymbol interference, diffraction, decline etc. in actual environment, that is location terminal is often nonlinear with the distance of locating base station or the position at angle and terminal place, location, very difficult geometric formula directly calculates, and neutral net exactly has the non-linear mapping capability of height.So neutral net is used for determining the average combined parameters J2P corresponding relation with optimum normalization thresholding X.Using average combined parameters J2P as the input layer of neutral net, optimum normalization thresholding X is as the output layer of neutral net, and when determining the number of neutral net hidden layer neuron, the distribution probability according to mean square deviation is estimated.Select mean square deviation less than 10-10Ratio more than 90% time corresponding neuronic number be hidden layer neuron number.Finally determine the corresponding relation of average combined parameters J2P and optimum normalization thresholding X.
What step C " according to the degree of bias S of live signal gathered, kurtosis K, maximum curl C and greatest gradient MS; try to achieve variables D and average combined parameters J2P; utilize D to obtain NLOS recognition result compared with the threshold value set in advance; and utilize corresponding relation F; calculate and obtain optimum normalization thresholding X; obtain TOA estimated value according to this thresholding, and send NLOS recognition result to location-server " can be detailed specifically is expressed as:
Certain integration step is adopted to be integrated obtaining several energy blocks the signal collected, try to achieve average combined parameters J2P, D is utilized to obtain NLOS recognition result compared with the threshold value set in advance, average combined parameters J2P is input to the neutral net trained simultaneously, corresponding optimum normalization thresholding X can be obtained, optimum normalization thresholding X is utilized to obtain first energy block exceeding this thresholding, the moment corresponding using the centre position of this energy block is as TOA estimated value, and sends NLOS recognition result to location-server.
The method is adopted to study under the IEEE802.15.3c channel model provided, find it is no matter that (closely, LOS, transmitting signal power big etc.) or communication condition bad (distance remote (< 20m), NLOS, transmitting signal power are low under the environment that communication condition is good) environment under, the accuracy of the result of calculation of propagation delay can be greatly improved, thus ensureing the accuracy of range measurement after using above-mentioned steps.Such as, it is the mean conditions of error after the various TOA based on energy acceptance method estimates to measure at 1000 times shown in table 1.It appeared that the result of the present invention will far better than other algorithm.Giving the accuracy rate of the NLOS recognizer based on energy measuring in table 2, result shows, under indoor environment, to most channel circumstance accuracy rate more than 85%.
The various energy acceptance Algorithm Error of table 1 compares (ns)
Table 2NLOS state recognition rate
SNR(dB) 16 18 20 22 24 26 28 30 32 34
CM1.1 99.3 99.3 99.3 99.3 99.3 99.3 99.4 99.4 99.4 99.4
CM2.1 78.9 79.6 81 81.4 80.7 81.4 81.3 81.5 81.5 81.7
CM1.2 100 100 100 100 100 100 100 100 100 100
CM2.2 84.8 86 86.3 86.3 86.4 87 87.3 87.1 87.2 87.6
CM1.3 100 100 100 100 100 100 100 100 100 100
CM2.3 87.3 88.8 87.8 88.3 88.5 88.8 90.6 91.4 90.0 90.0
CM1.4 100 100 100 100 100 100 100 100 100 100
CM2.4 85.9 87.1 87.2 88.2 87.5 87.6 88.2 87.3 88.2 87.5
Embodiment
When carrying out wireless location, terminal to be positioned is arranged according to it, and timing sends multiple 60GHz pulse train, in order to repeatedly measure.All locating base station receiving this pulse train, average combined parameters J2P is obtained by energy acceptance, utilize the identification only carrying out NLOS state of K, and obtain the estimated value of optimum normalization thresholding X according to the neutral net that precondition is good, in order to finally give TOA estimated value；NLOS recognition result and TOA estimated result are transferred to location-server；Then at location-server end, according to NLOS recognition result, measure the distance of gained or the coordinate position of range difference and reference base station, utilize TOA or TDOA location algorithm to determine the locus of terminal to be measured.As it is shown in fig. 7, mainly include following step:
(1), system initialization
System initialization, including installation and the relevant configuration of software and hardware.
The installation of base station: if two-dimensional localization, then at least need 3 locating base station；If three-dimensional localization then at least needs 4 locating base station.
The installation of location-server: require to be able to receive that the signal transmission delay that each base station sends at location-server end.Location-server requires function admirable, because location algorithm mainly runs on that server.
Location-server is arranged: the fingerprint database required for the location locating periodically of terminal, locating base station, the range finding number of times (sending the number of pulse train) every time positioned, the clock skew of each base station, signal velocity etc., and it is sent to terminal to be positioned by wireless transmission method, complete the setting to location terminal.
(2), the band location multiple 60GHz pulse train of terminal transmission
When terminal to be positioned to position, multiple pulse trains will be sent according to the value that arranges in advance.Each pulse train completes the estimation (namely distance estimations) of a normalization thresholding, and completing one-time positioning needs repeatedly to find range.
(3), locating base station receives signal signal calculated propagation delay
1., acquired integrated energy block, calculate the degree of bias, kurtosis, maximum curl and greatest gradient, calculate the value of D respectively and each variable is normalized, obtaining combined parameters；
According to the energy block that energy measuring collects, calculate the degree of bias S of energy block, kurtosis K, maximum curl C and greatest gradient MS respectively.Calculate the degree of bias and kurtosis compares KS=K/S.Combined parameters J=N*norm (KS)-M*norm (D) is obtained according to KS, D.In order to obtain better using the algorithm proposed in the present invention, when data are designed, set NLOS=NNLOSAnd MLOS=MNLOS。
2., identify required threshold value according to the NLOS set in advance in data base, utilize the D obtained to compare with it, carry out the identification of NLOS.
3., when reality is measured, return above-mentioned steps (2), then the average combined parameters J2P obtained that carries out curve fitting according to pre-determined fingerprint database calculates with the corresponding relation of optimum normalization thresholding X and obtains optimum normalization thresholding, finally give suitable threshold value, obtain the central value of the energy block exceeding this threshold value at first and integration period product as required TOA estimated value.
(4), propagation delay result of calculation and NLOS recognition result are sent to location-server by locating base station；
(5), location-server receives propagation delay and the NLOS recognition result of each base station；
According to the fingerprint database set before utilizing in (3), optimum the tried to achieve threshold value of normalization thresholding that average combined parameters J2P obtains, obtains the central value of the energy block exceeding this threshold value at first and integration period product as required TOA estimated value.
(6), location-server calculates range measurement and the NLOS recognition result of each base station；
Utilize the TOA estimated value tried to achieve in (5) to deduct owing to the clock skew sent and reception causes is multiplied by signal velocity again, be the range measurement of this locating base station.
(7), terminal to be positioned is positioned by location-server application TOA TDOA based on the location algorithm of distance.
According to the range measurement that all base stations are transmitted, calculate the coordinate at terminal place to be positioned.Its method mainly has TOA, TDOA etc., owing to location algorithm is not belonging to the content that this invention is protected, so not being described in detail at this.

Claims (4)

1., based on the 60GHz non line of sight identification of energy measuring and a wireless fingerprint localization method, comprise the following steps:
(1), alignment system is set up, involved alignment system includes multiple multiple locating base station being able to receive that signal that terminal to be positioned sends, and receive the location-server of location information that locating base station sends, and whole alignment system is initialized: include setting sample frequency and the integration period T of each locating base station；
(2), terminal transmission 60GHz pulse sequence signal to be positioned；
(3), locating base station receives propagation delay and the NLOS identification of above-mentioned signal signal calculated；
(4), propagation delay result of calculation and NLOS recognition result are sent to location-server by locating base station；
(5), location-server receives propagation delay and the NLOS recognition result of each base station；
(6), location-server calculates the range measurement of each base station；
(7), terminal to be positioned is positioned by location-server application TOA TDOA based on the location algorithm of distance；
It is characterized in that described step (3) includes tri-steps of following A-C:
A. the signal of step (2) is integrated computing and obtains integral energy block by locating base station, calculate the degree of bias S of this energy block, kurtosis K, maximum curl C and greatest gradient MS, and each variable above-mentioned is normalized, by each variable after normalization and then obtain combined parameters J and maximum curl/greatest gradient product D, try to achieve the fingerprint database setting up combined parameters meansigma methods J2P, TOA estimation difference, optimum normalization tri-parameters of thresholding X；
B. fingerprint database is carried out curve fitting, set up the corresponding relation F of the average combined parameters J2P corresponding to minimum TOA estimation difference and optimum normalization thresholding X；
C. according to the step A average combined parameters J2P obtained and maximum curl/greatest gradient product D, D is utilized to compare with the NLOS threshold value set in advance, judge that signal comes from LOS environment or NLOS environment, recognition result is preserved, utilize corresponding relation F, calculating obtains optimum normalization thresholding X, obtains propagation delay (i.e. TOA estimated value) according to this thresholding；
Specifically, step A is refined as following calculation procedure:
1), first setup parameter value, a signal to noise ratio snr is selected within the scope of 4-32dB, then under a selected SNR, determine different channel circumstances and multiple different integration period, described different channels environment is sighting distance and two kinds of varying environments of non line of sight, described multiple different integration periods are to select two or more value as integration period within the scope of 0.1ns 4ns, the quantity of selected different integration periods is designated as P, P and is greater than being equal to the natural number of 2；2P different environment and integration period combination then can be obtained at same SNR；
2) energy block, according to integral operation obtained, calculates 2P different environment and the degree of bias S of energy block of integration period combination, kurtosis K, maximum curl C and greatest gradient MS respectively；Calculate the ratio of degree of bias S and kurtosis K, be denoted as KS=K/S；Calculate the product of maximum curl C and greatest gradient MS, be denoted as D=C*MS；
A new combined parameters J=N*norm (KS)-M*norm (D) is measured according to KS and C two, wherein norm represents the normalized to parameter, N, M are that arithmetic number and N are be more than or equal to 6M, obtain 2P combined parameters J, average and be designated as average combined parameters J2P；
3) under same SNR, then calculate the optimum normalization thresholding under different channel circumstances and multiple different integration period:
First TOA estimation difference and best normalization thresholding are calculated:
Using the interval of (0:0.1:1) or less as normalization thresholding, calculate the integral energy block 1000 TOA errors under each thresholding respectively, and average as TOA estimation difference, thus obtaining the multiple TOA estimation differences corresponding with normalization threshold number, choose the minimum normalization thresholding corresponding to TOA error as best normalization thresholding；
Then can obtain 2P best normalization thresholding under different channels environment (sighting distance and non line of sight), different integration period, using the meansigma methods of 2P the best normalization thresholding as optimization thresholding X；
4) step 1 is returned) select next signal to noise ratio, and recalculate corresponding to the average combined parameters J2P under this signal to noise ratio, TOA estimation difference and optimization thresholding X, until all signal to noise ratios gone through within the scope of 4-32dB；
5) by step 4) value of 29 groups of average combined parameters J2P, TOA estimation difference and optimization thresholding X that obtains, as the fingerprint database being made up of three parameters；
Step B, fingerprint database is carried out curve fitting, utilize neutral net that above-mentioned fingerprint database is trained, finally set up the corresponding relation F of average combined parameters J2P and optimum normalization thresholding X, namely owing to average combined parameters J2P and SNR is relevant, and optimum normalization thresholding is calculated under certain specific SNR, the corresponding relation of J and optimum normalization thresholding therefore can be set up；
Step C, when signal transmission delay is carried out Practical Calculation, actual average combined parameters J2P and maximum curl/greatest gradient product D is obtained according to the degree of bias S of actual signal gathered, kurtosis K, maximum curl C and greatest gradient MS, D is utilized to compare with the NLOS threshold value set in advance, judge that signal comes from LOS environment or NLOS environment, recognition result is preserved；Utilize corresponding relation F, calculate the normalization thresholding obtained corresponding to this actual average combined parameters J2P, obtain TOA estimated value according to this normalization thresholding: be input to step B by the actual average combined parameters J2P of gained) own trained good neutral net, corresponding normalization thresholding can be obtained according to corresponding relation F, utilizing normalization thresholding to identify the energy block exceeding this thresholding at first, the moment corresponding using the centre position of this energy block is as TOA estimated value.
2. a kind of 60GHz non line of sight identification based on energy measuring and wireless fingerprint localization method as claimed in claim 1, it is characterised in that in step A) formula J=N*norm (KS)-M*norm (D) in, above-mentioned N value is less than 20.
3. a kind of 60GHz non line of sight identification based on energy measuring and wireless fingerprint localization method as claimed in claim 1, it is characterised in that in step A) formula J=N*norm (KS)-M*norm (D) in, N value 12, M value 2.
4. a kind of 60GHz non line of sight identification based on energy measuring and wireless fingerprint localization method as claimed in claim 1, it is characterized in that in above-mentioned steps C, when maximum curl according to gained/greatest gradient product D is to Practical Calculation, terminal transmission 60GHz pulse sequence signal to be positioned carries out the step of NLOS identification, specific as follows:
1) thresholding is set according to formula (1),
In IEEE802.15.3c60GHz channel model, 1000 channel impulse responses of each generation under LOS and NLOS environment respectively, corresponding 60GHz signal all can be obtained for each channel impulse response, then each 60GHz signal is all calculated and obtain corresponding maximum curl C and greatest gradient MS, and then obtain maximum curl/greatest gradient product D=C*MS, namely under LOS and NLOS environment, obtain the sample value of 1000 gradients/standard deviation product D respectively, be denoted as D respectivelyLOSWith DNLOS, for getting rid of the extremum of the minority of the existence impact on determining thresholding in sample value, respectively by DLOSWith DNLOSArrange by ascending order, choose DLOSWith DNLOSThe sample value of front 90% is denoted as respectivelyWithCalculate respectivelyWithMeansigma methods be denoted asWithChooseWithIn less value identify required thresholding α as final NLOSD；
2) step A-2 is utilized) calculated maximum curl/greatest gradient product D and the thresholding α set in advanceDRelatively judge that terminal transmission 60GHz pulse sequence signal to be positioned comes from LOS environment or NLOS environment；NLOS recognizer is expressed as:
When transmitting antenna is omnidirectional antenna (TX=360 °), as maximum curl D≤αDThen think that signal comes from LOS environment, otherwise, as maximum curl D > αDThen think that signal comes from NLOS environment；When transmitting antenna is non-omnidirectional antenna (TX < 360 °), as maximum curl D≤αDThen think that signal comes from NLOS environment, otherwise, as maximum curl D > αDThen think that signal comes from LOS environment.
3) in above-mentioned steps (7), utilize above-mentioned steps 2) LOS/NLOS recognition result, location-server will preferentially utilize the TOA estimated value under LOS environment, and in conjunction with the TOA estimated value under nlos environment, terminal to be positioned is positioned by application TOA TDOA based on the location algorithm of distance, thus obtaining more accurate positioning result.
CN201610065460.1A 2016-01-29 2016-01-29 A kind of identification of 60GHz non line of sight and wireless fingerprint positioning method based on energy measuring Expired - Fee Related CN105738866B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201610065460.1A CN105738866B (en) 2016-01-29 2016-01-29 A kind of identification of 60GHz non line of sight and wireless fingerprint positioning method based on energy measuring

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201610065460.1A CN105738866B (en) 2016-01-29 2016-01-29 A kind of identification of 60GHz non line of sight and wireless fingerprint positioning method based on energy measuring

Publications (2)

Publication Number Publication Date
CN105738866A true CN105738866A (en) 2016-07-06
CN105738866B CN105738866B (en) 2018-03-13

Family

ID=56247096

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201610065460.1A Expired - Fee Related CN105738866B (en) 2016-01-29 2016-01-29 A kind of identification of 60GHz non line of sight and wireless fingerprint positioning method based on energy measuring

Country Status (1)

Country Link
CN (1) CN105738866B (en)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108663717A (en) * 2018-04-08 2018-10-16 青岛科技大学 Unmanned plane signal recognition method and system
CN110178043A (en) * 2016-11-11 2019-08-27 弗劳恩霍夫应用研究促进协会 User equipment positioning in mobile communications network based on delay and path intensity
CN110225449A (en) * 2019-05-22 2019-09-10 东南大学 It is a kind of based on millimeter wave CRAN 3D positioning, test the speed and environment mapping method

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP2542001A1 (en) * 2011-06-30 2013-01-02 Naser El-Sheimy Method and apparatus for determining position of a wireless device
CN104219761A (en) * 2014-10-09 2014-12-17 中国石油大学(华东) Ultra-wideband wireless positioning method based on maximum slope

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP2542001A1 (en) * 2011-06-30 2013-01-02 Naser El-Sheimy Method and apparatus for determining position of a wireless device
CN104219761A (en) * 2014-10-09 2014-12-17 中国石油大学(华东) Ultra-wideband wireless positioning method based on maximum slope

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
R.C.DANIELS和R.W.HEALTH: "60GHz wireless communications: emerging requirements and design recommendations", 《IEEE VEHICULAR STRATEGY MAGAZINE》 *
徐凌伟 等: "60GHz室内信道下一种快速天线选择算法", 《微波学报》 *

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110178043A (en) * 2016-11-11 2019-08-27 弗劳恩霍夫应用研究促进协会 User equipment positioning in mobile communications network based on delay and path intensity
CN110178043B (en) * 2016-11-11 2023-06-20 弗劳恩霍夫应用研究促进协会 User equipment positioning in a mobile communication network based on delay and path strength
CN108663717A (en) * 2018-04-08 2018-10-16 青岛科技大学 Unmanned plane signal recognition method and system
CN108663717B (en) * 2018-04-08 2019-09-10 青岛科技大学 Unmanned plane signal recognition method and system
CN110225449A (en) * 2019-05-22 2019-09-10 东南大学 It is a kind of based on millimeter wave CRAN 3D positioning, test the speed and environment mapping method

Also Published As

Publication number Publication date
CN105738866B (en) 2018-03-13

Similar Documents

Publication Publication Date Title
Choi et al. Deep learning based NLOS identification with commodity WLAN devices
CN105554882B (en) The identification of 60GHz non line of sight and wireless fingerprint positioning method based on energy measuring
CN105611629A (en) 60GHz millimeter wave non-line of sight identification and wireless fingerprint positioning method based on energy detection
CN105929378A (en) Direct tracking method based on external radiation source joint time delay and Doppler frequency
CN109490826B (en) Ranging and position positioning method based on radio wave field intensity RSSI
CN106851821B (en) Indoor three-dimensional positioning method based on wireless communication base station
CN101883426A (en) Ultra-wideband wireless positioning method with high precision
CN106488557A (en) LDCC PDF based on WiFi signal is classified delay time estimation method
CN105510948A (en) Multi-target directly positioning method for complex envelope and carrier phase information of allied radio signals
CN103297924A (en) Novel positioning fingerprint design method and indoor positioning system based on channel estimation
Liu et al. LC-DNN: Local connection based deep neural network for indoor localization with CSI
CN110636436A (en) Three-dimensional UWB indoor positioning method based on improved CHAN algorithm
CN108226912B (en) Sparse network-based non-contact object perception positioning method and system
CN104219761A (en) Ultra-wideband wireless positioning method based on maximum slope
CN105611628A (en) High precision pulse 60GHz wireless fingerprint positioning method based on energy detection
Li et al. An indoor positioning algorithm based on RSSI real-time correction
CN105738866A (en) 60GHz Non-Line-of-Sight identification and wireless fingerprint positioning method based on energy detection
Tan et al. UAV localization with multipath fingerprints and machine learning in urban NLOS scenario
CN114205906A (en) Indoor positioning method based on UWB channel quality assessment
CN105589063B (en) Impulse radio 60GHz distance-finding methods based on the degree of bias
Tarrio et al. An RSS localization method based on parametric channel models
Wei et al. A closed-form location algorithm without auxiliary variables for moving target in noncoherent multiple-input and multiple-output radar system
Xu et al. Variance-based fingerprint distance adjustment algorithm for indoor localization
Bryan et al. Geolocation of multiple noncooperative emitters using received signal strength: Sparsity, resolution, and detectability
CN105759242A (en) High-precision pulse 60GHz wireless fingerprint positioning method based on energy detection

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant
CF01 Termination of patent right due to non-payment of annual fee

Granted publication date: 20180313

Termination date: 20190129

CF01 Termination of patent right due to non-payment of annual fee