CN103744051B - Base station identification approach in a kind of SFN positioning system - Google Patents

Abstract

The present invention proposes the base station identification approach in a kind of SFN positioning system.Identification of base stations problem, from positioning system, is converted into Data classification problem, realizes identification of base stations by minimum distance classifier by described method; Utilization state forecast model estimates the kinematic parameter of transfer table, and then reduces the hsrdware requirements of system; In order to obtain better identification of base stations effect under nlos environment, using Interactive Multiple-Model method to alleviate non line of sight effect, thus obtaining transfer table target information more accurately.Location estimation and identification of base stations two different position system functions are closely linked by the present invention, economical, effectively, be applicable to all digital television standards.

Description

Base station identification approach in a kind of SFN positioning system

Technical field

The invention belongs to communication field of locating technology, the base station identification approach particularly in a kind of SFN positioning system.

Background technology

DTTB (DTTB) starts to widely use in August, 2007 in China as a kind of brand-new information transmission technology.User, except can receiving effective video sound signal, can also utilize the carrier wave of ground digital television signal and digital code stream to measure the spatial parameters such as the distance from launching tower to receiving end, realize wireless location.Due to the huge advantage of digital television signal on location, increasing people starts to pay close attention to this location new technology.

Digital tv ground broadcasting have employed SFN technology for saving frequency spectrum, is namely in the transmitting station under the synchronous regime of different location, launches same signal at one time with same frequency.This technology can save valuable frequency resource greatly, improves the availability of frequency spectrum; But for positioning function exploitation, mean the systematic difficult problem being difficult to avoid, namely transfer table itself cannot identify the signal from different base station as global position system GPS receiver and network based positioning system.Therefore, the identification of base stations problem of SFN positioning system will be a challenging problem.

Current single frequency network base stations knows method for distinguishing to be had:

(1) based on the base station identification approach that pseudo-random sequence is injected

This method solve the identification of base stations problem of the digital television ground broadcast signal of United States advanced television systems committee ATSC standard.At transmitting terminal, the Kasami sequence of 16, is also called radio frequency watermark (RFWM) signal injection in ATSC standard signal, will ensure that the signal injected does not produce any impact to Digital TV Receiving simultaneously.At receiving end, the sequence that Received signal strength and local Kasami sequencer produce is carried out related operation.If local sequence is exactly the RFWM signal being injected into i-th transmitter, so auto-correlation computation will there will be correlation peak, and different transmitting base stations just can be distinguished.

(2) based on the time-multiplexed base station identification approach of transmission parameter signaling TPS

The cardinal principle of the method adopts time division multiplexing tdm mode at transmitting terminal to TPS.For a certain signal frame, regulation only has a transmitting base station to send TPS signal, and other transmitting base station all sends 0 at the TPS correspondence position of this signal frame; In next signal frame, another transmitting base station sends TPS, and the correspondence position of all the other base stations all sends 0.By that analogy.Therefore, at transmitting terminal, multiple transmitting base station sends homotactic signal frame simultaneously, only has a transmitting base station to send TPS by agreement; At receiving end, in the signal frame to the same sequence number be superimposed from multiple transmitting base station, receiver only detects the TPS signal from a certain particular transmission base station.When transmitting base station known by receiver and send the reference table of TPS signal frame structure sequence number, just transmitting base station can be identified.

(3) based on the identification of base stations algorithm of speed of mobile terminal

The velocity information that the method provides from mobile terminal, changes into data interconnection problem by identification of base stations problem, utilizes velocity information to the restriction relation of base station time delay, thus obtains the pair relationhip between base station and time delay, complete identification of base stations.

(4) based on the base station identification approach of fuzzy logic

The method is based upon on the basis of Kalman (Kalman) localization method equally.The basic ideas of its algorithm are divided into 4 parts: base station-time delay pairing arranges; Degree of membership calculates and compares; Location estimation; Location verification and feedback.

(5) based on the base station identification approach under sighting distance LOS/ non line of sight NLOS

The method utilizes non line of sight statistical property Gaussian distributed, and under follow-up location-estimation algorithm can provide the condition of environmental information, utilizes specific location-estimation algorithm to provide channel environmental information, builds identification of base stations algorithm.

Summary of the invention

The object of the invention is to overcome the deficiencies in the prior art, propose the base station identification approach in a kind of SFN positioning system.

In order to solve the problems of the technologies described above, the technical solution used in the present invention is: the base station identification approach in a kind of SFN positioning system, comprises step as follows:

Steps A, the measurements of arrival time value that receiver estimates i-th base station signal from Received signal strength is defined as x _i; By the set of the measurements of arrival time value of N number of base station signal composition X, X={x _i, i=1,2...N}; Base station measurements of arrival time value subset is built according to set X;

Step B, first the population of measured values in set X and home base stations set of measurements number are compared, home base stations set of measurements is stored in receiver; Then from home base stations set of measurements, choose the candidate base station subset participating in location;

Step C, distinguishes the distance of the base station in calculated candidate subset of base stations to moving target, using the reference value of this distance as base station;

Step D, candidate base station sub-set size and measured value sub-set size are consistent, to each candidate base station subset, the Euclidean distance sum of the base station reference value of correspondence position in each measured value and candidate base station subset in computation and measurement value subset; Search Euclidean distance and minimum subset of base stations, as identification of base stations result;

Step e, carries out measured value to above-mentioned identification of base stations result level and smooth;

Step F, is used for location estimation and the status predication of moving target by level and smooth measured value.

In step C, described base station is Euclidean distance to the distance of moving target.

In step D, described Euclidean distance summation process is as follows:

Step D-1, chooses subset successively from described measured value subset set and candidate base station subset set;

Step D-2, according to the corresponding order of element in subset, calculates the Euclidean distance of element on respective correspondence position successively;

Step D-3, carries out read group total by the Euclidean distance of above-mentioned calculating, summation gained result be exactly this measured value subset sums candidate base station subset Euclidean distance and.

Beneficial effect of the present invention: the present invention proposes the base station identification approach in a kind of SFN positioning system.Identification of base stations problem, from positioning system, is converted into Data classification problem, realizes identification of base stations by minimum distance classifier by described method; Utilization state forecast model estimates the kinematic parameter of transfer table, and then reduces the hsrdware requirements of system; In order to obtain better identification of base stations effect under nlos environment, using Interactive Multiple-Model method to alleviate non line of sight effect, thus obtaining transfer table target information more accurately.Location estimation and identification of base stations two different position system functions are closely linked by the present invention, economical, effectively, be applicable to all digital television standards.

Accompanying drawing explanation

Fig. 1 is the general frame of base station identification approach of the present invention.

Fig. 2 be in the present invention minimum distance classifier for the process flow diagram of identification of base stations.

Fig. 3 is the identification of base stations analysis of the present invention under unfixed collection of base stations.

Fig. 4 is the present invention cumulative distribution function CDF at different conditions.

Fig. 5 is the cumulative distribution function of the present invention under difference drives noise.

Fig. 6 is the cumulative distribution function of the present invention under different mobile stations speed.

Fig. 7 is the computation process of the Euclidean distance sum of p measured value subset sums q candidate base station subset.

Embodiment

Below in conjunction with accompanying drawing, describe the base station identification approach in a kind of SFN positioning system of the present invention in detail.

As shown in Figure 1, the base station identification approach in a kind of SFN positioning system of the present invention, be common algorithm because non line of sight NLOS alleviates algorithm location-estimation algorithm, State Forecasting Model is also the routine techniques in Radar Signal Processing.Specific descriptions minimum distance classifier is used for identification of base stations below.

Minimum distance classifier is used for the process flow diagram of identification of base stations as shown in Figure 2.Suppose that the DTV number of base stations of the priori stored in the TOA measured value that obtains and receiver is respectively N and M, concrete implementation step is as follows: (1), according to the TOA number of measurements obtained, obtains all possible measured value subset.N number of TOA measured value, can obtain N! Individual measured value subset.

(2) if N=M, stop collection of base stations and select, turn (3).Select for collection of base stations, we need to select N number of base station from M base station., lined up the vector with measured value subset same dimension, in vector, the sort method difference of each element can not affect recognition performance.According to permutation and combination principle, obtain the subset of base stations of individual candidate.This step is for solving because localizing objects moves the base station uncertainty caused.

(3) for the subset of base stations of each candidate, the prediction measured value of each base station in subset of computations.For the i-th base station in the set of jth candidate base station, prediction measured value is

$D_{i,j}=\sqrt{{(x_i-\hat{x})}^2+{(y_i-\hat{y})}^2},i=1...N;j=1...C_M^N$

Wherein (x _i, y _i) be i-th base station coordinates, it is the position prediction value of transfer table.

(4) to each TOA measured value subset, calculate the Euclidean distance that measured value is predicted in its base station corresponding with each candidate base station subset, and distance is sued for peace.The process of the distance summation of p measured value subset sums q candidate base station subset is as follows: first A. chooses down the subset being designated as p and q successively from measured value subset set and candidate base station subset set.B. according to the corresponding order of element in subset, the Euclidean distance of element on respective correspondence position is calculated successively.C. the Euclidean distance of above-mentioned calculating is carried out read group total.D. the result of gained of suing for peace be exactly this measured value subset sums candidate base station subset Euclidean distance and.Note: detailed process as shown in Figure 7.

When obtain Euclidean distance and after, then use minimum distance criterion, extract there is the TOA measured value subset sums candidate base station subset having minimum Distance geometry.In measured value subset, each measured value just belongs to the base station with subset of base stations correspondence position, thus completes identification of base stations.

Experimental result

Respectively checking is compared in varied situations to identification of base stations algorithm of the present invention below.Experiment parameter is chosen as follows:

(1) suppose to have 4 DTV base stations in emulation experiment.Base station 1, base station 2, the coordinate of base station 3 and base station 4 is respectively (0m, 0m), (8000m, 0m), (15000m, 30000m) and (45000m, 30000m).

(2) under fixing collection of base stations condition, base station 1, base station 2 and base station 3 participate in the whole position fixing process of transfer table.

(3) under unfixed collection of base stations condition, only the measured value of base station 1 is subject to NLOS impact.From LOS to NLOS, condition changes once every 200 time samples this measured value.Other measured value is all LOS.

(4) track definition of transfer table is by right value, x (t)=9.7t, y (t)=16.8t. measured value adds that measurement noises and NLOS mistake produce respectively.Measurement noises is assumed to Gaussian noise, and (average is 0 standard deviation is σ _m=2m).NLOS mistake is assumed to Gaussian distribution equally, and (average is m _nLOS=45m standard deviation is σ _nLOS=5m) under fixing collection of base stations and unfixed collection of base stations condition, it is 0.05s that the time sample number of emulation was respectively for 1000 and 3000. sampling intervals.Carry out 50 MonteCarlo experiments altogether.

(5) NLOS alleviates algorithm employing Interactive Multiple-Model (IMM) algorithm; Optimum configurations is as follows: the variance of driving noise is the probability of each model is u _{k, 1}(0)=u _{k, 2}(0) the initial covariance matrix of=0.5fork=1,2,3.LOS and NLOS wave filter is 100 × σ _m× Iand100 × σ _nlosthe Markov transition probability matrix of × I.LOS/NLOS pattern is

$P=\left[\begin{array}{cc}{p}_{11}& p_{12}\\ p_{21}& p_{22}\end{array}\right]=\left[\begin{array}{cc}0.995& 0.005\\ 0.005& 0.995\end{array}\right]$

Storage can be there is wherein in simulated conditions supposition transfer table the position estimated.State Forecasting Model parameter is as follows: the exponent number of state model is 1, and the location estimation feedback number of use is 2, and the time of prediction is the next time interval.Location-estimation algorithm adopts algorithm. and the initial value of location estimation is the mean value of three base station coordinates.

Experiment 1: checking identification of base stations performance.

Fig. 3 is under unfixed collection of base stations, the base station of real participation and identification of base stations result (1 represents that base station participates in location, and 0 represents that base station does not participate in locating).Although the measured value of base station 1 contains NLOS mistake, under the inventive method, the identification that base station 1 still can be correct.Pass through the recognition result of base station 3 and base station 4, we can see, the present invention has the ability when base station kind changes, work that still can be correct simultaneously.

Experiment 2: location estimation performance evaluation.

In order to describe the relation closely of the present invention and location estimation step, Fig. 4 describes CDF at different conditions.We see, in both cases, positioning performance is similar and can meet practical application.Therefore the present invention can provide correct identification of base stations and location estimation result simultaneously.

Experiment 3: stability analysis.

We, for fixing collection of base stations analysis, verify under different driving noise variances and mobile station speed, stability of the present invention.Experiment parameter is as follows: drive noise variance to get 1m/s ²and 5m/s ²emulate different transfer table acceleration environments.The speed parameter of transfer table high speed and low speed is v _x=9.7m/s, v _y=16.8m/s and v _x=2m/s, v _y=3m/s. Fig. 5 and Fig. 6 respectively describes the location estimation performance of the present invention under different parameters.

For those skilled in the art, according to above implementation type can be easy to association other advantage and distortion.Therefore, the present invention is not limited to above-mentioned instantiation, and it carries out detailed, exemplary explanation as just example to a kind of form of the present invention.Not deviating from the scope of present inventive concept, the technical scheme that those of ordinary skill in the art are obtained by various equivalent replacement according to above-mentioned instantiation, all should be included within right of the present invention and equivalency range thereof.

Claims (3)

1. the base station identification approach in SFN positioning system, is characterized in that, comprise step as follows:

Steps A, the measurements of arrival time value that receiver estimates i-th base station signal from Received signal strength is defined as x _i; By the set of the measurements of arrival time value of N number of base station signal composition X, X={x _i, i=1,2...N}; Base station measurements of arrival time value subset is built according to set X;

Step B, first the population of measured values in set X and home base stations set of measurements number are compared, home base stations set of measurements is stored in receiver; Then from home base stations set of measurements, choose the candidate base station subset participating in location;

Step C, distinguishes the distance of the base station in calculated candidate subset of base stations to moving target, using the reference value of this distance as base station;

Step D, candidate base station sub-set size and measured value sub-set size are consistent, to each candidate base station subset, the Euclidean distance sum of the base station reference value of correspondence position in each measured value and candidate base station subset in computation and measurement value subset; Search Euclidean distance and minimum subset of base stations, as identification of base stations result;

Step e, carries out measured value to above-mentioned identification of base stations result level and smooth;

Step F, is used for location estimation and the status predication of moving target by level and smooth measured value.

2. the base station identification approach in a kind of SFN positioning system according to claim 1, is characterized in that, in step C, described base station is Euclidean distance to the distance of moving target.

3. the base station identification approach in a kind of SFN positioning system according to claim 1, is characterized in that, in step D, described Euclidean distance summation process is as follows:

Step D-1, chooses subset successively from described measured value subset set and candidate base station subset set;

Step D-2, according to the corresponding order of element in subset, calculates the Euclidean distance of element on respective correspondence position successively;

Step D-3, carries out read group total by the Euclidean distance of above-mentioned calculating, summation gained result be exactly this measured value subset sums candidate base station subset Euclidean distance and.