CN104935388B - Background interference noise level prediction method and system - Google Patents
Background interference noise level prediction method and system Download PDFInfo
- Publication number
- CN104935388B CN104935388B CN201510221994.4A CN201510221994A CN104935388B CN 104935388 B CN104935388 B CN 104935388B CN 201510221994 A CN201510221994 A CN 201510221994A CN 104935388 B CN104935388 B CN 104935388B
- Authority
- CN
- China
- Prior art keywords
- noise level
- average
- interference noise
- ambient interferences
- sequence
- 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.)
- Active
Links
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to a background interference noise level prediction method and system. The method comprises: setting relevant background interference noise level mean sequences of a prediction moment and a known time quantum in dependence on the historic background interference noise level mean value data set of a predicted frequency point and a preset perception data effective scope; obtaining a Euclidean distance between the relevant background interference noise level mean sequence of the known time quantum and the relevant background interference noise level mean sequence of the prediction moment; screening a plurality of background interference noise level mean values of the known time quantum corresponding with the Euclidean distance and less than a preset threshold to generate a background interference noise level prediction value set; and performing weight calculation in dependence on the background interference noise level prediction value set to obtain the background interference noise level prediction value of the predicted frequency point at the prediction moment. The background interference noise level prediction method and system analyze historic data through a similarity coupling method, predict a future background interference noise level based on historical data change rules, and possess high prediction precision.
Description
Technical field
The present invention relates to frequency management technology, more particularly to a kind of ambient interferences noise level Forecasting Methodology and system.
Background technology
Ambient interferences noise is one of determinant of influence receiver signal interference-to-noise ratio, the direct shadow of its precision of prediction
Ring the accuracy and validity to dynamic frequency management.Ambient interferences noise influences or destroys wireless device normal work
The summation of electromagnetic radiation, is made up of interference signal and ambient noise.At present around the main skill of ambient interferences noise level prediction
Art has following several：
(1) P.372 recommendation, should there is provided the reference value of 0.1Hz~100GHz wide background noise levels for ITU-R
Ambient noise is defined as recommendation the summation of the unwanted radiation in various sources, and is specially eliminated from single recognizable transmitting
Signal, this mode is to the actual environment particularly noise emission equipment City Regions such as city and residential block and does not apply to, estimation
Noise floor value can be universal relatively low.
(2) ITU-R SM.1753 recommendations give measurement and assessment ambient noise in a kind of actual radio application
Method.Ambient noise is divided into white Gaussian noise, three parts of impulsive noise and single carrier noise by the recommendation.To Gauss
The measurement of white noise uses " 20% method ", i.e., the frequency range studied is scanned first, chooses one or more and does not exist
The frequency of transmitting, records the level value of certain period of time, and using wherein minimum 20% level value intermediate value as noise
Level.The method has given up 80% sample, and precision is not high, and needs the artificial idle frequency of participation selection, implements complex.
(3) at present also have a kind of electromagnetic background noise extracting method based on layering-clustering algorithm, the method by collection
Spectrum monitoring data carry out data subset and divide and field intensity layering, the center field intensity value of each clustering cluster are obtained, by central field
Intensity values sequence obtains ambient noise sample value, forms ambient noise sample sequence, obtains the background-noise level in the unit interval.
The method compares above two method, and precision is significantly improved, and need not manually be participated in.But the method and above two method
The prediction to ambient noise is only provided, and in actual dynamic frequency management application, in addition it is also necessary to consider interference signal
Influence.
In spectrum monitoring, the electromagnetic energy of each frequency collection is the interference noise level summation of the frequency.Due to dry
The randomness disturbed, particularly shortwave frequency range, interference noise level often in interior fluctuation in a big way, using measured data or one section
Used as interference noise level predicted value in a short time, precision is also difficult to meet and requires average in time.
The content of the invention
Based on this, to solve problem present in existing ambient interferences noise level Predicting Technique, the present invention provides a kind of
Ambient interferences noise level Forecasting Methodology and system, analyse the historical spectrum Monitoring Data of accumulation scientifically, draw degree of precision
Ambient interferences noise level predicted value.
To achieve the above object, adopted the following technical scheme that in the embodiment of the present invention：
A kind of ambient interferences noise level Forecasting Methodology, comprises the following steps：
According to the historical background interference noise level mean data collection and default perception data effective range of prediction frequency,
Build the background context interference noise electricity sequence of average of prediction time and the background context interference noise electricity of known time section
Sequence of average；The historical background interference noise level mean data collection includes the prediction frequency in several different times
The ambient interferences noise level average of section；The known time section is concentrated for the historical background interference noise level mean data
Time period；
Obtain the background context interference noise electricity sequence of average of the known time section and the background context of prediction time
Euclidean distance between interference noise electricity sequence of average；
Correspondence Euclidean distance is filtered out to be made an uproar less than the background context interference of several known time sections of pre-determined threshold
Acoustic-electric sequence of average, extracts the ambient interferences noise level average of several known time sections, generates the prediction
Ambient interferences noise level predicted value collection of the frequency in prediction time；
It is weighted according to the ambient interferences noise level predicted value collection, the acquisition default frequency is in prediction
The ambient interferences noise level predicted value at quarter.
The present invention also provides a kind of ambient interferences noise level forecasting system, including：
Data sequence build module, for according to prediction frequency historical background interference noise level mean data collection and
Default perception data effective range, builds the background context interference noise electricity sequence of average and known time section of prediction time
Background context interference noise electricity sequence of average；The historical background interference noise level mean data collection includes the prediction
Ambient interferences noise level average of the frequency in several different time sections；The known time section is disturbed for the historical background
The time period that noise level mean data is concentrated；
Oldham distance calculating module, the background context interference noise electricity sequence of average for obtaining the known time section
And the Euclidean distance between the background context interference noise electricity sequence of average of prediction time；
Predicted value collection generation module, for filtering out several known times section of the correspondence Euclidean distance less than pre-determined threshold
Background context interference noise electricity sequence of average, extract several known times section ambient interferences noise level average,
Ambient interferences noise level predicted value collection of the generation prediction frequency in prediction time；
Weighted calculation module, for being weighted according to the ambient interferences noise level predicted value collection, obtains institute
State ambient interferences noise level predicted value of the default frequency in prediction time.
The present invention carries out similitude the matching analysis according to historical background interference noise level mean data collection, when analyzing one section
The ambient interferences noise level mean data of interior similar time point, finds the Changing Pattern between data, and based on Changing Pattern
Following ambient interferences noise level is predicted, thus the present invention can obtain the precision of prediction higher compared to existing other technologies.
Brief description of the drawings
Fig. 1 is the schematic flow sheet of ambient interferences noise level Forecasting Methodology in the embodiment of the present invention；
Fig. 2 is the principle schematic of similitude matching method in the embodiment of the present invention；
Fig. 3 illustrates for the ambient interferences noise level of 6MHz in the embodiment of the present invention predicts the outcome with the contrast of actual result
Figure；
Fig. 4 is the ambient interferences noise level predicated error distribution map of 6MHz in the embodiment of the present invention；
Fig. 5 shows for the ambient interferences noise level of 11MHz in the embodiment of the present invention predicts the outcome with the contrast of actual result
It is intended to；
Fig. 6 is the ambient interferences noise level predicated error distribution map of 11MHz in the embodiment of the present invention；
Fig. 7 shows for the ambient interferences noise level of 15MHz in the embodiment of the present invention predicts the outcome with the contrast of actual result
It is intended to；
Fig. 8 is the ambient interferences noise level predicated error distribution map of 15MHz in the embodiment of the present invention；
Fig. 9 is the structural representation of ambient interferences noise level forecasting system in the embodiment of the present invention；
Figure 10 is the structural representation of another ambient interferences noise level forecasting system in the embodiment of the present invention.
Specific embodiment
Present disclosure is further described below in conjunction with the accompanying drawings.
As shown in figure 1, providing a kind of ambient interferences noise level Forecasting Methodology in the present embodiment, comprise the following steps：
S100 is effective according to the historical background interference noise level mean data collection and default perception data of prediction frequency
Scope, the background context interference for building the background context interference noise electricity sequence of average and known time section of prediction time is made an uproar
Acoustic-electric sequence of average；
The background context interference noise electricity sequence of average that S200 obtains the known time section is related to prediction time
Euclidean distance between the equal value sequence of ambient interferences noise level；
S300 filters out correspondence Euclidean distance and is done less than the background context of several known time sections of pre-determined threshold
The equal value sequence of noise level is disturbed, the ambient interferences noise level average of several known time sections is extracted, generation is described
Ambient interferences noise level predicted value collection of the prediction frequency in prediction time；
S400 is weighted according to the ambient interferences noise level predicted value collection, obtains the default frequency pre-
Survey the ambient interferences noise level predicted value at moment.
Specifically, the general principle reference picture 2 of the present embodiment, if the background of preceding several time periods of certain time period is done
Disturb noise level average and the essentially identical (ash of the ambient interferences noise level average of preceding several time periods of another time period
Color dash area is identical), then it is considered that the ambient interferences noise level average of the two time periods also will be essentially identical.Cause
This, first according to the historical background interference noise level mean data collection and default perception data of prediction frequency in the present embodiment
Effective range, the background context of the background context interference noise electricity sequence of average and known time section that build prediction time is done
Disturb the equal value sequence of noise level.Wherein, historical background interference noise level mean data concentrates the history for covering prediction frequency
Data, that is, predict ambient interferences noise level average of the frequency in multiple different time sections, it is known that the time period refers to historical background
The time period that interference noise level mean data is concentrated.Prediction time and the background context interference noise level of known time section
Equal value sequence is made up of the ambient interferences noise level average of different time sections, according to the historical background interference noise of prediction frequency
Level mean data collection and default perception data effective range determine.Then the background context interference for calculating known time section is made an uproar
Euclidean distance between the background context interference noise electricity sequence of average of acoustic-electric sequence of average and prediction time；Using default
Thresholding (generally takes 3~5dB), filters out background context of the correspondence Euclidean distance less than several known times section of pre-determined threshold
Interference noise electricity sequence of average, extracts the ambient interferences noise level average of several time periods, arranges in certain sequence,
Ambient interferences noise level predicted value collection of the generation prediction frequency in prediction time.Then it is pre- further according to ambient interferences noise level
The element that measured value is concentrated is weighted, and obtains ambient interferences noise level predicted value of the default frequency in prediction time.
In a kind of specific embodiment, the historical background interference noise level of prediction frequency can be by the following method obtained
Mean data collection：
The ambient interferences noise level instantaneous value in collection several observation cycles of prediction frequency；
One observation cycle is divided into several time periods according to presetting granularity, is done according to the background in each time period
Noise level instantaneous value is disturbed, the ambient interferences noise level average of prediction each time period of frequency is counted；
Predict that the historical background interference of frequency is made an uproar described in ambient interferences noise level average generation according to each time period
Acoustic-electric average data collection.
For example, it is assumed that be one with one day and observe the cycle, if it is P that the Monitoring Data of prediction frequency integrates0, P0Expression
Formula is：
P0=[p0(d, n)] d=1,2 ..., D；N=1,2 ..., N
Wherein：
p0It is the element of Monitoring Data concentration, i.e. the ambient interferences noise of the prediction frequency that each observation cycle monitoring is arrived is electric
Flat instantaneous value (unit is dB)；
D is the number of days of collection；
N is the sample number of the prediction frequency ambient interferences noise level instantaneous value in a day, N=24 × 60/t0, wherein, t0
It is the scan period of spectrum monitoring equipment.
Then one day (24 hours) are divided into several time periods, each time period of statistical forecast frequency by presetting granularity
Ambient interferences noise level average.
If it is P to predict that the historical background interference noise level mean data of frequency integrates1, P1Expression formula be：
P1=[p1(d, m)] d=1,2 ..., D；M=1,2 ..., M
Wherein：
p1For the element that historical background interference noise level mean data is concentrated (unit is dB)；
M is the time hop count being divided into a day, and M=24 × 60/T, wherein T are presetting granularity (unit is minute).
Thus p can be derived1Expression formula be：
Wherein：
According to historical background interference noise level mean data collection P1, ambient interferences noise is carried out using similitude matching method
Level is predicted.The advantage of similitude matching method is computational accuracy high, and tell about carries out background in detail below according to similitude matching method
The process of interference noise level prediction.
First, according to the historical background interference noise level mean data collection P of prediction frequency1, prediction time is built respectively
Background context interference noise electricity sequence of average and known time section background context interference noise electricity sequence of average.Hold
Connect above example, be an observation cycle still with one day, for prediction time, w time period before taking the same day, a few days ago
The ambient interferences noise level average of same time period and adjacent w time period is used as background context interference noise level average sequence
Row, wherein putting in order for time period does not influence result, but the arrangement of all background context interference noises electricity sequence of average is suitable
Sequence must be represented by equation below according to same rule, the background context interference noise electricity sequence of average of prediction time：
P1.d,m={ p1.p,q∈P1| 0 ＜ d-p≤w, | m-q |≤w } ∪ { p1.p,q∈P1| d=p, 0 ＜ m-q≤w }
Wherein, P1.d,mRepresent background context interference noise electricity sequence of average of the prediction frequency in prediction time：Predict
The background context interference noise level that frequency observes m-th time period in cycle (being the d days in present embodiment) at d-th is equal
Value sequence, P1.p,qRepresent p-th ambient interferences noise of q-th time period of observation cycle (being pth day in present embodiment)
Level average, w is default perception data effective range, P1To predict the historical background interference noise level mean data of frequency
Collection.
Equally, for known time section, it is also possible to obtain background context interference noise electricity sequence of average by above-mentioned formula.
In a kind of specific embodiment, when build prediction time background context interference noise electricity sequence of average and
During the background context interference noise electricity sequence of average of known time section, concentrated in historical background interference noise level mean data
P1If have being lacked for building the ambient interferences noise level average of the time period of background context interference noise electricity sequence of average,
The back of the body of missing is then substituted using the ambient interferences noise level mean of mean of several corresponding time periods in observation cycle
Scape interference noise level average.If for example, d-th ambient interferences noise level average of the 3rd time period in observation cycle lacks
Lose, then can use the 1st observation the cycle the 3rd time period, the 2nd observation the cycle the 3rd time period ..., the d-1 observe
The ambient interferences noise level mean of mean of the 3rd time period in cycle substitutes d-th the 3rd time period in observation cycle
Ambient interferences noise level average calculated.
In a kind of specific embodiment, when the background context interference noise electricity sequence of average for building known time section
When, in historical background interference noise level mean data collection P1If in have for build the background context of currently known time period do
Disturb the ambient interferences noise level average missing of several time periods of the equal value sequence of noise level, and deleted background interference noise
The quantity of the time period of level average and the background context interference noise electricity sequence of average for building the currently known time period
The ratio of total amount of time period be more than threshold value, then cancel that to build the background context interference noise level of currently known time period equal
Value sequence.For example, the background context interference noise electricity sequence of average for building the currently known time period needs the n back of the body of time period
Scape interference noise level average, wherein there is the b ambient interferences noise level average missing (b of time period<N), when b/n is more than threshold
During value, then cancel the equal value sequence of ambient interferences noise level for building the currently known time period, can again select known time section
Include calculating.General, the threshold value takes 30%.
Secondly, according to the expression formula of above-mentioned background context interference noise electricity sequence of average：
P1.d,m={ p1.p,q∈P1| 0 ＜ d-p≤w, | m-q |≤w } ∪ { p1.p,q∈P1| d=p, 0 ＜ m-q≤w }
As can be seen that each background context interference noise electricity sequence of average can be as a multidimensional variable, dimension：2
× w × (w+1), the background context interference noise electricity sequence of average that each known time section is calculated respectively is related to prediction time
Euclidean distance between the equal value sequence of ambient interferences noise level, calculating formula is：
distdm,ck=| | p1.d,m-p1.c,k||
Wherein, p1.d,mRepresent the background context interference noise electricity sequence of average of prediction time, p1.c,kRepresent known time
The background context interference noise electricity sequence of average of section, distdm,ckRepresent the background context interference noise level of known time section
Euclidean distance between equal value sequence and the background context interference noise of prediction time electricity sequence of average.
Based on pre-determined threshold (generally taking 3~5dB), known time section of the correspondence Euclidean distance less than pre-determined threshold is filtered out
Background context interference noise electricity sequence of average, extract such known time section ambient interferences noise level average, by one
Fixed order arrangement (such as ascending order), ambient interferences noise level predicted value of the generation prediction frequency in prediction time
Collection.Then the element that ambient interferences noise level predicted value is concentrated is weighted, draws prediction frequency in prediction time
Ambient interferences noise level predicted value.Ambient interferences noise level predicted value concentrates the putting in order to predicting the outcome of element
Influence is only dependent upon weight coefficient, and weight coefficient can be adjusted according to actual conditions, general to set and known time section
Background context interference noise electricity sequence of average to prediction time background context interference noise electricity sequence of average Euclidean away from
From correlation, apart from smaller, weight coefficient is bigger.Specific formula for calculation is as follows：
Wherein, I is the element number that ambient interferences noise level predicted value is concentrated, biIt is weight coefficient, p1.iRepresent background
The element that interference noise level predicted value is concentrated, p1It is the ambient interferences level predicted value for predicting frequency in prediction time.
The beneficial effect of ambient interferences noise level Forecasting Methodology of the invention is illustrated below in conjunction with a specific case.
In present case, using 1 day as an observation cycle, Yanqing County of Beijing area May 13 to May 20 in 2014 is gathered
The spectrum monitoring data of continuous 8 days of day.Spectrum monitoring equipment working frequency range is 2~30MHz, a width of 3kHz of analytic band, scanning speed
Rate is that 500 channels are per second.With 5 minutes for a time period, 288 time periods were divided into by 24 hours one day, calculated respectively
Tri- ambient interferences noise level averages of prediction each time period of frequency of 6MHz, 11MHz, 15MHz, obtain each prediction frequency
Historical background interference noise level mean data collection.Then, phase is used based on historical background interference noise level mean data collection
Ambient interferences noise level prediction is carried out like property matching method：
(1) preset perception data effective range and take 2, i.e. w=2, build 6MHz all in 13 to 19 May in 2014
Know the background context interference noise level average sequence of time period and first time period (as prediction time) of May 20 day in 2014
Row.
(2) for the prediction time (first time period of i.e. 2014 May 20 day) of current selected, its background context is calculated
Euclidean distance between the background context interference noise electricity sequence of average of interference noise electricity sequence of average and each known time section.
(3) pre-determined threshold is taken for 3dB, and the background context interference for filtering out known time section of the Euclidean distance less than 3dB is made an uproar
Acoustic-electric sequence of average, extracts the ambient interferences noise of these background context interference noise level average sequences corresponding time period
Level, and ascending sequence, form ambient interferences noise level predicted value collection.
(4) weight coefficient is taken for { 1,0,0 ... } (takes the ambient interferences of the minimum known time section of correspondence Euclidean distance
Noise level average is predicted as prediction frequency 6MHz in the first ambient interferences noise level of time period of May 20 day in 2014
Value, the weighting scheme accuracy is higher in the case of valid data abundance, but can than in the case of more serious in shortage of data
By the way of average weighted), the element that ambient interferences noise level predicted value is concentrated is weighted, draw 2014 years
First ambient interferences noise level predicted value of time period of May 20 day.
(5) using first ambient interferences noise level predicted value of time period of May 20 day in 2014 as known background
Interference noise level average includes calculating, builds the background context interference noise level of on May 20th, 2014 next time period
Equal value sequence.
Repeat (2) to (5), you can predict 288 ambient interferences noises of time period of the 6MHz whole days of on May 20th, 2014
Level average.Predict the outcome with the contrast of measured result as shown in figure 3, error distribution is as shown in Figure 4.
11MHz and 15MHz can respectively be tried to achieve in 288 time periods of the whole day of on May 20th, 2014 by same method
Ambient interferences noise level predicted value.11MHz's predicts the outcome with measured result contrast as shown in figure 5, error is distributed such as Fig. 6 institutes
Show.15MHz's predicts the outcome with measured result contrast as shown in fig. 7, error distribution is as shown in Figure 8.
In sum, ambient interferences noise level of the present invention based on historical data analysis interior similar time point for a period of time
Mean data, finds the Changing Pattern between data, and predicts the ambient interferences noise level of future time instance based on Changing Pattern,
With high precision of prediction.Thinking of the present invention is novel, takes into full account the influence of interference signal and ambient noise, based on the history back of the body
Scape interference noise level mean data collection carries out similitude the matching analysis, with good precision of prediction, and suitable for full frequency band.
Accordingly, the present invention also provides a kind of ambient interferences noise level forecasting system, as shown in figure 9, including：
Data sequence builds module 100, for the historical background interference noise level mean data collection according to prediction frequency
And default perception data effective range, build prediction time background context interference noise electricity sequence of average and it is known when
Between section background context interference noise electricity sequence of average；The historical background interference noise level mean data collection includes described
Ambient interferences noise level average of the prediction frequency in several different time sections；The known time section is the historical background
The time period that interference noise level mean data is concentrated；
Oldham distance calculating module 200, the background context interference noise level average for obtaining the known time section
Euclidean distance between the background context interference noise electricity sequence of average of sequence and prediction time；
Predicted value collection generation module 300, for filter out correspondence Euclidean distance less than pre-determined threshold several it is known when
Between section background context interference noise electricity sequence of average, extract several known times section ambient interferences noise level it is equal
Value, generates ambient interferences noise level predicted value collection of the prediction frequency in prediction time；
Weighted calculation module 400, for being weighted according to the ambient interferences noise level predicted value collection, obtains
Ambient interferences noise level predicted value of the default frequency in prediction time.
In the present embodiment, data sequence structure module 100 is equal according to the historical background interference noise level of prediction frequency
Value Data collection and default perception data effective range, build prediction time background context interference noise electricity sequence of average with
And the background context interference noise electricity sequence of average of known time section.Wherein, historical background interference noise level mean data
Concentration covers the historical data of prediction frequency, that is, predict that frequency is equal in the ambient interferences noise level of multiple different time sections
Value, it is known that the time period refers to the time period that historical background interference noise level mean data is concentrated.Prediction time and it is known when
Between section background context interference noise electricity sequence of average be made up of the ambient interferences noise level average of different time sections, according to
Predict that the historical background interference noise level mean data collection and default perception data effective range of frequency determine.Then Euclidean
The background context interference noise electricity sequence of average that distance calculation module 200 calculates each known time section is related to prediction time
Euclidean distance between the equal value sequence of ambient interferences noise level；Predicted value collection generation module 300 (is generally taken using pre-determined threshold
3~5dB), filter out background context interference noise electricity of the correspondence Euclidean distance less than several known times section of pre-determined threshold
Sequence of average, extracts the ambient interferences noise level average of several time periods, in certain sequence arrangement (for example from it is small to
Big order), ambient interferences noise level predicted value collection of the generation prediction frequency in prediction time.Weighted calculation module 400
The element concentrated according to ambient interferences noise level predicted value is weighted, so as to obtain the back of the body of the default frequency in prediction time
Scape interference noise level predicted value.
In a kind of specific embodiment, as shown in Figure 10, ambient interferences noise level forecasting system of the invention is also wrapped
History data set module 500 is included, history data set module 500 includes：
Acquisition module 501, the ambient interferences noise level for gathering several observation cycles of prediction frequency is instantaneous
Value；
Average statistical module 502, for according to presetting granularity to be divided into several time periods one observation cycle, according to
Ambient interferences noise level instantaneous value in each time period, counts the ambient interferences noise of prediction each time period of frequency
Level average；
Dataset generation module 503, for pre- described in the ambient interferences noise level average generation according to each time period
The historical background interference noise level mean data collection of frequency measurement point.
For example, it is assumed that be one with one day and observe the cycle, if it is P that the Monitoring Data of prediction frequency integrates0, P0Expression
Formula is：
P0=[p0(d, n)] d=1,2 ..., D；N=1,2 ..., N
Wherein：
p0It is the element that Monitoring Data is concentrated, i.e., the prediction frequency in each observation cycle that acquisition module 501 is collected
Ambient interferences noise level instantaneous value (unit is dB)；
D is the number of days of collection；
N is the sample number of the prediction frequency ambient interferences noise level instantaneous value in a day, N=24 × 60/t0, wherein, t0
It is the scan period of spectrum monitoring equipment.
Then one day (24 hours) are divided into several time periods by average statistical module 502 by presetting granularity, and statistics is pre-
The ambient interferences noise level average of frequency measurement each time period of point.Background of the dataset generation module 503 according to each time period
Interference noise level average generation predicts the historical background interference noise level mean data collection of frequency.
If it is P to predict that the historical background interference noise level mean data of frequency integrates1, P1Expression formula be：
P1=[p1(d, m)] d=1,2 ..., D；M=1,2 ..., M
Wherein：
p1For the element that historical background interference noise level mean data is concentrated (unit is dB)；
M is the time hop count being divided into a day, and M=24 × 60/T, wherein T are presetting granularity (unit is minute).
Thus p can be derived1Expression formula be：
Wherein：
In a kind of specific embodiment, it is equal in structure background context interference noise level that data sequence builds module 100
Below equation is used during value sequence：
P1.d,m={ p1.p,q∈P1| 0 ＜ d-p≤w, | m-q |≤w } ∪ { p1.p,q∈P1| d=p, 0 ＜ m-q≤w }
Wherein, P1.d,mRepresent that the ambient interferences noise level of prediction d-th m-th time period in observation cycle of frequency is equal
Value, P1.p,qP-th ambient interferences noise level average of q-th time period in observation cycle is represented, w is that default perception data has
Effect scope, P1To predict the historical background interference noise level mean data collection of frequency.
Data sequence builds module 100 when background context interference noise electricity sequence of average is built, and is done in historical background
Disturb noise level mean data collection P1If in have for build background context interference noise electricity sequence of average time period the back of the body
Scape interference noise level average is lacked, then the ambient interferences noise level using several corresponding time periods in observation cycle is equal
The average value of value substitutes the ambient interferences noise level average of missing.If for example, d-th the 3rd time period in observation cycle
Ambient interferences noise level average lack, then data sequence build module 100 can with the 1st observe the cycle the 3rd time period,
2nd the 3rd time period in observation cycle ..., the 3rd ambient interferences noise level of time period in the d-1 observation cycle
The ambient interferences noise level average that mean of mean substitutes d-th the 3rd time period in observation cycle is calculated.
In a kind of specific embodiment, data sequence builds module 100 and is done in the background context for building known time section
When disturbing the equal value sequence of noise level, in historical background interference noise level mean data collection P1If in have currently known for building
The ambient interferences noise level average missing of several time periods of the background context interference noise electricity sequence of average of time period,
And the quantity of the time period of deleted background interference noise level average is done with for building the background context of currently known time period
The ratio of total amount of the time period of the equal value sequence of noise level is disturbed more than threshold value, then data sequence structure module 100 cancellation builds
The background context interference noise electricity sequence of average of currently known time period.For example, data sequence builds module 100 building currently
The background context interference noise electricity sequence of average of known time section needs the n ambient interferences noise level average of time period,
Wherein there is the b ambient interferences noise level average missing (b of time period<N), when b/n is more than threshold value, then data sequence builds
Module 100 cancels the equal value sequence of ambient interferences noise level for building the currently known time period, can again select known time section
Include calculating.General, the threshold value takes 30%.
The realization of above modules its concrete functions, can refer to above-mentioned background and bothers retouching for noise prediction method part
State, here is omitted.
Each technical characteristic of embodiment described above can be combined arbitrarily, to make description succinct, not to above-mentioned reality
Apply all possible combination of each technical characteristic in example to be all described, as long as however, the combination of these technical characteristics is not deposited
In contradiction, the scope of this specification record is all considered to be.
Embodiment described above only expresses several embodiments of the invention, and its description is more specific and detailed, but simultaneously
Can not therefore be construed as limiting the scope of the patent.It should be pointed out that coming for one of ordinary skill in the art
Say, without departing from the inventive concept of the premise, various modifications and improvements can be made, these belong to protection of the invention
Scope.Therefore, the protection domain of patent of the present invention should be determined by the appended claims.
Claims (10)
1. a kind of ambient interferences noise level Forecasting Methodology, it is characterised in that comprise the following steps：
According to the historical background interference noise level mean data collection and default perception data effective range of prediction frequency, build
The background context interference noise electricity sequence of average of prediction time and the background context interference noise level of known time section are equal
Value sequence；The historical background interference noise level mean data collection includes the prediction frequency in several different time sections
Ambient interferences noise level average；Known time section for the historical background interference noise level mean data concentrate when
Between section；All background context interference noises electricity sequence of average puts in order according to same rule；
The background context interference noise electricity sequence of average and the background context of prediction time for obtaining the known time section are disturbed
Euclidean distance between the equal value sequence of noise level；
Filter out background context interference noise electricity of the correspondence Euclidean distance less than several known time sections of pre-determined threshold
Sequence of average, extracts the ambient interferences noise level average of several known time sections, generates the prediction frequency
In the ambient interferences noise level predicted value collection of prediction time；
It is weighted according to the ambient interferences noise level predicted value collection, obtains the default frequency in prediction time
Ambient interferences noise level predicted value.
2. ambient interferences noise level Forecasting Methodology according to claim 1, it is characterised in that obtain as follows
The historical background interference noise level mean data collection of the prediction frequency：
The ambient interferences noise level instantaneous value in collection several observation cycles of prediction frequency；
One observation cycle is divided into several time periods according to presetting granularity, is made an uproar according to the ambient interferences in each time period
The ambient interferences noise level average of each time period of frequency is predicted described in the instantaneous Data-Statistics of vocal level；
The historical background interference noise electricity of frequency is predicted described in ambient interferences noise level average generation according to each time period
Average data collection.
3. ambient interferences noise level Forecasting Methodology according to claim 2, it is characterised in that built according to equation below
The background context interference noise electricity sequence of average of prediction time and the background context interference noise level of known time section are equal
Value sequence：
P1.d,m={ p1.p,q∈P1| 0 ＜ d-p≤w, | m-q |≤w } ∪ { p1.p,q∈P1| d=p, 0 ＜ m-q≤w }
Wherein, P1.d,mRepresent d-th background context interference noise level average of m-th time period in observation cycle of prediction frequency
Sequence, P1.p,qP-th ambient interferences noise level average of q-th time period in observation cycle is represented, w is default perception data
Effective range, P1To predict the historical background interference noise level mean data collection of frequency.
4. the ambient interferences noise level Forecasting Methodology according to Claims 2 or 3, it is characterised in that when prediction is built
The background context interference noise electricity sequence of average at quarter and the background context interference noise electricity sequence of average of known time section
When, if concentrated in the historical background interference noise level mean data having for building background context interference noise level average
The ambient interferences noise level average missing of the time period of sequence, then using the back of the body of several corresponding time periods in observation cycle
Scape interference noise level mean of mean substitutes the ambient interferences noise level average of missing.
5. the ambient interferences noise level Forecasting Methodology according to Claims 2 or 3, it is characterised in that when building known
Between section background context interference noise electricity sequence of average when, if the historical background interference noise level mean data concentrate
The background for having several time periods for the background context interference noise electricity sequence of average for building the currently known time period is done
Noise level average missing is disturbed, and the quantity of the time period of deleted background interference noise level average is currently known with for building
The ratio of the total amount of the time period of the background context interference noise electricity sequence of average of time period is more than threshold value, then cancel building and work as
The background context interference noise electricity sequence of average of preceding known time section.
6. a kind of ambient interferences noise level forecasting system, it is characterised in that including：
Data sequence builds module, for the historical background interference noise level mean data collection according to prediction frequency and default
Perception data effective range, builds the background context interference noise electricity sequence of average of prediction time and the phase of known time section
Close the equal value sequence of ambient interferences noise level；The historical background interference noise level mean data collection includes the prediction frequency
In the ambient interferences noise level average of several different time sections；The known time section is the historical background interference noise
The time period that level mean data is concentrated；All background context interference noises electricity sequence of average puts in order according to identical rule
Then；
Oldham distance calculating module, for obtain the background context interference noise electricity sequence of average of known time section with it is pre-
The Euclidean distance surveyed between the background context interference noise electricity sequence of average at moment；
Predicted value collection generation module, for filtering out phase of the correspondence Euclidean distance less than several known times section of pre-determined threshold
The equal value sequence of ambient interferences noise level is closed, the ambient interferences noise level average of several known times section, generation is extracted
Ambient interferences noise level predicted value collection of the prediction frequency in prediction time；
Weighted calculation module, for being weighted according to the ambient interferences noise level predicted value collection, obtains described pre-
If frequency is in the ambient interferences noise level predicted value of prediction time.
7. ambient interferences noise level forecasting system according to claim 6, it is characterised in that also including history data set
Module, the history data set module includes：
Acquisition module, the ambient interferences noise level instantaneous value for gathering several observation cycles of prediction frequency；
Average statistical module, for according to presetting granularity to be divided into several time periods one observation cycle, during according to each
Between ambient interferences noise level instantaneous value in section, the ambient interferences noise level for counting prediction each time period of frequency is equal
Value；
Dataset generation module, for predicting frequency described in the ambient interferences noise level average generation according to each time period
Historical background interference noise level mean data collection.
8. ambient interferences noise level forecasting system according to claim 7, it is characterised in that the data sequence builds
Module builds the background context interference noise electricity sequence of average of prediction time and the phase of known time section according to equation below
Close the equal value sequence of ambient interferences noise level：
P1.d,m={ p1.p,q∈P1| 0 ＜ d-p≤w, | m-q |≤w } ∪ { p1.p,q∈P1| d=p, 0 ＜ m-q≤w }
Wherein, P1.d,mRepresent d-th background context interference noise level average of m-th time period in observation cycle of prediction frequency
Sequence, P1.p,qP-th ambient interferences noise level average of q-th time period in observation cycle is represented, w is default perception data
Effective range, P1To predict the historical background interference noise level mean data collection of frequency.
9. ambient interferences noise level forecasting system according to claim 8, it is characterised in that the data sequence builds
Background context interference of the module in the background context interference noise electricity sequence of average and known time section for building prediction time
During the equal value sequence of noise level, if concentrated in the historical background interference noise level mean data having for building background context
The ambient interferences noise level average missing of the time period of interference noise electricity sequence of average, then using several observation cycles pair
The ambient interferences noise level mean of mean of the time period answered substitutes the ambient interferences noise level average of missing.
10. ambient interferences noise level forecasting system according to claim 9, it is characterised in that the data sequence structure
Block is modeled when the background context interference noise electricity sequence of average of known time section is built, in the historical background interference noise
If if level mean data is concentrated having for building the electric sequence of average of the background context interference noise of currently known time period
The dry ambient interferences noise level average missing of time period, and the time period of deleted background interference noise level average quantity
Ratio with the total amount of the time period of the background context interference noise electricity sequence of average for building the currently known time period is big
In threshold value, then cancel the background context interference noise electricity sequence of average for building the currently known time period.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510221994.4A CN104935388B (en) | 2015-04-30 | 2015-04-30 | Background interference noise level prediction method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510221994.4A CN104935388B (en) | 2015-04-30 | 2015-04-30 | Background interference noise level prediction method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104935388A CN104935388A (en) | 2015-09-23 |
| CN104935388B true CN104935388B (en) | 2017-05-24 |
Family
ID=54122376
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510221994.4A Active CN104935388B (en) | 2015-04-30 | 2015-04-30 | Background interference noise level prediction method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104935388B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106535232B (en) * | 2016-12-16 | 2019-11-19 | 京信通信系统（中国）有限公司 | A kind of frequency point optimization method and device |
| CN110190924A (en) * | 2019-05-17 | 2019-08-30 | 中科融通物联科技无锡有限公司 | A kind of frequency hopping unmanned plane defence method and device |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP2566074A1 (en) * | 2010-05-06 | 2013-03-06 | ZTE Corporation | Method and device for predicting signal to interference and noise ratio |
| CN103023590A (en) * | 2012-12-08 | 2013-04-03 | 北京工业大学 | Acquisition and processing system of global system for mobile communications for railway (GSM-R) network interference signals |
| CN103384173A (en) * | 2013-01-07 | 2013-11-06 | 沈阳远方计控系统有限公司 | Interference signal avoiding method of wireless remote transmission multi-frequency point set copy system |
Family Cites Families (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8676216B2 (en) * | 2010-06-29 | 2014-03-18 | Qualcomm Incorporated | Method and apparatus for mitigating interference in femtocell deployments |
| CN102546052B (en) * | 2011-12-12 | 2015-03-25 | 中国船舶重工集团公司第七二四研究所 | Self-adaptive threshold controlling and parameter calculating and measuring technology and method |
| CN103440871B (en) * | 2013-08-21 | 2016-04-13 | 大连理工大学 | A kind of method that in voice, transient noise suppresses |
| CN104467996A (en) * | 2013-09-24 | 2015-03-25 | 天津市鸿岚科技有限公司 | Electromagnetic interference monitoring device |
| CN104184530B (en) * | 2014-08-29 | 2016-05-11 | 大唐移动通信设备有限公司 | A kind of detection method of uplink interference and device |
| CN104467995B (en) * | 2014-10-24 | 2016-08-17 | 西安电子科技大学 | Blind primary user detection based on HMM and transmission level recognition methods |
-
2015
- 2015-04-30 CN CN201510221994.4A patent/CN104935388B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP2566074A1 (en) * | 2010-05-06 | 2013-03-06 | ZTE Corporation | Method and device for predicting signal to interference and noise ratio |
| CN103023590A (en) * | 2012-12-08 | 2013-04-03 | 北京工业大学 | Acquisition and processing system of global system for mobile communications for railway (GSM-R) network interference signals |
| CN103384173A (en) * | 2013-01-07 | 2013-11-06 | 沈阳远方计控系统有限公司 | Interference signal avoiding method of wireless remote transmission multi-frequency point set copy system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104935388A (en) | 2015-09-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104868960B (en) | Ambient interferences noise level Forecasting Methodology and system | |
| CN102624468B (en) | Automatic broadband detection method based on dual fast Fourier transformation (FFT) | |
| CN102147429B (en) | Broadband microwave power meter and interference signal filtering method | |
| CN104168075B (en) | Frequency spectrum detecting method and device in the case of a kind of without knowledge of noise covariance | |
| CN109474355A (en) | Adaptive noise THRESHOLD ESTIMATION and method for extracting signal based on spectrum monitoring data | |
| CN113009566B (en) | Local earthquake motion prediction model and construction method thereof | |
| CN104935388B (en) | Background interference noise level prediction method and system | |
| CN112287796B (en) | Radiation source identification method based on VMD-Teager energy operator | |
| CN110632386B (en) | Solar radio interference filtering method, readable storage medium and electronic equipment | |
| CN108760310A (en) | Accidental resonance Fault Diagnosis of Roller Bearings based on novel signal-to-noise ratio index | |
| CN103700109B (en) | SAR image change detection based on multiple-objection optimization MOEA/D and fuzzy clustering | |
| CN102158868B (en) | Based on acquisition methods and the system of the interference matrix of frequency sweep | |
| CN109933933B (en) | Noise treatment method and equipment | |
| CN107526064A (en) | Adaptive LFM modulated parameter estimating methods based on two dimensional character | |
| CN116821685B (en) | Sound signal identification method and system for cruising robot | |
| CN110189064A (en) | The recognition methods and system of section overproof water quality risk pressure source based on big data | |
| CN109034088A (en) | A kind of unmanned plane signal detection method and device | |
| CN117172601A (en) | Non-invasive load monitoring method based on residual total convolution neural network | |
| CN104977602A (en) | Control method and apparatus for earthquake data acquisition construction | |
| CN117390378A (en) | Intelligent management method and system for dual-carbon platform data | |
| CN107040269A (en) | Pole based on variance medium filtering/ultralow frequency channel atmospheric noise suppressing method | |
| Sharma et al. | Optimised CNN for human counting using spectrograms of probabilistic WiFi CSI | |
| KR100851038B1 (en) | Multiple method for removing noise included in partial discharge signal | |
| CN113702612B (en) | River bank pollution detection and investigation site selection method and system based on artificial intelligence | |
| CN105721081B (en) | A kind of cognitive wireless radio cooperation frequency spectrum sensing method for gathering gesture estimation based on user |
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 |