CN103402220B - A kind of device and method obtained and optimize DTMB SFN coverage rate - Google Patents

Abstract

The invention provides a kind of device and method obtained and optimize DTMB SFN coverage rate, for ground digital television single frequency network network planning and optimization.Device includes that module, statistical module, optimization module, GIS module, geographic information database and covering result database are analyzed in station message processing module, station data base, radio wave propagation budget module, interference.The present invention defines synchronization thresholding to embody SFN gain, use K LNM method that the available power and jamming power that receive place are synthesized, the relevant moments method of coupling is utilized to ask synthesis available signal power and the correlation coefficient of cumulative interference power, the prediction making the spot probability of coverage cell considers the dependency that SFN self-interference and shadow fading cause, and utilizes the spot probability of CCDF function computation grid.Budget result of the present invention is actual, accurate, and amount of calculation is little, makes SFN self-interference minimum under conditions of not changing other parameters of transmitter, and without increasing cost.

Description

A kind of device and method obtained and optimize DTMB SFN coverage rate

Technical field

The present invention relates to a kind of acquisition and optimization DTMB (Digital Terrestrial Television Multimedia Broadcasting) method and apparatus of SFN coverage rate, belong to ground digital television single frequency network network rule Draw and optimisation technique field.

Background technology

SFN disposes form because its plurality of advantages becomes main terrestrial DTV overlay network, all in SFN Transmitter uses identical frequency to launch same signal simultaneously.In addition to natural multipath, SFN itself produces " artificial multi-path ", OFDM (Orthogonal Frequency Division Multiplexing, orthogonal frequency division multiplexi) makes SFN become May, OFDM protect the size at interval determine in SFN the ultimate range between transmitter, when the distance of transmitter is big When the distance that protection interval is corresponding, interference will be produced from.The method reducing SFN self-interference region has a lot, such as adjusts Transmitting station site, transmitting power, antenna parameter and emission delay, the cost adjusting emission delay in these methods is 0, so Adjust emission delay and become a kind of common method reducing SFN self-interference.But, current engineering use manual type adjust Whole emission delay, poor accuracy；During SFN is built in China, transmitter number is more, each transmitter additional emission The possible combined number of time delay is huge and has dependency, and this makes to manually adjust the most complicated, it is achieved difficulty.The each transmitting of visible adjustment Machine additional emission time delay thus to improve SFN coverage rate be a multiparameter, nonlinearity, complicated combinatorial optimization problem, Therefore seek a kind of method solving transmitter additional emission time delay optimum combination efficiently to necessitate.

Accurately setting up of SFN coverage rate budget model is to realize premise and the basis that SFN coverage rate optimizes.At list In frequency net, receiver receives from different transmitters, the multiple signals with different delay simultaneously, signal be useful signal or Interference signal depends on this signal and fast Fourier change (FFT, Fast Fourier Transform) in OFDM receiver The delay inequality of window, also needs to synthesize each available power and jamming power at place receiving.

SFN is applied to digital audio broadcasting (DAB, the Digital Audio in Europe the earliest the nineties in last century Broadcasting) with DVB (DVB, Digital Video Broadcasting), cover at budget SFN During the covering spot probability of community, document [1]-[4] utilize weighting function signal calculated receiving the useful work produced at place Rate and jamming power, have ignored SFN certainly when calculating the variance of the difference of dB value of synthesis available power and cumulative interference power Interference and signal shadow fading cause dependency, and the calculating making coverage cell cover spot probability produces error ([1] R.Rebhan,Jens Zander.On the Outage Probability in Single Frequency Networks for Digital Broadcasting[J].IEEE Transactions on Broadcasting,1993,39(4):395- 400;[2]Agnes Ligeti,Jens Zander,Minimal Cost Coverage Planning for Single Frequency Networks[J].IEEE Transactions on Broadcasting,1999,45(1):78-87;[3] G.Koutitas.DVB network optimisation for energy efficiency[C].//Proc.IEEE Int.Conf.Adv.Commun.Technol.(ICACT),2010,2:7–10;[4]Lanza M,Gutierrez A L, Barriuso I,Perez J R.Coverage Optimization in Frequency Networks Using Simulated Annealing[C].//IEEE International Symposium on Antennas and Propagation and USNC/URSI National Radio Science Meeting(APSURSI),2011,7: 2789-2792).Existing commercial, professional terrestrial digital broadcasting signal coverage prediction software, such as LS company of Germany in the world CHIRplus_bc and the ICS-telecom of ADTI company of France, these softwares, in addition to supporting DAB and DVB standard, also prop up Hold the ATSC standard of the U.S. and the ISDB-T standard of Japan.

In August, 2006, China has promulgated the terrestrial DTV multimedia broadcasting standards with Chinese independent intellectual property right DTMB, China's DTMB standard do not supported by commercial, the terrestrial digital broadcasting signal coverage prediction software of specialty in the world, and this A little business software interface cipherings and fixing, buy and upgrade cost is expensive.At present, the DTMB SFN coverage prediction of domestic specialty System discloses a kind of digital single frequency network signal almost without, Publication No. 101795162 Chinese patent on August 4th, 2010 Covering quality method for measuring, calculating and evaluating, the method uses the spot probability of Monte Carlo method computation grid point, and the accuracy of calculating takes Certainly in number of sampling points, the most amounts of calculation of number of sampling points are the biggest；Secondly, the method uses power and method to carry out signal syntheses, makes Became estimating field strength intensity, document [4] by calculate lognormal variable sum method contrasted, power and, simplify In phase multiplication, LNM method, the degree of accuracy that LNM method is estimated is the highest.

Document [5]-[6] utilize population (PSO, Particle Swarm Optimization) optimized algorithm ([5] J.Morgade,J.Pérez,J.Basterrechea,M.Toca García,A.Arrinda,P.Angueira,Coverage optimization for DVB T/H single frequency networks using a PSO algorithm[C]// Pablo Angueira.IEEE International Symposium on Broadband Multimedia Systems and Broadcasting,BMSB2009.USA:IEEE Computer Society,2009:13–15;[6]J.Morgade, J.Pérez,J.Basterrechea,A.Arrinda,P.Angueira,Optimization of the coverage area for DVB-T single frequency networks using a particle swarm based method[C]// WilliamC.Y.Lee.IEEE Vehicular Technology Conference,VTC Spring2009,USA: Institute of Electrical and Electronics Engineers Inc.,2009:26–29).Genetic algorithm (GA, genetic algorithm) and PSO algorithm are all bionic Algorithm, but PSO is mainly used in continuous problem, once grain Sub-undue concentration, it is possible to be absorbed in local minimum.Document [7]-[8] being pointed out, GA can also be applied to dispersed problem, and Overcome conventional search methods and be absorbed in the shortcoming of locally optimal solution, there is preferable global search performance, the operation machine of its uniqueness Reason makes it solve, with limited cost, the combinatorial optimization problem that search volume is big, complicated.Therefore, genetic algorithm be widely used with In engineering optimization.Being optimized SFN coverage rate based on genetic algorithm is a direction studied.([7] Song Dan, Zhang Xiao Woods. the research of multisystem compatible receiver frequency point select permeability based on fixpoint theory realizes [J] with genetic algorithm. physics Report, 2010,59 (9): 6697-6704;[8] Wu Zhiyong, Guo Hong, Lv Zhenhua, Qian Hao. double margin brushless based on genetic algorithm are straight Stream Motor Optimizing Design [J]. BJ University of Aeronautics & Astronautics journal .2011,37 (12): 1541-1545).

Summary of the invention

The present invention, in order to improve SFN coverage rate predictablity rate, overcomes transmitter number to manually adjust time more additional The difficult point of emission delay, it is provided that a kind of acquisition and the apparatus and method optimizing DTMB SFN coverage rate.

The present invention obtains and the device optimizing DTMB SFN coverage rate, including: station message processing module, station number Module, statistical module, optimization module, GIS(Geographic is analyzed according to storehouse, radio wave propagation budget module, interference Information System, GIS-Geographic Information System) module, geographic information database and covering result database.

Station message processing module is each information launching the station in collecting SFN, and the information launching the station includes: site Title, longitude and latitude, transmitting power, antenna radiation pattern, signal modulation system, tranmitting frequency, transmitting time delay, state and network are compiled Number, and the information classification launching the station is stored in station data base.Launch the transmitting power of the station, antenna radiation pattern, letter Number modulation system, tranmitting frequency, transmitting time delay, state and network numbering, add in terminal and configure.Geographic information database is used In storage electronic chart.GIS module loading electronic chart, and generate grid screen figure layer according to the computational accuracy arranged, by result It is stored in geographic information database.

Radio wave propagation budget module obtains each information launching the station of state of activation from station data base, from GIS mould Block obtains the geography information of each grid central point in selection area, calculates state of activation each and launch the station in the range of selecting The link load of each grid central point, reception power, field intensity value, propagation time and the time of advent, and send result to GIS Module and statistical module are analyzed in module, interference.GIS module is according to the maximum field intensity values computation grid color value of each grid and one by one Coloring, covers output result on electronic chart.

Interference is analyzed module and is used for implementing function such as: first, arrange reference time t₀Selection standard, frame head and frame Length, receiver background noise N₀, synchronize threshold CNR Margn, synchronize thresholding P_syn, minimum intermediate value field intensity E_medProtect with radio frequency Protect rate PR, select or weighting function w (Δ τ is set_i), selecting non-receiver to be received and situational variables, situational variables is that load is made an uproar dry Than (CNIR) value or covering spot probability；Secondly, reference time t is determined₀, determine that each transmitter is in the range of selected in each grid The time of advent of heart point is relative to reference time t₀Delay inequality, and obtain the weighting letter of each grid available power and jamming power Number w (Δ τ_i) value, Δ τ_iRepresent the ToA time of advent of i-th transmitter_iWith t₀Delay inequality；Then, signal level is set Synchronized by receiver, according to weighting function w (Δ τ at the signal synchronizing more than thresholding_i) calculate each transmitter at each grid center The available power of point and jamming power, use K factor lognormal method (K-LNM method), and the synthesis obtaining each grid central point has By power U and the average of the dB value of cumulative interference power I and variance；Finally, if situational variables is CNIR, calculate covering field intensity big In the CNIR value in the region equal to minimum intermediate value field intensity, the CNIR value in other regions represents with 255, the CNIR value calculated is sent To GIS module and statistical module；If situational variables is for covering spot probability, calculates and cover field intensity more than or equal to minimum intermediate value field intensity Region in the synthesis available power of each grid and the correlation coefficient of cumulative interference power and cover spot probability, in other regions The covering spot probability of grid is 0, and the spot probability of grid is sent to GIS module and statistical module.Reference time t₀Have two Kind of selection standard: the first is to arrive the earliest and power is more than synchronizing thresholding P_synSignal the time of advent on the basis of, second Planting is on the basis of the time of advent of level peak signal；Synchronize thresholding P_syn=N₀+Margn.GIS module is according to each grid Cover the color comparator scope that spot probability value or CNIR value define with system to compare, carry out each grid color value calculating one by one Coloring, output result covers on electronic chart.

Statistical module, according to the statistical condition arranged, filters out the grid meeting statistical condition in statistical regions, and will The selection result is sent to GIS module, the percentage ratio of total grid number in calculating statistical regions shared by the grid meeting statistical condition, And be sent to optimize module.Statistical module is sent to GIS module selected statistical regions feature, obtains statistics from GIS module Grid in region and total number thereof.The budget result meeting the grid that statistical condition requires is covered electronically by GIS module On figure, it is unsatisfactory for the budget result not display of the grid that statistical condition requires.

Optimize module to calculate based on genetic algorithm and optimization areal coverage can be made to reach in the highest SFN respectively launch The emission delay optimum combination of the station and coverage rate.In every generation evolutionary process of genetic algorithm, by each in present age population The value of time delay vector is sent to station data base, respectively launches the transmitting time delay of the station at station database update, from GIS module In middle acquisition optimization region, the geography information of each grid central point, recalculates state of activation by radio wave propagation budget module Each station of launching to optimizing the time of advent of each grid central point in region, by interference analyze module select state for activating and Nonactivated transmitter is as non-desire receiver/transmitter, and calculates covering spot probability or the CNIR value of each grid, by statistics mould Block is the percentage ratio of total grid number in calculating favored area shared by the grid meeting statistical condition.

Cover result database for storing the calculating knot of radio wave propagation budget module, interference analysis module and statistical module Really.

The present invention obtains and the method optimizing DTMB SFN coverage rate, including below step:

Step A: load planning region electronic chart and arrange computational accuracy, GIS module generates according to the computational accuracy arranged Grid screen figure layer, transfers the geography information of each grid central point in grid screen from geographic information database, and geography information includes Longitude, latitude, height and increased surface covering.

Step B: add SFN and respectively launch the station and configure the parameter of each transmitting station, parameter includes launching power, sending out Radio frequency rate, transmitter antenna gain (dBi), transmitting antenna feeder loss, transmitting antenna radiation pattern, signal modulation system, emission delay, shape State and network numbering, state is for activating or not activating.

Step C: arrange link loss budget parameter, including arranging propagation model, mulching material decay and link load Budget distance, and reception antenna height, gain and minimum intermediate value field intensity are set.

Step D: calculate each transmitting station link load to each grid central point of state of activation, receive power, field intensity Value, propagation time and the time of advent.

Step E: according to the maximum field intensity values computation grid color value of each grid and colour, output result covers at electricity On sub-map and at terminal demonstration.

Step F: first, is set to state of activation by transmitter each in SFN, and the outside co-channel transmitter of SFN is arranged For unactivated state.Then, the reference time t of FFT window original position is set₀Selection standard, frame head and frame length, connect Receipts machine background noise N₀, synchronize threshold CNR Margn, synchronize thresholding P_syn, minimum intermediate value field intensity E_med, radio-frequency protection ratio PR, Select or weighting function w (Δ τ is set_i).Finally, selecting non-receiver to be received and situational variables, situational variables is that load is made an uproar dry ratio Value (CNIR, Carrier to Noise plus Interference Ratio) or covering spot probability.

Reference time t₀There are two kinds of selection standards: the first is to arrive the earliest and power is more than P_synThe arrival of signal On the basis of time, the second is on the basis of the time of advent of level peak signal.Synchronize thresholding P_syn=N₀+Margn。

Step G: determine reference time t₀, and each transmitter time of advent is relative to reference time t₀Delay inequality Δ τ_i, Obtain weighting function w (the Δ τ of each grid available power and jamming power_i) value；Signal level is set and is synchronizing more than thresholding Signal synchronized by receiver, according to weighting function w (Δ τ_i) calculate each transmitter each grid central point available power and Jamming power, obtains synthesis available power U and the lognormal distribution parameter of cumulative interference power I of each grid central point.If Situational variables is to carry dry ratio of making an uproar, and calculates the CNIR value covering field intensity more than or equal to the region of minimum intermediate value field intensity, other regions CNIR value represents with 255, and the CNIR value of calculating is sent to GIS module；If covering spot probability, interference is analyzed module and is calculated Cover field intensity more than or equal to the synthesis available power of each grid in the region of minimum intermediate value field intensity and being correlated with of cumulative interference power Coefficient and covering spot probability, in other regions, the covering spot probability of grid is 0, is sent to by the covering spot probability of grid GIS module.

Step H:GIS module is according to the color comparator model covering spot probability value or CNIR value and system definition of each grid Enclosing and compare, carry out each grid color value and calculate and colour, output result covers on electronic chart and at terminal demonstration.

Step I: according to statistical regions needed for geographical feature selection, statistical condition is set.

Step J: meet the grid of statistical condition in filtering out selected areas, and the selection result is sent to GIS module, and The percentage ratio of total grid number in meeting statistical regions shared by the grid number of statistical condition in calculating selected areas.

The budget result meeting the grid that statistical condition requires is covered on electronic chart and shows by step K:GIS module In terminal, it is unsatisfactory for the budget result not display of the grid that statistical condition requires.

Step L: optimize region according to needed for geographical feature selection, arranges maximum genetic algebra, population scale.

Step M: in the case of not changing transmitter parameter, calculates based on genetic algorithm and can make optimization areal coverage Reach emission delay optimum combination and the coverage rate of each transmitter in the highest SFN.

Step N: export emission delay optimum combination and the coverage rate of each transmitter.

In described step D, can use ITU-R P.1546, ITU-R P.370, ITU-R P.526 or Okumura- In Hata radio waves propagation model and calibration model budget SFN thereof, each transmitting station is to the reception power of each grid central point And field intensity value.

In step G, synthesis available power U of grid central point and the cumulative interference power I of grid central point are respectively as follows:

$U=\underset{i=1}{\overset{N_1}{\mathrm{\Σ}}}{U}_i=\underset{i=1}{\overset{N_1}{\mathrm{\Σ}}}{P}_{\mathrm{ri}}w\left(\mathrm{\Δ}{\mathrm{\τ}}_i\right),$ $I=\underset{i=1}{\overset{N_1}{\mathrm{\Σ}}}{I}_i+\underset{j=1}{\overset{N_2}{\mathrm{\Σ}}}{I}_j+\underset{k=1}{\overset{N^{\′\′}}{\mathrm{\Σ}}}{I}_k=\underset{i=1}{\overset{N_1}{\mathrm{\Σ}}}{P}_{\mathrm{ri}}[1-w\left({\mathrm{\Δ\τ}}_i\right)]+\underset{j=1}{\overset{N_2}{\mathrm{\Σ}}}{P}_{\mathrm{rj}}+\underset{k=1}{\overset{N^{\′\′}}{\mathrm{\Σ}}}{P}_{\mathrm{rk}};$

Wherein, N₁For receiving power in SFN more than P_synTransmitting number of stations, N₂Little for receiving power in SFN In P_synTransmitting number of stations, N'' is the outside co-channel transmitter number of SFN, U_iPlace product is being received for i-th transmitter Raw available signal power, P_riThe local power received for i-th transmitter, I_iPlace generation is being received for i-th transmitter Interfering signal power.

P_riStatistical property obeys logarithm normal distribution, and the dB value changed is W_ri=10 log₁₀(P_ri)；" " represents multiplication Computing；

W_riAverageStandard varianceWherein,It it is the standard side of shadow fading Difference, takes 5.5dB in ground digital television broadcast；Represent the local average power that i-th transmitter receives；

U_iDB value X converted_i=10 log₁₀(U_i) averageAnd standard varianceIt is respectively as follows:

$m_{X_i}=m_{W_{\mathrm{ri}}}+10\·{\log }_{10}\left(w\left({\mathrm{\Δ\τ}}_i\right)\right)=10\·{\log }_{10}\left(\stackrel{\‾}{P_{\mathrm{ri}}}\right)+10\·{\log }_{10}\left(w\left({\mathrm{\Δ\τ}}_i\right)\right),$ ${\mathrm{\σ}}_{X_i}={\mathrm{\σ}}_{W_{\mathrm{ri}}}=5.5\mathrm{dB};$

I_iDB value Y converted_i=10 log₁₀(I_i) averageAnd standard varianceIt is respectively as follows:

$m_{Y_i}=m_{W_{\mathrm{ri}}}+10\·{\log }_{10}(1-w\left({\mathrm{\Δ\τ}}_i\right))=10\·{\log }_{10}\left(\stackrel{\‾}{P_{\mathrm{ri}}}\right)+10\·{\log }_{10}(1-w\left({\mathrm{\Δ\τ}}_i\right)),$ ${\mathrm{\σ}}_{Y_i}={\mathrm{\σ}}_{W_{\mathrm{ri}}}=5.5\mathrm{dB};$

DB value X and Y that synthesis available power U and cumulative interference power I convert are respectively as follows:

X=10log₁₀(U), Y=10log₁₀(I)；

The relevant moments method of coupling is utilized to calculate synthesis available power and the correlation coefficient of cumulative interference power of each grid, specifically It is:

First, X is made_i'=ln (U_i), then X_i'=λ X_i, parameterThen X_i' averageWith Standard varianceIt is respectively as follows: In like manner, Y is made_i'=ln (I_i), X'=ln (U), Y'=ln (I), Then Y_i' and Y_i, X' Yu X, Y' Yu Y be respectively provided with X_i' and X_iBetween identical relation, and X_i' and Y_i' correlation coefficientEqual to X_i And Y_iCorrelation coefficientThe correlation coefficient r of X' Yu Y'_X'Y'Correlation coefficient r equal to X and Y_XY；

Make variable φ=X'+Y', then average m of φ_φ=m_X'+m_Y', variances sigma_φ ²=σ_X' ²+2r_X'Y'σ_X'σ_Y'+σ_Y' ²；Coupling U Correlation Moment with I $E\left[\mathrm{UI}\right]=E\left[e^{X^{\′}}{e}^{Y^{\′}}\right]=E\left[e^{X^{\′}+Y^{\′}}\right]=E\left[e^{\mathrm{\φ}}\right]=e^{m_{\mathrm{\φ}}+{{\mathrm{\σ}}_{\mathrm{\φ}}}^2/2},$ Correlation Moment E [UI] can also be expressed as:

$E\left[\mathrm{UI}\right]=E\left[e^{X^{\′}}{e}^{Y^{\′}}\right]=E\left[e^{X^{\′}+Y^{\′}}\right]$

$=E\left[(e^{{X_1}^{\′}}+e^{{X_2}^{\′}}+...+e^{{X_{N_1}}^{\text{\′}}})(e^{{Y_1}^{\′}}+e^{{Y_2}^{\′}}+...+e^{{Y_{N_1+N_2+N^{\′\′}}}^{\′}})\right]$

$=\underset{i=1}{\overset{N_1}{\mathrm{\Σ}}}\underset{i=1}{\overset{N_1+N_2+N^{\′\′}}{\mathrm{\Σ}}}E\left[\left(e^{{X_i}^{\′}+{Y_i}^{\′}}\right)\right]$

$=\underset{i=1}{\overset{N_1}{\mathrm{\Σ}}}\underset{i=1}{\overset{N_1+N_2+N^{\′\′}}{\mathrm{\Σ}}}{e}^{m_{{X_i}^{\′}}+m_{{Y_i}^{\′}}+1/2\·({\mathrm{\σ}}_{X_i^{\′}}^2+{\mathrm{\σ}}_{Y_i^{\′}}^2+2r_{X_i^{\′}{Y}_i^{\′}}{\mathrm{\σ}}_{X_i^{\′}}{\mathrm{\σ}}_{Y_i^{\′}})}$

$=v$

Thus obtain: $r_{\mathrm{XY}}=r_{X^{\′}{Y}^{\′}}=\frac{2[\ln \left(v\right)-(m_{X^{\′}}+m_{Y^{\′}})]-({{\mathrm{\σ}}_{X^{\′}}}^2+{{\mathrm{\σ}}_{Y^{\′}}}^2)}{{2\mathrm{\σ}}_{X^{\′}}{\mathrm{\σ}}_{Y^{\′}}},$ WithIt is respectively Y_i' equal Value and standard variance.

The invention has the beneficial effects as follows: obtaining and device and the side optimizing DTMB SFN coverage rate according to the present invention Method, definition synchronizes thresholding P_synTo embody SFN gain, power is at P_synAbove signal could be synchronized by receiver, simultaneously root Border DTMB receiver interference model calculates each station at the available signal power of grid central point and interfering signal power factually, adopts Carry out signal syntheses by K-LNM method, utilize the relevant moments method of coupling to calculate synthesis available power and the phase relation of cumulative interference power Number, utilizes the spot probability of CCDF function computation grid, and budget result is actual, accurate, and amount of calculation is little, it is possible to increase SFN Coverage rate predictablity rate, is simultaneously based on genetic algorithm and calculates the emission delay optimum combination of each transmitter so that SFN covers Rate is the highest, it is not necessary to change the parameter such as transmitter power, antenna, cost is 0, overcome transmitter number more time manually adjust attached Add the difficult point of emission delay, and accuracy is high.Apparatus and method of the present invention can apply in China's DTMB SFN planning.

Accompanying drawing explanation

Fig. 1 is obtaining and the composition structural representation of the device optimizing DTMB SFN coverage rate of the embodiment of the present invention；

Fig. 2 is obtaining and the functional schematic of the device optimizing DTMB SFN coverage rate of the embodiment of the present invention；

Fig. 3 is obtaining and the overall flow figure of the method optimizing DTMB SFN coverage rate of the embodiment of the present invention；

Fig. 4 is the workflow diagram of step 103 Wave Propagation Prediction of the inventive method；

Fig. 5 is the workflow diagram of the step 106 interference analysis module of the inventive method；

Fig. 6 is the workflow diagram of the optimization module of the step 112 of the inventive method.

Detailed description of the invention

Hereinafter, obtaining and device and the side optimizing DTMB SFN coverage rate of the present invention is described in detail with reference to accompanying drawing 1～6 Method.

Apparatus and method of the present invention definition synchronizes thresholding P_synTo embody SFN gain, power is at P_synAbove signal Could be synchronized by receiver, calculate each transmitter in SFN according to actual DTMB receiver interference model and receiving place product Raw available power and jamming power, use K-LNM method to carry out signal syntheses, utilizes the relevant moments method of coupling to seek synthesis useful signal Power and the correlation coefficient of cumulative interference power, make the prediction of the spot probability of coverage cell consider SFN self-interference and the moon The dependency that shadow decline causes, utilizes complementary Cumulative Distribution Function (CCDF, Complementary Cumulative Distribution Function) calculate the spot probability of each grid, budget result is actual, accurate, and amount of calculation is little.This Bright optimum combination based on transmitter emission delay each in genetic algorithm for solving SFN, is not changing other parameters of transmitter Under the conditions of make SFN self-interference minimum, the method without increasing cost, and overcome transmitter number more time manually adjust Transmitter emission delay is complicated and realizes difficult difficult point, and accuracy is high.

As depicted in figs. 1 and 2, obtaining of the embodiment of the present invention is tied with the device composition optimizing DTMB SFN coverage rate Structure, including: the transmitting station 1 in SFN, station message processing module 2, station data base 3, radio wave propagation budget module 4, Interference is analyzed module 5, statistical module 6, is optimized module 7, GIS module 8, geographic information database 9, covering result database 10 and Terminal 11.

The transmitting station 1 in SFN is for the object analyzed, and adds in terminal 11 and configures the ginseng of each transmitting station 1 Number, parameter includes launching power, antenna radiation pattern, signal modulation system, tranmitting frequency, transmitting time delay, state and network numbering Deng.State refers to that this transmitting station is in activation or unactivated state.

Station message processing module 2, launches the relevant information of the station 1 for collecting each SFN, including: site title, Longitude and latitude, transmitting power, antenna radiation pattern, signal modulation system, tranmitting frequency, transmitting time delay, state and network numbering etc., and The information launching the station 1 is stored in station data base 3.

Station data base 3, for storing and manage the data after station message processing module 2 classification processes, can be to the station Information batch processing, such as change launch power, tranmitting frequency, emission delay and state etc..

Radio wave propagation budget module 4, for obtaining each information launching the station 1 of state of activation from station data base 3 Data, and the geographic information data of each grid central point in the range of GIS module 8 obtains link loss budget, select ITU-R P.1546, ITU-R P.370, ITU-R P.526 or Okumura-Hata radio waves propagation model and calibration model budget activation thereof The each of state launches the station 1 link load, reception power, field intensity value, propagation time and arrival to each grid central point Time, result of calculation is sent to GIS module 8, interference analysis module 5 and statistical module 6.

Module 5 is analyzed in interference, first, arranges reference time t₀Selection standard, frame head and frame length, receiver background Noise N₀, synchronize threshold CNR Margn, synchronize thresholding P_syn, minimum intermediate value field intensity E_medWith radio-frequency protection ratio PR, select or set Put weighting function w (Δ τ_i), selecting non-receiver to be received and situational variables, it is dry than (CNIR) value or covering that situational variables is that load is made an uproar Spot probability；Secondly, reference time t is determined₀, determine that each transmitter (namely launching the station) is in the range of selected in each grid The time of advent of heart point is relative to reference time t₀Delay inequality, and obtain the weighting letter of each grid available power and jamming power Number w (Δ τ_i) value, Δ τ_iRepresent the ToA time of advent of i-th transmitter_iWith t₀Delay inequality.Then arrange signal level to exist The signal synchronizing more than thresholding is synchronized by receiver, according to weighting function w (Δ τ_i) calculate each transmitter at grid central point product Raw available power and jamming power, use K-LNM method to the available signal power of each grid central point and interference signal merit Rate carries out synthesis analysis, calculates the synthesis available power of each grid central point and the average of the dB value of cumulative interference power and side Difference.Finally, if situational variables is CNIR, calculate the CNIR value covering field intensity more than or equal to the region of minimum intermediate value field intensity, other The CNIR value in region represents with 255, and the CNIR value of calculating is sent to GIS module 8 and statistical module 6；If situational variables is for covering Lid spot probability, calculates covering field intensity and is more than or equal to synthesis available power and the synthesis of each grid in the region of minimum intermediate value field intensity The correlation coefficient of jamming power and covering spot probability, in other regions, the covering spot probability of grid is 0, by the place of grid Probability is sent to GIS module 8 and statistical module 6.The relevant moments method of coupling is utilized to calculate the synthesis available power of each grid central point With the correlation coefficient of cumulative interference power, concrete methods of realizing is as shown in step G of Fig. 2 and the inventive method.Utilize complementation Covering of Cumulative Distribution Function (CCDF, Complementary Cumulative Distribution Function) computation grid Lid spot probability.

Reference time t₀There are two kinds of selection standards: the first is to arrive the earliest and power is more than synchronizing thresholding P_synLetter Number the time of advent on the basis of, the second is on the basis of the time of advent of level peak signal；Synchronize thresholding P_syn=N₀+ Margn。

Statistical module 6, in statistical regions, according to the statistical condition arranged, filters out the grid meeting statistical condition, will The selection result is sent to GIS module 8, and the percentage of total grid number in calculating statistical regions shared by the grid meeting statistical condition Ratio, and be sent to optimize module 7.Wherein, 6 statistical regions features of statistical module are sent to GIS module 8, and from GIS module 8 Obtain the grid in statistical regions and total number thereof；

Optimize module 7, to maximize the coverage rate optimizing region as target, calculate in SFN each based on genetic algorithm Stand the optimum combination of emission delay, during optimizing, time delay vector value each in each generation population is sent to station data Storehouse 3, is sent to GIS module 8 by required optimization provincial characteristics, calls radio wave propagation budget module 4 and again predicts that each station is to excellent In change region, the time of advent of each grid central point, interference are analyzed module 5 and are selected state to be to activate and nonactivated transmitter work For non-desire receiver/transmitter, the calculating covering spot probability of each grid, statistical module 6 is selected to calculate coverage rate, thus calculate each The target function value of time delay vector.

GIS module 8 loads electronic chart, according to the computational accuracy information arranged by the digital map grid of selected areas Latice layer, and store the result in geographic information database 9.GIS module 8 is transferred selected from geographic information database 9 The geography information such as the height of each grid central point, longitude and latitude, increased surface covering in region, and it is sent to radio wave propagation budget module 4.GIS module 8 is according to the maximum field intensity values computation grid color value of each grid of the budget of radio wave propagation budget module 4 and one by one Coloring, covers output result on electronic chart and shows in terminal 11.GIS module 8 is analyzed module 5 according to interference and is calculated The covering spot probability value of each grid or CNIR value, compare with the color comparator scope of system definition, carry out each grid color Value calculates and colours one by one, and output result covers on electronic chart and shows in terminal 11.6 selected systems of statistical module Meter provincial characteristics is sent to GIS module 8, obtains the grid in statistical regions and total number thereof from GIS module 8, according to statistics mould The result of calculation of block 6, covers the budget result meeting the grid that statistical condition requires on electronic chart, is unsatisfactory for adding up bar The not display of the budget result of the grid that part requires, budget result refers to cover spot probability value or CNIR value.GIS module 8 from Optimize module 7 and obtain optimization region, by the geographical letter such as the height of grid central point, longitude and latitude, increased surface covering in optimization region Breath is to radio wave propagation budget module 4.

Geographic information database 9 is for storing electronic map information.

Cover result database 10 for storing radio wave propagation budget module 4, interference analyze module 5 and statistical module 6 Result of calculation.

Terminal 11 predicts the outcome for display.

It addition, the present embodiment at exercisable ITU-R P.370, ITU-R P.1564, ITU-R P.526, Okumura- Having selected Okumura-Hata model in the electric wave propagation prediction such as Hata, this model has taken into full account urban architecture covering Thing, the receiving area impact on radio wave propagation, be suitable for city and densely inhabited district's field intensity prediction.

It addition, the present invention uses GIS carry out the storage of geographic information data, obtain, analyze and show, on the one hand improve The accuracy of Wave Propagation Prediction, on the other hand shows covering and the interference calculation result of planning region intuitively.

It addition, the present invention is when loading electronic chart, user can arrange computational accuracy, generates new grid on electronic chart Latice layer, utilizes grid screen figure layer color method explicitly face digital television single-frequency network and covers and interference calculation result, user Can customize the size of grid screen, density, the most flexibly, practical；And grid colouring is simple, quickly.

With reference to accompanying drawing 2 and 3, the embodiment of the present invention obtain the schematic diagram with the method optimizing DTMB SFN coverage rate and Flow chart, the method comprises the following steps:

Step 100: start system, loads planning region electronic map data, including Terrain Elevation figure layer, striograph layer and ground Table covering figure layer, VectorLayer etc., and computational accuracy is set, the size of the most each grid, GIS module 8 is according to the meter arranged Calculate precision information and generate grid screen figure layer, from geographic information database 9, transfer the geography of each grid central point in grid screen Information, including longitude, latitude, highly, the information such as increased surface covering.

Step 101: load SFN and launch station information and configure each transmitting station parameter, including launching power, transmitting Frequency, transmitter antenna gain (dBi), transmitting antenna feeder loss, transmitting antenna radiation pattern, signal modulation system, additional transmitting are prolonged Time, state and network numbering etc..State refers to that this transmitting station is in activation or not state of activation.Employing GIS can be at map The information such as the upper position showing the respectively transmitting station in SFN directly perceived.

Step 102: link loss budget parameter is set in radio wave propagation budget module 4, including arrange propagation model, The parameters such as face covering decay, link loss budget distance, arrange reception antenna height, gain and minimum intermediate value field intensity.

Step 103: radio wave propagation budget module 4 calculates each station 1 of launching that state in each SFN is activated state and arrives each The link load of grid central point, reception power, field intensity value, propagation time and the time of advent.The result of calculating is sent to GIS Module 5 is analyzed in module 8 and interference.

Step 104:GIS module 8 carries out grid color value calculating according to the maximum field intensity values of each grid and colours, defeated Go out result cover on electronic chart and show in terminal 11.

Step 105: arrange transmitter state, is set to state of activation by transmitter each in SFN, and SFN is outside same Channel transmitter is set to unactivated state.The reference time t of FFT window original position is set in module 5 is analyzed in interference₀Choosing Select standard, frame head and frame length, receiver background noise N₀In (unit is dBw), synchronization threshold CNR Margn, minimum Value field intensity E_med, radio-frequency protection ratio (PR, Protection Ratio), select or weighting function w (Δ τ be set_i), selection analysis Variable is made an uproar dry ratio (CNIR, Carrier to Noise plus Interference for covering spot probability or load Ratio), non-desire receiver/transmitter is selected.

In order to embody the gain of SFN, the present invention is also provided with synchronize thresholding P_syn, power is at P_synAbove signal Can be synchronized by receiver, usual P_synLess than or equal to receiver sensitivity P_min。

Wherein synchronize thresholding P_syn=N₀+ Margn, unit is dBw, P_min=N₀+ C/N, unit be dBw, C/N be carrier-to-noise ratio Thresholding, usual Margn≤C/N, minimum equivalent field intensity E_minWith P_minBetween transformational relation be:

E_min(dB μ v/m)=P_min+L_f-G+20logf_c+105.06

Wherein, L_fFeeder loss (unit is dB), G for reception antenna are the antenna gain relative to half-wave dipole (unit is dBd), f_cFor operating frequency (unit is MHz).

Minimum intermediate value field intensity E_medWith minimum equivalent field intensity E_minBetween relation be:

E_med=E_min+P_mmn+C₁, fixed outdoor reception

E_med=E_min+P_mmn+C₁+L_h, mobile reception

E_med=E_min+P_mmn+C₁+L_h+L_b, indoor fixed reception

Wherein, C₁For the location correction factor；P_mmnFor artificial noise margin (unit is dB)；L_hFor High consumption, (unit is DB), point is received at ground more than 1.5m；L_bFor building penetration loss (unit is dB).

C₁=μ σ_t

Whereinσ_tFor total mean square deviation (unit is dB)；σ_m=5.5dB, for large scale mean square deviation, σ_b For building penetration loss mean square deviation (unit is dB)；μ is distribution factor, is 0.52 during spot probability 70%, is 1.28 when 90%, It is to be 2.33 when 1.64 and 99% when 95%.

The reference time t of FFT window original position in the present invention₀There are two kinds of selection standards: the first is to arrive the earliest And power is more than P_synSignal the time of advent on the basis of, the second is on the basis of the time of advent of level peak signal.

The present invention selects non-desire receiver/transmitter by the state of transmitter.Non-desire receiver/transmitter state can be:

1. activate；

2. activate with inactive；

The most inactive.

Step 106: determine reference time t₀, and each transmitter ToA time of advent_iRelative to reference time t₀Time delay Difference, if Δ τ_iRepresent the ToA time of advent of i-th transmitter_iWith t₀Delay inequality, and according to the weighting function w (Δ of step 105 τ_i) obtain each grid available power and the weighting function value of jamming power.Signal level is set at the signal synchronizing more than thresholding Synchronized by receiver, according to weighting function w (Δ τ_i) calculate each transmitter in the available power of grid central point and jamming power. Then synthesis available power and the lognormal distribution parameter of cumulative interference power of each grid central point are obtained.The present invention implements In example, situational variables is for covering spot probability, and interference is analyzed module 5 and calculated each grid covering spot probability.

Interference is analyzed module 5 and is determined the reference time t of FFT window original position₀And calculate each transmitter ToA time of advent_i Relative to this reference time t₀Delay inequality.In SFN, receive place and receive from different transmitters simultaneously, there is difference Multiple signals of time delay, from i-th transmitter signal receive place whether be that useful signal depends on its time of advent ToA_iWith t₀Delay inequality.In the embodiment of the present invention, each transmitting station is provided with a transmitter.If ToA_iWith t₀Delay inequality For Δ τ_i=ToA_i-t₀。

By the anti-echo interference to the commercial DTMB receiver that domestic primary receiver manufacturer produces in the present invention Performance test and emulation draw DTMB receiver interference model, with weighting function w (Δ τ_i) represent, it is used for calculating each SFN and sends out Penetrate station available power U at grid central point_iWith jamming power I_i.Weighting function w (Δ τ_i) it is:

$w\left({\mathrm{\&Delta;\&tau;}}_i\right)=\left\{\begin{array}{c}0,{\mathrm{\&Delta;\&tau;}}_i<-T_{\mathrm{GI}}\\ 1,-T_{\mathrm{GI}}\&le;{\mathrm{\&Delta;\&tau;}}_i\&le;T_{\mathrm{GI}}\\ 0,{\mathrm{\&Delta;\&tau;}}_i>T_{\mathrm{GI}}\end{array}\right.;$

Wherein, T_GIRepresent the protection interval of OFDM symbol, i.e. frame head.

Except both the above weighting function w (Δ τ_i) outward, w (Δ τ also can be write as required_i)。

From Okumura-Hata radio wave propagation theory, owing to shadow fading affects, the basis received from i-th transmitter Ground power P_riWith place change at random, can be expressed as:

$P_{\mathrm{ri}}=\frac{P_{\mathrm{ti}}{G}_{\mathrm{ti}}{G}_r}{L\left(d_i\right)}\&CenterDot;{10}^{{\mathrm{\&zeta;}}_i/10}=\stackrel{\&OverBar;}{P_{\mathrm{ri}}}\&CenterDot;{10}^{{\mathrm{\&zeta;}}_i/10}$

Wherein, P_ti、G_tiIt is ERP and the antenna gain of i-th transmitter respectively, d_iWith L (d_i) it is respectively The distance of i transmitted from transmitter to receiver and path loss, G_rRepresent the antenna gain of receiver,Represent that i-th transmitter connects The local average power received, ζ_iBeing zero-mean gaussian stochastic variable, be used for describing the shadow fading near receiver, unit is dB。

Visible, P_riStatistical property obeys logarithm normal distribution, dB value W of order conversion_ri=10 log₁₀(P_ri), then:

W_riAverageStandard varianceWherein,It it is the standard side of shadow fading Difference, takes 5.5dB in ground digital television broadcast.

I-th transmitter is receiving available signal power U that place produces_i=P_ri·w(Δτ_i), interfering signal power I_i =P_ri(1-w (Δ τ_i)), w (Δ τ in grid_i) vary less, can ignore, so U_iAnd I_iObedience lognormal is divided Cloth, then available signal power is converted into dB value X_i=10 log₁₀(U_i) average, standard variance is respectively as follows:

$m_{X_i}=m_{W_{\mathrm{ri}}}+10\&CenterDot;{\log }_{10}\left(w\left({\mathrm{\&Delta;\&tau;}}_i\right)\right)=10\&CenterDot;{\log }_{10}\left(\stackrel{\&OverBar;}{P_{\mathrm{ri}}}\right)+10\&CenterDot;{\log }_{10}\left(w\left({\mathrm{\&Delta;\&tau;}}_i\right)\right),$ ${\mathrm{\&sigma;}}_{X_i}={\mathrm{\&sigma;}}_{W_{\mathrm{ri}}}=5.5\mathrm{dB}$

Interfering signal power is converted into dB value Y_i=10 log₁₀(I_i) average, standard variance is respectively as follows:

$m_{Y_i}=m_{W_{\mathrm{ri}}}+10\&CenterDot;{\log }_{10}(1-w\left({\mathrm{\&Delta;\&tau;}}_i\right))=10\&CenterDot;{\log }_{10}\left(\stackrel{\&OverBar;}{P_{\mathrm{ri}}}\right)+10\&CenterDot;{\log }_{10}(1-w\left({\mathrm{\&Delta;\&tau;}}_i\right)),$ ${\mathrm{\&sigma;}}_{Y_i}={\mathrm{\&sigma;}}_{W_{\mathrm{ri}}}=5.5\mathrm{dB}$

In SFN, receive the synthesis available power in placeCumulative interference power N' represents SFN internal transmitter total number, and N'' is the outside co-channel transmitter number of SFN.The present invention utilizes K factor Lognormal method (K-LNM) carries out signal syntheses, and signal intensity is estimated actual, tries to achieve the synthesis available power of grid central point U, the Parameters of Normal Distribution of dB value of cumulative interference power I.In the embodiment of the present invention, situational variables is for covering spot probability, interference Analysis module 5 calculates covering field intensity and does with synthesis more than or equal to the synthesis available power of each grid in the region of minimum intermediate value field intensity Disturbing the correlation coefficient of power and cover spot probability, in other regions, the covering spot probability of grid is 0, by the covering ground of grid Point probability is sent to GIS module 8.The present invention calculates synthesis available power and cumulative interference power by the relevant moments method of coupling Correlation coefficient, makes the prediction of the spot probability of coverage cell consider the dependency that SFN self-interference and shadow fading cause, Utilizing complementary Cumulative Distribution Function (CCDF) to calculate the spot probability of each grid, budget result is more accurately and amount of calculation is little.

When in step 105, the situational variables of selection is CNIR, interference is analyzed module 5 calculating covering field intensity and is more than or equal to The CNIR value in the region of little intermediate value field intensity, the CNIR value in other regions represents with 255, all CNIR values calculated is sent to GIS module 8.

Step 107:GIS module 8 is by the color comparator covering spot probability value or CNIR value and system definition of each grid Scope compares, and carries out each grid color value and calculates and colour, and output result covers on electronic chart and shows in terminal 11。

Step 108: in statistical module 6, selects required Statistical Area according to geographical feature (increased surface covering, geographical height) Territory, arranges statistical condition.In the embodiment of the present invention, statistical condition is set to cover minima and the maximum of spot probability value.? When situational variables is CNIR, statistical condition is set to minima and the maximum of CINR value；Statistical condition can also is that covering field The minima of intensity values and maximum.

Step 109: statistical module 6 meets the grid of statistical condition in filtering out statistical regions, and the selection result is sent To GIS module 8, and meet in statistical regions shared by the grid number of statistical condition the hundred of total grid number in calculating selected areas Proportion by subtraction.

For overcoming " cliff effect " of digital broadcasting, under fixed reception mode, when covering spot probability value more than or equal to 70% Think that this grid is capped, make covering parameter Cov of this jth grid_j=1, otherwise, Cov_j=0；

The then coverage rate of statistical regionsWherein, M is the grid number that statistical regions is total.

Step 110:GIS module 8 by meet statistical condition require grid budget result cover on electronic chart also Show in terminal, be unsatisfactory for the grid not display that statistical condition requires.Budget result refers to the covering spot probability value of grid Or CNIR value.

Step 111: optimize district needed for selecting according to geographical feature (increased surface covering, geographical height) in optimizing module 7 Territory, arranges maximum genetic algebra G_maxWith population scale P_s。

Step 112: in the case of not changing transmitter parameter, such as power, antenna etc., optimize module 7 based on heredity Algorithm calculates and coverage rate can be made to cover after reaching the emission delay optimum combination of each transmitter in the highest SFN and optimizing Lid rate.

Step 113: each optimum time delay vector launching the station 1 and coverage rate after optimizing in output SFN.

The present invention is based on the optimum combination of each transmitter additional emission time delay, the method in genetic algorithm for solving SFN Without changing the parameters such as transmitting station power, site, antenna, cost is 0, overcome transmitter number more time manually adjust transmitting The difficult point of machine additional delay, and accuracy is high.

As shown in Figure 4, the workflow of step 103 is described in detail, after performing step 102, then performs step 201.

Step 201: judge whether the most non-computation grid, if it is, perform step 202；Otherwise, step 205 is performed.

Step 202: take a uncalculated grid, it is judged that whether also have uncalculated activation to launch the station under this grid, as Fruit is, then perform step 203；Otherwise, step 201 is performed.

Step 203: take a uncalculated transmitting station, calculate this transmitting station to this grid central point link load, Receive power, field intensity, propagation time and the time of advent.

(1) link loss budget；

The present embodiment is as a example by Okmura-Hata model.

L_p(d_i)=69.55+26.16lgf_c-13.82lgh_ti+(44.9-6.55lgh_r)lgd_i-α(h_r)+C_clutter

Wherein:

L_p(d_i) the level terrain urban district that is as the criterion geographic and geomorphic conditions under propagation loss, unit is dB；

d_iLaunching the station propagation distance to grid central point for i-th, unit is km, can be according to launching station longitude and latitude With grid central point calculation of longitude & latitude；

f_cFor operating frequency, unit is MHz；

h_tiFor i-th transmitter antenna effective depth, unit is m, approximates station elevation and adds height of transmitting antenna sum；

h_rFor reception antenna height, unit is m；

α(h_r) it is reception antenna height correction factor, unit is dB；

C_clutterFor the increased surface covering correction for attenuation factor, unit is dB；

It addition, the present invention can also utilize method of least square to carry out radio waves propagation model correction by importing measured data, Thus utilize the radio waves propagation model after correction to carry out link budget, make to predict the outcome more accurate.

(2) power and field intensity budget are received；

The local average power of reception antenna outputFor:

$\stackrel{\&OverBar;}{P_{\mathrm{ri}}}=P_i-L_p\left(d_i\right)+G_r-L_r$

Wherein:Unit is dBw；L_p(d_i) it is propagation loss, unit is dB；G_rFor receiving antenna gain, unit is dBi, L_rFor reception antenna feeder loss, unit is dB, P_iFor the ERP of i-th transmitter, unit is dBw；P_iBy following Formula is calculated:

P_i=P_ti+G_ti-L_ti

Wherein, P_tiFor the transmitting power of i-th transmitter, unit is dBw, G_tiFor the antenna gain of i-th transmitter, Unit is dBi, L_tiFor the transmitting antenna feeder loss of i-th transmitter, unit is dB.

Receive equivalent received field intensity E in place_riThe computing formula of (dB (μ V/m)):

$E_{\mathrm{ri}}=\stackrel{\&OverBar;}{P_{\mathrm{ri}}}+L_r-G_r+20{\log f}_c+107.2$

(3) path travel time t_iCalculating:

t_i=d_i/c;

Wherein: d_iLaunching the station propagation distance to grid central point for i-th, unit is km, can be according to launching the station Longitude and latitude and grid central point calculation of longitude & latitude；C is the light velocity, c=3 × 10⁸m/s。

(4) calculate the time of advent of i-th transmitting station arrival grid central point:

Time of arrival (toa) ToA_iBe calculated as follows:

ToA_i=t_i+delay_i

delay_iThe emission delay of the i-th transmitting station for arranging in step 101.

Step 204: preserve power, field intensity, propagation time and the budget result of the time of advent；Then 202 execution are gone to step.

Step 205: export the maximum power value of each grid point and maximum field intensity values to GIS module 8.

As it is shown in figure 5, the workflow of step 106 is described in detail.After performing step 105, perform step 301；

Step 301: judge whether the most non-computation grid；If it is, perform step 302；Otherwise, step 315 is performed.

Step 302: according to the reference time t of the selection FFT window original position that step 105 is arranged₀Benchmark, and respectively Transmitting station arrives the performance number of grid central point and determines t the time of advent₀。

Step 303: take a uncalculated grid, it is judged that whether this grid also has the uncalculated transmitting station, if it is, Perform step 304；Otherwise, perform step 310, carry out signal syntheses.

Step 304: take a uncalculated transmitting station, according to this transmitting station state, it is judged that whether this transmitting station is SFN internal transmitter (activated state is then SFN internal transmitter, and inactive state is not the most SFN internal transmitter), If it is, perform step 305；Otherwise, step 309 is performed.

Step 305: calculate the ToA time of advent of this transmitting station signal_iRelative to t₀Delay inequality Δ τ_i: Δ τ_i=ToA_i- t₀。

Step 306: calculate weighting function w (the Δ τ of this transmitting station signal_i) value；

In SFN, arbitrarily reception place receives simultaneously and launches the station from difference, have multiple letters of different delay Number, if it is the useful signal delay inequality Δ τ that depends on that step 304 calculates_i.Can select in step 105 to represent that DTMB connects W (the Δ τ of receipts machine interference model_i) calculation expression, it is also possible to it is arranged as required to w (Δ τ_i) calculation expression, this function is referred to as Weighting function, launches, for calculating, available power and the jamming power that station signal produces at grid central point.

$w\left({\mathrm{\&Delta;\&tau;}}_i\right)=\left\{\begin{array}{c}0,{\mathrm{\&Delta;\&tau;}}_i<-T_{\mathrm{GI}}\\ 1,{-T}_{\mathrm{GI}}\&le;{\mathrm{\&Delta;\&tau;}}_i\&le;T_{\mathrm{GI}}\\ 0,{\mathrm{\&Delta;\&tau;}}_i>T_{\mathrm{GI}}\end{array}\right.$

Wherein, T_GIRepresent the protection interval of OFDM symbol, i.e. frame head respectively.

Step 307: judge that whether this transmitting station arrives the power of grid central point more than synchronizing thresholding P_syn, if it is, Perform step 308；Otherwise, then step 309 is performed.

Step 308: utilize weighting function w (the Δ τ that step 306 calculates_i) value calculates this transmitting station signal in grid The available power of heart point generation and jamming power.

Available signal power U_i=P_ri·w(Δτ_i), it is converted into dB value X_i=10 log₁₀(U_i)；

Interfering signal power I_i=P_ri·(1-w(Δτ_i)), it is converted into dB value Y_i=10 log₁₀(I_i)；

Wherein P_riThe local power produced at grid central point for i-th transmitter.

Then 303 execution are gone to step.

Step 309: regarding this transmitting station signal all interference signal.

Available signal power U_i=0, it is converted into dB value X_i=-inf；

Interfering signal power I_i=P_ri, it is converted into dB value Y_i=10 log₁₀(P_ri)。

Then 303 execution are gone to step.

Step 310: utilize K-LNM method to carry out signal syntheses, asks synthesis available power U of each grid central point and synthesis dry Disturbing the lognormal distribution parameter of power I, computational methods are as follows:

Prediction power P_riDB value i.e. W_ri=10 log₁₀(P_ri) it is the random number with place change Normal Distribution, Then its averageThe i.e. performance number of step 103 predictionStandard varianceWhereinIt is the standard variance of shadow fading, depends on the seriousness of shadow attenuation, ground digital television broadcast takes 5.5dB.

W (Δ τ in grid_i) vary less, can ignore, so U_iAnd I_iObey logarithm normal distribution, then X_i= 10·log₁₀(U_i) average, standard variance be:

$m_{X_i}=m_{W_{\mathrm{ri}}}+10\&CenterDot;{\log }_{10}\left(w\left({\mathrm{\&Delta;\&tau;}}_i\right)\right)=10\&CenterDot;{\log }_{10}\left(\stackrel{\&OverBar;}{P_{\mathrm{ri}}}\right)+10\&CenterDot;{\log }_{10}\left(w\left({\mathrm{\&Delta;\&tau;}}_i\right)\right),$ ${\mathrm{\&sigma;}}_{X_i}={\mathrm{\&sigma;}}_{W_{\mathrm{ri}}}=5.5\mathrm{dB}$

Y_i=10 log₁₀(I_i) average, standard variance be

$m_{Y_i}=m_{W_{\mathrm{ri}}}+10\&CenterDot;{\log }_{10}(1-w\left({\mathrm{\&Delta;\&tau;}}_i\right))=10\&CenterDot;{\log }_{10}\left(\stackrel{\&OverBar;}{P_{\mathrm{ri}}}\right)+10\&CenterDot;{\log }_{10}(1-w\left({\mathrm{\&Delta;\&tau;}}_i\right)),$ ${\mathrm{\&sigma;}}_{Y_i}={\mathrm{\&sigma;}}_{W_{\mathrm{ri}}}=5.5\mathrm{dB}$

Synthesis available power U of grid central point: $U=\underset{i=1}{\overset{N_1}{\mathrm{\&Sigma;}}}{U}_i=\underset{i=1}{\overset{N_1}{\mathrm{\&Sigma;}}}{P}_{\mathrm{ri}}w\left(\mathrm{\&Delta;}{\mathrm{\&tau;}}_i\right);$

The cumulative interference power I of grid central point: $I=\underset{i=1}{\overset{N_1}{\mathrm{\&Sigma;}}}{I}_i+\underset{j=1}{\overset{N_2}{\mathrm{\&Sigma;}}}{I}_j+\underset{k=1}{\overset{N^{\&prime;\&prime;}}{\mathrm{\&Sigma;}}}{I}_k=\underset{i=1}{\overset{N_1}{\mathrm{\&Sigma;}}}{P}_{\mathrm{ri}}[1-w\left({\mathrm{\&Delta;\&tau;}}_i\right)]+\underset{j=1}{\overset{N_2}{\mathrm{\&Sigma;}}}{P}_{\mathrm{rj}}+\underset{k=1}{\overset{N^{\&prime;\&prime;}}{\mathrm{\&Sigma;}}}{P}_{\mathrm{rk}}.$

Wherein, N' is SFN internal transmitter total number, N₁For receiving power in SFN more than P_synThe transmitting station Number, N₂For receiving power in SFN less than P_synTransmitting number of stations, N'=N₁+N₂, N'' is that SFN is outside with taking place frequently Penetrate machine number.

N independent logarithm normal distribution variable sum, can be considered approximate log normally distributed variable, and its distributed constant can be by K-LNM method calculates, and makes X=10log₁₀(U), Y=10log₁₀(I), then the calculating process of X, Y average and variance is as follows:

(1) will Value substitute into X respectively_dB, utilize following formula to be converted into napier's value.

X_Neper=1/10log₁₀(e)·X_dB≈0.23X_dB

(2) napier is utilized to be worth average m calculating X_XAnd variance

$m_X=\ln \left[\underset{i=1}{\overset{N_1}{\mathrm{\&Sigma;}}}\exp (m_{X_i}+\frac{{{\mathrm{\&sigma;}}_{X_i}}^2}{2})\right]-\frac{{{\mathrm{\&sigma;}}_X}^2}{2};$

${{\mathrm{\&sigma;}}_X}^2=\ln [1+k\frac{\underset{i=1}{\overset{N_1}{\mathrm{\&Sigma;}}}\{\exp \left({{\mathrm{\&sigma;}}_{X_i}}^2\right)-1\}\exp (2m_{X_i}+{{\mathrm{\&sigma;}}_{X_i}}^2)}{{\left(\underset{i=1}{\overset{N_1}{\mathrm{\&Sigma;}}}\exp (m_{X_i}+\frac{{{\mathrm{\&sigma;}}_{X_i}}^2}{2})\right)}^2}]$

(3) average m of Y can in like manner be obtained_YAnd variances sigma_Y ²:

$m_Y=\ln \left[\underset{i=1}{\overset{N_1+N_2+N^{\&prime;\&prime;}}{\mathrm{\&Sigma;}}}\exp (m_{Y_i}+\frac{{{\mathrm{\&sigma;}}_{Y_i}}^2}{2})\right]-\frac{{{\mathrm{\&sigma;}}_Y}^2}{2};$

${{\mathrm{\&sigma;}}_Y}^2=ln[1+k\frac{\underset{i=1}{\overset{N_1+N_2+N^{\&prime;\&prime;}}{\mathrm{\&Sigma;}}}\{\exp \left({{\mathrm{\&sigma;}}_{Y_i}}^2\right)-1\}\exp ({2m}_{Y_i}+{{\mathrm{\&sigma;}}_{Y_i}}^2)}{{\left(\underset{i=1}{\overset{N_1+N_2+N^{\&prime;\&prime;}}{\mathrm{\&Sigma;}}}\exp (m_{Y_i}+\frac{{{\mathrm{\&sigma;}}_{Y_i}}^2}{2})\right)}^2}]$

(4) m that will obtain_X、σ_X、m_Y、σ_YNapier's value substitute into X respectively_Neper, it is converted into dB value according to following formula.

X_dB=10log₁₀(e)X_Neper

K is correction factor, depends on the standard variance receiving signal, for standard variance less than 6dB, generally takes 0.7.

Step 311: discriminatory analysis variable is still CNIR for covering spot probability, if covering spot probability, then holds Row step 312；Otherwise, step 314 is performed.

Step 312: utilize the relevant moments method of coupling to calculate synthesis available power and the correlation coefficient of cumulative interference power, calculate Method is as follows:

As Δ τ_iWithin the specific limits, U_iAnd I_iIt is not the most 0, normal variate X_iAnd Y_iBoth have the biggest dependency, cover Special Caro simulation result shows X_iAnd Y_iCorrelation coefficientEven if therefore under the influence of ignoring shadow fading, X and Y it Between there is dependency.For calculating the correlation coefficient r of X and Y_XY, make X_i'=ln (U_i), then X_i'=λ X_i, Wherein parameterIn like manner, Y is made_i'=ln (I_i), X'=ln (U), Y'=ln (I), Then Y_i' and Y_i, X' Yu X, Y' Yu Y be respectively provided with X_i' and X_iBetween identical relation, and haver_X′Y′=r_XY。

Make φ=X'+Y', the most then average m of φ_φ=m_X'+m_Y', variances sigma_φ ²=σ_X' ²+2r_X'Y'σ_X'σ_Y'+σ_Y' ².Coupling U with The Correlation Moment of I:

$E\left[\mathrm{UI}\right]=E\left[e^{X^{\&prime;}}{e}^{Y^{\&prime;}}\right]=E\left[e^{X^{\&prime;}+Y^{\&prime;}}\right]=E\left[e^{\mathrm{\&phi;}}\right]=e^{m_{\mathrm{\&phi;}}+{{\mathrm{\&sigma;}}_{\mathrm{\&phi;}}}^2/2}$

$E\left[\mathrm{UI}\right]=E\left[e^{X^{\&prime;}}{e}^{Y^{\&prime;}}\right]=E\left[e^{X^{\&prime;}+Y^{\&prime;}}\right]$

$=E\left[(e^{{X_1}^{\&prime;}}+e^{{X_2}^{\&prime;}}+...+e^{{X_{N_1}}^{\&prime;}})(e^{{Y_1}^{\&prime;}}+e^{{Y_2}^{\&prime;}}+...+e^{{Y_{N_1+N_2+N^{\&prime;\&prime;}}}^{\&prime;}})\right]$

$=\underset{i=1}{\overset{N_1}{\mathrm{\&Sigma;}}}\underset{i=1}{\overset{N_1+N_2+N^{\&prime;\&prime;}}{\mathrm{\&Sigma;}}}E\left[\left(e^{{X_i}^{\&prime;}{+Y_i}^{\&prime;}}\right)\right]$

$=\underset{i=1}{\overset{N_1}{\mathrm{\&Sigma;}}}\underset{i=1}{\overset{N_1+N_2+N^{\&prime;\&prime;}}{\mathrm{\&Sigma;}}}{e}^{m_{{X_i}^{\&prime;}}+m_{{Y_i}^{\&prime;}}+1/2\&CenterDot;({\mathrm{\&sigma;}}_{X_i^{\&prime;}}^2+{\mathrm{\&sigma;}}_{Y_i^{\&prime;}}^2+{2r}_{X_i^{\&prime;}{Y}_i^{\&prime;}}{\mathrm{\&sigma;}}_{X_i^{\&prime;}}{\mathrm{\&sigma;}}_{Y_i^{\&prime;}})}$

$=v$

Thus can obtain, $r_{\mathrm{XY}}=r_{X^{\&prime;}{Y}^{\&prime;}}=\frac{2[\ln \left(v\right)-(m_{X^{\&prime;}}+m_{Y^{\&prime;}})]-({{\mathrm{\&sigma;}}_{X^{\&prime;}}}^2+{{\mathrm{\&sigma;}}_{Y^{\&prime;}}}^2)}{{2\mathrm{\&sigma;}}_{X^{\&prime;}}{\mathrm{\&sigma;}}_{Y^{\&prime;}}}.$ WithIt is respectively Y_i' average and Standard variance.

Be can be seen that by above procedure m_X'、m_Y'It is exactly m_X、m_YHow Training value, therefore the method can utilize the result of calculation of step 310, thus reduces amount of calculation.

Step 313: calculating and cover spot probability, computational methods are as follows:

The spot probability that covers of grid is defined as carrying dry radio-frequency protection ratio PR being more than under this mode of operation than CNIR The probability of (Protection Ratio).

Carry to be dried and represent with γ than (CINR), be defined as:

Wherein, U is available signal power sum, and I is interfering signal power sum, N₀For receiver background noise.

The covering spot probability of this grid is:

P_c(x_i,y_i)=P{ γ ＞ PR}=P{ γ_dB＞ PR_dB}

Wherein subscript dB represents the dB value taking γ Yu PR, (x_i,y_i) represent i-th grid center point coordinate.

Can be obtained by step 310,Obey independent logarithm normal distribution, make ψ=X-Y, then distributed constant average m_ψ=m_X-m_Y, Variances sigma_ψ ²=σ_X ²+σ_Y ²-2r_XYσ_Xσ_Y.If N₀, I separate, by P_c(x_i,y_i) it is abbreviated as P_c, then have:

$P_c=P\{{\left(\frac{U}{I}\right)}_{\mathrm{dB}}>{\mathrm{PR}}_{\mathrm{dB}}\}\&times;P\{{\left(\frac{U}{N_0}\right)}_{\mathrm{dB}}>{\mathrm{PR}}_{\mathrm{dB}}\}$

For synthesizing the useful field intensity region less than minimum intermediate value field intensity, order covers spot probability P_c=0；Synthesis is had Use is by force more than or equal to the region of minimum intermediate value field intensity, if cumulative interference field intensity exists, then

$P_c=Q\left(\frac{{\mathrm{PR}}_{\mathrm{dB}}-(m_X-m_Y)}{\sqrt{{{\mathrm{\&sigma;}}_X}^2+{{\mathrm{\&sigma;}}_Y}^2-2r_{\mathrm{XY}}{\mathrm{\&sigma;}}_X{\mathrm{\&sigma;}}_Y}}\right)\&times;Q\left(\frac{{\mathrm{PR}}_{\mathrm{dB}}-(m_X-N_0)}{\sqrt{{{\mathrm{\&sigma;}}_X}^2}}\right)$

Otherwise $P_c=Q\left(\frac{{(C/N)}_{\mathrm{dB}}-(m_X-N_0)}{\sqrt{{{\mathrm{\&sigma;}}_X}^2}}\right).$

Wherein function $Q\left(y\right)=\frac{1}{\sqrt{2\mathrm{\&pi;}}}{\&Integral;}_y^{\&infin;}{e}^{-\frac{x^2}{2}}\mathrm{dx}.$

The present invention utilizes Cumulative Distribution Function to calculate the spot probability of this grid, and amount of calculation is little.

After having calculated the covering spot probability of this grid, go to step 301 execution.

Step 314: calculating CNIR, computational methods are as follows:

For synthesizing the useful field intensity region less than minimum intermediate value field intensity, make CNIR=255；It is useful powerful for synthesis In the region equal to minimum intermediate value field intensity, if cumulative interference field intensity exists, then CNIR=m_X-10log10(10^m_Y/10+10^ N₀/ 10), otherwise CNIR=m_X-N₀。

After having calculated the CNIR of this grid, go to step 301 execution.

Step 315: export the covering spot probability of each grid or CNIR to GIS module 8.

As shown in Figure 6, the workflow of step 112 is described in detail, after performing step 111, starts to perform step 401；

Step 401: initialize genetic algorithm parameter.

Initialization genetic algorithm parameter includes: population scale size, maximum genetic algebra, the transmitter that is active Number N_opt, coding figure place l, generation gap GGAP, mutation probability P_mWith crossover probability P_cr.Wherein population scale size, maximum heredity generation Number, transmitter state arrange in step 111, optimize module 7 can obtain from station data base 3 be active Penetrate machine number N'.

Population scale size P_s, mutation probability P_mWith crossover probability P_crIt it is the control ginseng affecting effect of optimization in genetic algorithm Number, P_mBeing worth the biggest, result is the poorest, because P_mIt is worth the biggest, produces new individual probability the biggest, more similar in appearance to random search algorithm, institute With P_mValue is the least, and desirable 0.01.Because P_crBeing worth the least, the probability of the individual survival of high fitness is the biggest, is equivalent to solution space Reducing, result is deteriorated.By the optimum results under a large amount of different scenes is contrasted, it was therefore concluded that: during scene change, need According to transmitter number, P is set_s, the most desirable P_m=0.01、P_cr>=0.7 ensures that optimum results is no longer deteriorated.

The present embodiment encodes figure place l=20, generation gap GGAP=0.9, P_m=0.01, P_cr=0.7。

Step 402: select the decision variable of genetic algorithm.

The embodiment of the present invention selects the N that SFN is internal and is active_optIndividual transmitter emission delay is as certainly Plan variable, N_optIndividual decision variable one N of composition_optDimension line delay vector delay:

Step 403: determine span and the constraints of decision variable.

Because relative time delay is more than T_GITime, signal, the therefore span of i-th transmitter emission delay are disturbed in generation For:

0≤delay_i≤T_GI, and delay_i=n × 0.1 μ s, 1≤i≤N_opt, n is positive integer.

Step 404: coding.

The present invention uses Gray code to delay_iEncode.

Step 405: generate initial population.

Stochastic generation P_sIndividual a length of l × N_optTime delay vector as initial population.

Step 406: calculating target function value.

Object function is set in the area of coverage not cover grid percentage ratio.

$f\left(\mathrm{delay}\right)=(1-\underset{j=1}{\overset{M^{\&prime;}}{\mathrm{\&Sigma;}}}{\mathrm{Cov}}_j/M^{\&prime;})\&times;100\%$

M' is grid number total in optimizing region.

Time delay vector value each in each generation population is sent to station data base 3, required optimization region is sent to GIS Module 8, calls radio wave propagation budget module 4 and again predicts that each transmitting station is to when optimizing the arrival of each grid central point in region Between, interference is analyzed module 5 and is selected state to be to activate with nonactivated transmitter as non-desire receiver/transmitter, in calculation optimization region Covering the field intensity covering spot probability more than or equal to each grid of minimum intermediate value field intensity, statistical module 6 calculates coverage rate, thus counts Calculate the target function value of each time delay vector.

Step 407: distribution fitness function value.

Using fitness assignment algorithm based on linear ordering to distribute fitness value, selecting pressure reduction is 2, and target function value is more Little, fitness value is the biggest.Using time delay vector minimum for the contemporary target function value calculated as optimum time delay vector, and preserve Coverage rate corresponding to excellent time delay vector.

Step 408: judge whether to meet end condition, if it is, perform step 113, the optimum delay variable of output with And coverage rate after optimizing；If it is not, then perform step 409.End condition refers to: reached maximum genetic algebra G_maxOr When excellent time delay vector is during evolution without changing.

Step 409: select.The present invention uses random ergodic sampling operator, generation gap GGAP=0.9.

Step 410: intersecting, the present invention uses single-point crossover operator.

Step 411: variation, the present invention uses Discrete mutation operator.

Afterwards, heavily insertion based on relevance grade is used to guarantee P_sThe individual time delay vector adapted to most of × (1-GGAP) is always connected Resume and be multicast to the next generation.After execution of step 411, go to step 406 execution.

Repeat step 406-step 411 until meeting end condition.

In sum, according to the present invention obtain with optimize DTMB SFN coverage rate apparatus and method, this device and Method definition synchronizes thresholding P_synTo embody SFN gain, power is at P_synAbove signal could be synchronized by receiver, according to Actual DTMB receiver interference model calculates each station at the available signal power of grid central point and interfering signal power, uses K-LNM method carries out signal syntheses, utilizes the relevant moments method of coupling to calculate synthesis available power and the correlation coefficient of cumulative interference power, Utilizing the spot probability of CCDF function computation grid, budget result is actual, and amount of calculation is little, it is possible to increase SFN coverage rate is pre- The correctness surveyed and accuracy rate, be simultaneously based on genetic algorithm calculate each transmitter additional emission time delay so that self-interference region Littleization, this optimization method is without changing the parameter such as transmitter power, antenna, and cost is 0, overcome transmitter number more time people Work adjusts the difficult point that transmitter emission delay is complicated and realization is difficult, and accuracy is high.Apparatus of the present invention and method can be applied In China's DTMB SFN is built.

Claims (6)

1. the device obtained and optimize DTMB SFN coverage rate, it is characterised in that including: station message processing module, Module, statistical module, optimization module, GIS module, geography information number are analyzed in station data base, radio wave propagation budget module, interference According to storehouse and covering result database；

Station message processing module, in collecting SFN, each launches the information of the station, and the information launching the station includes: stand Location title, longitude and latitude, transmitting power, antenna radiation pattern, signal modulation system, tranmitting frequency, transmitting time delay, state, network are compiled Number, and the classification of each transmitting station is stored in station data base；Geographic information database is used for storing electronic chart, GIS mould Block loads electronic chart, and generates grid screen figure layer according to the computational accuracy arranged, and stores the result into geographic information database In；

Radio wave propagation budget module obtains each information launching the station of state of activation from station data base, from GIS module Obtain the geography information of grid central point of selection area, calculate state of activation each and launch the station in grid in the range of selected The link load of heart point, reception power, field intensity value, propagation time and the time of advent, and send result to GIS module, do Disturb analysis module and statistical module；GIS module according to the maximum field intensity values computation grid color value of each grid and colours one by one, will Output result covers on electronic chart；

The function of module realization is analyzed in interference: first, arrange reference time t₀Selection standard, frame head and frame length, reception Machine background noise N₀, synchronize threshold CNR Margn, synchronize thresholding P_syn, minimum intermediate value field intensity E_medWith radio-frequency protection ratio PR, choosing Select or arrange weighting function w (Δ τ_i), select non-receiver to be received and situational variables, situational variables be load make an uproar dry than CNIR value or Cover spot probability；Secondly, reference time t is determined₀, when determining the arrival of each transmitter each grid central point in the range of selected Between relative to reference time t₀Delay inequality, and obtain weighting function w (the Δ τ of each grid available power and jamming power_i) Value, Δ τ_iRepresent the ToA time of advent of i-th transmitter_iWith t₀Delay inequality；Then, arrange signal level synchronize thresholding with On signal synchronized by receiver, according to weighting function w (Δ τ_i) calculate each transmitter available power at each grid central point And jamming power, use K factor lognormal method, obtain synthesis available power U and the cumulative interference power I of each grid central point The average of dB value and variance；Finally, if situational variables is CNIR value, calculates and cover field intensity more than or equal to minimum intermediate value field intensity The CNIR value in region, the CNIR value in other regions represents with 255, and the CNIR value of calculating is sent to GIS module and statistical module； If situational variables is for covering spot probability, calculates and cover field intensity more than or equal to the synthesis of each grid in the region of minimum intermediate value field intensity Available power and the correlation coefficient of cumulative interference power and cover spot probability, in other regions, the covering spot probability of grid is 0, the spot probability of grid is sent to GIS module and statistical module；Wherein, reference time t₀There are two kinds of selection standards: first Kind it is to arrive the earliest and power is more than synchronizing thresholding P_synSignal the time of advent on the basis of, the second is the strongest with level On the basis of the time of advent of signal；Synchronize thresholding P_syn=N₀+ Margn, N₀For receiver background noise, Margn is for synchronizing thresholding Carrier-to-noise ratio；The color comparator scope that GIS module defines with system according to covering spot probability value or the CNIR value of each grid compares, Carrying out each grid color value calculate and colour one by one, output result covers on electronic chart；

Described weighting function w (Δ τ_i) it is: $w\left({\mathrm{\&Delta;\&tau;}}_i\right)=\left\{\begin{array}{c}0,\mathrm{\&Delta;}{\mathrm{\&tau;}}_i<-T_{GI}\\ 1,-T_{GI}\&le;{\mathrm{\&Delta;\&tau;}}_i\&le;T_{GI}\\ 0,\mathrm{\&Delta;}{\mathrm{\&tau;}}_i>T_{GI}\end{array}\right.,$ Wherein, T_GIRepresent frame head length；

Synthesis available power U and the cumulative interference power I of the grid central point that described interference analysis module obtains are respectively as follows:

$U=\underset{i=1}{\overset{N_1}{\mathrm{\&Sigma;}}}{U}_i=\underset{i=1}{\overset{N_1}{\mathrm{\&Sigma;}}}{P}_{ri}w\left({\mathrm{\&Delta;\&tau;}}_i\right),I=\underset{i=1}{\overset{N_1}{\mathrm{\&Sigma;}}}{I}_i+\underset{j=1}{\overset{N_2}{\mathrm{\&Sigma;}}}{I}_j+\underset{k=1}{\overset{N^{\&prime;\&prime;}}{\mathrm{\&Sigma;}}}{I}_k=\underset{i=1}{\overset{N_1}{\mathrm{\&Sigma;}}}{P}_{ri}\&lsqb;1-w\left({\mathrm{\&Delta;\&tau;}}_i\right)\&rsqb;+\underset{j=1}{\overset{N_2}{\mathrm{\&Sigma;}}}{P}_{rj}+\underset{k=1}{\overset{N^{\&prime;\&prime;}}{\mathrm{\&Sigma;}}}{P}_{rk};$

Wherein, N₁For receiving power in SFN more than P_synTransmitting number of stations, N₂For receiving power in SFN less than P_syn Transmitting number of stations, N " be the outside co-channel transmitter number of SFN, U_iHaving of place generation is being received for i-th transmitter With signal power, P_riThe local power received for i-th transmitter, I_iThe interference that place produces is being received for i-th transmitter Signal power；

P_riStatistical property obeys logarithm normal distribution, and the dB value changed is W_ri=10 log₁₀(P_ri)；

W_riAverageStandard varianceWherein,It is the standard variance of shadow fading, Ground digital television broadcast takes 5.5dB；Represent the local average power that i-th transmitter receives；

U_iDB value X converted_i=10 log₁₀(U_i) averageAnd standard varianceIt is respectively as follows:

$m_{X_i}=m_{W_{ri}}+10\&CenterDot;{\log }_{10}\left(w\right({\mathrm{\&Delta;\&tau;}}_i\left)\right)=10\&CenterDot;{\log }_{10}\left(\stackrel{\&OverBar;}{P_{ri}}\right)+10\&CenterDot;{\log }_{10}\left(w\right({\mathrm{\&Delta;\&tau;}}_i\left)\right),{\mathrm{\&sigma;}}_{X_i}={\mathrm{\&sigma;}}_{W_{ri}}=5.5dB;$

I_iDB value Y converted_i=10 log₁₀(I_i) averageAnd standard varianceIt is respectively as follows:

$m_{Y_i}=m_{W_{ri}}+10\&CenterDot;{\log }_{10}(1-w({\mathrm{\&Delta;\&tau;}}_i\left)\right)=10\&CenterDot;{\log }_{10}\left(\stackrel{\&OverBar;}{P_{ri}}\right)+10\&CenterDot;{\log }_{10}(1-w({\mathrm{\&Delta;\&tau;}}_i\left)\right),{\mathrm{\&sigma;}}_{Y_i}={\mathrm{\&sigma;}}_{W_{ri}}=5.5dB;$

DB value X and Y that synthesis available power U and cumulative interference power I convert are respectively as follows:

X=10log₁₀(U), Y=10log₁₀(I)；

Wherein, interference is analyzed module and is used K factor lognormal method, obtains synthesis available power U and the conjunction of each grid central point The average of dB value of jamming power I and the implementation method of variance is become to be:

(1) willValue substitute into X respectively_dB, utilize following formula to be converted into napier's value；

X_Neper=1/10log₁₀(e)·X_dB≈0.23X_dB；

(2) napier is utilized to be worth average m calculating X_XAnd variance

$m_X=ln\&lsqb;\underset{i=1}{\overset{N_1}{\&Sigma;}}\exp (m_{X_i}+\frac{{{\mathrm{\&sigma;}}_{X_i}}^2}{2})\&rsqb;-\frac{{{\mathrm{\&sigma;}}_X}^2}{2};$

${{\mathrm{\&sigma;}}_X}^2=ln\&lsqb;1+k\frac{\underset{i=1}{\overset{N_1}{\mathrm{\&Sigma;}}}\{\exp \left({{\mathrm{\&sigma;}}_{X_i}}^2\right)-1\}\exp (2m_{X_i}+{{\mathrm{\&sigma;}}_{X_i}}^2)}{{\left(\underset{i=1}{\overset{N_1}{\mathrm{\&Sigma;}}}\exp \right(m_{X_i}+\frac{{{\mathrm{\&sigma;}}_{X_i}}^2}{2}\left)\right)}^2}\&rsqb;;$

K is correction factor；

(3) average m of Y it is calculated_YAnd variances sigma_Y ²:

$m_Y=ln\&lsqb;\underset{i=1}{\overset{N_1+N_2+N^{\&prime;\&prime;}}{\&Sigma;}}\exp (m_{Y_i}+\frac{{{\mathrm{\&sigma;}}_{Y_i}}^2}{2})\&rsqb;-\frac{{{\mathrm{\&sigma;}}_Y}^2}{2};$

${{\mathrm{\&sigma;}}_Y}^2=ln\&lsqb;1+k\frac{\underset{i=1}{\overset{N_1+N_2+N^{\&prime;\&prime;}}{\mathrm{\&Sigma;}}}\{\exp \left({{\mathrm{\&sigma;}}_{Y_i}}^2\right)-1\}\exp (2m_{Y_i}+{{\mathrm{\&sigma;}}_{Y_i}}^2)}{{\left(\underset{i=1}{\overset{N_1+N_2+N^{\&prime;\&prime;}}{\mathrm{\&Sigma;}}}\exp \right(m_{Y_i}+\frac{{{\mathrm{\&sigma;}}_{Y_i}}^2}{2}\left)\right)}^2}\&rsqb;;$

(4) m that will obtain_X、σ_X、m_Y、σ_YNapier's value substitute into X respectively_Neper, it is converted into dB value according to following formula；

X_dB=10log₁₀(e)X_Neper；

Described interference is analyzed module and is utilized the relevant moments method of coupling to calculate synthesis available power and the cumulative interference power of each grid Correlation coefficient, specifically:

First, X is made_i'=ln (U_i), then X_i'=λ X_i, parameterThen X_i' averageWith standard side DifferenceIt is respectively as follows:In like manner, Y is made_i'=ln (I_i), X'=ln (U), Y'=ln (I), then Y_i' With Y_i, X' Yu X, Y' Yu Y be respectively provided with X_i' and X_iBetween identical relation, and X_i' and Y_i' correlation coefficientEqual to X_iAnd Y_i's Correlation coefficientThe correlation coefficient r of X' Yu Y'_X'Y'Correlation coefficient r equal to X and Y_XY；

Make variable φ=X'+Y', then average m of φ_φ=m_X'+m_Y', variances sigma_φ ²=σ_X' ²+2r_X'Y'σ_X'σ_Y'+σ_Y' ²；Coupling U's Yu I Correlation Moment $E\&lsqb;UI\&rsqb;=E\&lsqb;e^{X^{\&prime;}}{e}^{Y^{\&prime;}}\&rsqb;=E\&lsqb;e^{X^{\&prime;}+Y^{\&prime;}}\&rsqb;=E\&lsqb;e^{\mathrm{\&phi;}}\&rsqb;=e^{{}_{m_{\mathrm{\&phi;}}+{{\mathrm{\&sigma;}}_{\mathrm{\&phi;}}}^2\mathrm{/2}}},$ Correlation Moment E [UI] can also be expressed as:

$\begin{array}{c}E\&lsqb;UI\&rsqb;=E\&lsqb;e^{X^{\&prime;}}{e}^{Y^{\&prime;}}\&rsqb;=E\&lsqb;e^{X^{\&prime;}+Y^{\&prime;}}\&rsqb;\\ =E\&lsqb;(e^{{X_1}^{\&prime;}}+e^{{X_2}^{\&prime;}}+...+e^{{X_{N_1}}^{\&prime;}})(e^{{Y_1}^{\&prime;}}+e^{{Y_2}^{\&prime;}}+...+e^{{Y_{N_1+N_2+N^{\&prime;\&prime;}}}^{\&prime;}})\&rsqb;\\ =\underset{i=1}{\overset{N_1}{\mathrm{\&Sigma;}}}\underset{i=1}{\overset{N_1+N_2+N^{\&prime;\&prime;}}{\mathrm{\&Sigma;}}}E\&lsqb;\left(e^{{X_i}^{\&prime;}+{Y_i}^{\&prime;}}\right)\&rsqb;\\ =\underset{i=1}{\overset{N_1}{\mathrm{\&Sigma;}}}\underset{i=1}{\overset{N_1+N_2+N^{\&prime;\&prime;}}{\mathrm{\&Sigma;}}}{e}^{m_{{X_i}^{\&prime;}}+m_{{Y_i}^{\&prime;}}+1/2\&CenterDot;({\mathrm{\&sigma;}}_{X_i^{\&prime;}}^2+{\mathrm{\&sigma;}}_{Y_i^{\&prime;}}^2+2r_{X_i^{\&prime;}{Y}_i^{\&prime;}}{\mathrm{\&sigma;}}_{X_i^{\&prime;}}{\mathrm{\&sigma;}}_{Y_i^{\&prime;}})}\\ =v\end{array}$

Thus obtain: $r_{XY}=r_{X^{\&prime;}{Y}^{\&prime;}}=\frac{2\&lsqb;\ln \left(v\right)-(m_{X^{\&prime;}}+m_{Y^{\&prime;}})\&rsqb;-({{\mathrm{\&sigma;}}_{X^{\&prime;}}}^2+{{\mathrm{\&sigma;}}_{Y^{\&prime;}}}^2)}{2{\mathrm{\&sigma;}}_{X^{\&prime;}}{\mathrm{\&sigma;}}_{Y^{\&prime;}}},$ WithIt is respectively Y_i' average and standard Variance；

Module is analyzed in described interference, utilizes complimentary cumulative function CCDF to calculate the covering spot probability of each grid, specifically:

The covering spot probability P of i-th grid_c(x_i,y_i) it is: P_c(x_i,y_i)=P{ γ ＞ PR}=P{ γ_dB＞ PR_dB}；(x_i, y_i) represent i-th grid central point coordinate；γ is CINR value,γ_dBAnd PR_dBRepresent the dB value taking γ Yu PR；

Make ψ=X-Y, then average m of ψ_ψ=m_X-m_Y, variances sigma_ψ ²=σ_X ²+σ_Y ²-2r_XYσ_Xσ_Y；If N₀, I separate, thenm_XAnd σ_XFor average and the standard variance of X, m_YAnd σ_YFor Y average and Standard variance；

For synthesizing the useful field intensity grid less than minimum intermediate value field intensity, the covering spot probability P of this grid_c=0；

Synthesis is useful by force more than or equal to the region of minimum intermediate value field intensity, if cumulative interference field intensity exists, then this grid Covering spot probability:

$P_c=Q\left(\frac{{\mathrm{PR}}_{dB}-(m_X-m_Y)}{\sqrt{{{\mathrm{\&sigma;}}_X}^2+{{\mathrm{\&sigma;}}_Y}^2-2r_{XY}{\mathrm{\&sigma;}}_X{\mathrm{\&sigma;}}_Y}}\right)\&times;Q\left(\frac{{\mathrm{PR}}_{dB}-(m_X-N_0)}{\sqrt{{{\mathrm{\&sigma;}}_X}^2}}\right)$

Otherwise, $P_c=Q\left(\frac{{(C/N)}_{dB}-(m_X-N_0)}{\sqrt{{{\mathrm{\&sigma;}}_X}^2}}\right);$

Wherein, function(C/N)_dBDB value for carrier-to-noise ratio thresholding；

Statistical module, according to the statistical condition arranged, filters out the grid meeting statistical condition in statistical regions, and will screening Result is sent to GIS module, the percentage ratio of total grid number in calculating statistical regions shared by the grid meeting statistical condition, concurrently Give optimization module；The budget result meeting the grid that statistical condition requires is covered on electronic chart by GIS module, is unsatisfactory for The not display of the budget result of the grid that statistical condition requires；

Optimize module to calculate based on genetic algorithm optimization areal coverage can be made to reach each transmitter in the highest SFN Emission delay optimum combination and coverage rate；In every generation evolutionary process of genetic algorithm, by each time delay in present age population to The value of amount is sent to station data base, respectively launches the transmitting time delay of the station at station database update, obtains from GIS module The geography information of each grid central point in optimization region, recalculates each transmitting of state of activation by radio wave propagation budget module The station, to the time of advent of each grid central point in optimization region, is analyzed module by interference and is selected state for activation with inactive Transmitter as non-desire receiver/transmitter, and the covering spot probability of each grid or CNIR value in calculation optimization region, by system Meter module is the percentage ratio of total grid number in calculating favored area shared by the grid meeting statistical condition；

Optimize module to calculate based on genetic algorithm optimization areal coverage can be made to reach each transmitter in the highest SFN Emission delay optimum combination and coverage rate, implementing step is:

Step 401: initialize genetic algorithm parameter, including: population scale P_s, maximum genetic algebra G_max, be in SFN sharp The transmitter number N of the state of living_opt, coding figure place l, generation gap GGAP, mutation probability P_mWith crossover probability P_cr；

Wherein, l=20, GGAP=0.9, P are set_m=0.01, P_cr≥0.7；

Step 402: select the N that SFN is internal and is active_optIndividual transmitter emission delay, as genetic algorithm Decision variable, N_optIndividual decision variable composition time delay vector delay:

Step 403: determine span and the constraints of decision variable；

I-th transmitter emission delay delay is set_iSpan be: 0≤delay_i≤T_GI, and delay_i=n × 0.1 μ S, 1≤i≤N_opt, n is positive integer；

Step 404: use Gray code to delay_iEncode；

Step 405: stochastic generation P_sIndividual a length of l × N_optTime delay vector as initial population；

Step 406: the value of each time delay vector in contemporary population is sent to each transmitting station, calculates each transmitting station to excellent The time of advent of each grid central point in change region, selection state is penetrated as non-transmitting-receiving with nonactivated transmitter for activating Machine, covers the field intensity covering spot probability more than or equal to the grid of minimum intermediate value field intensity, finally calculates mesh in calculation optimization region Offer of tender numerical value f (delay):

$f\left(delay\right)=(1-\underset{j=1}{\overset{M^{\&prime;}}{\&Sigma;}}{\mathrm{Cov}}_j/M^{\&prime;})\&times;100\%$

Wherein, M' is grid number total in optimizing region；

Step 407: use fitness assignment algorithm based on linear ordering to distribute fitness value, selecting pressure reduction is 2, object function Being worth the least, fitness value is the biggest；Using time delay vector minimum for target function value in the present age as optimum time delay vector, and preserve optimum Coverage rate corresponding to time delay vector；

Step 408: judge whether to meet end condition, if it is, the coverage rate that the optimum time delay of output is vectorial and corresponding, optimum Time delay vector i.e. emission delay optimum combination；If it does not, continue executing with step 409；End condition refers to: reached maximum When genetic algebra or optimum time delay vector are during evolution without changing；

Step 409: use random ergodic sampling operator；

Step 410: use single-point crossover operator；

Step 411: use Discrete mutation operator；Afterwards, heavily insertion based on relevance grade is used to guarantee P_s× (1-GGAP) is individual the suitableeest The time delay vector answered always is traveled to the next generation continuously；Then 406 execution are gone to step；

Cover result database for storing radio wave propagation budget module, interference analysis module and the result of calculation of statistical module.

2. the method obtained and optimize DTMB SFN coverage rate, it is characterised in that comprise the following steps:

Step A: load planning region electronic chart and arrange computational accuracy, GIS module generates grid according to the computational accuracy arranged Net figure layer, transfers the geography information of each grid central point in grid screen from geographic information database, geography information include longitude, Latitude, height and increased surface covering；

Step B: add SFN and respectively launch the station and configure the parameter of each transmitting station, parameter includes launching power, launching frequency Rate, transmitter antenna gain (dBi), transmitting antenna feeder loss, launch antenna radiation pattern, signal modulation system, emission delay, state and Network numbering；

Step C: arrange link loss budget parameter, including arranging propagation model, mulching material decay and link loss budget Distance, and reception antenna height, gain and minimum intermediate value field intensity are set；

Step D: calculate each transmitting station link load to each grid central point of state of activation, receive power, field intensity value, biography Between sowing time and the time of advent；

Step E: according to the maximum field intensity values computation grid color value of each grid and colour, output result covers electronically On figure and at terminal demonstration；

Step F: transmitter each in SFN is set to state of activation, the outside co-channel transmitter of SFN is set to inactive shape State；The reference time t of fast Fourier change FFT window original position is set₀Selection standard, frame head and frame length, connect Receipts machine background noise N₀, synchronize threshold CNR Margn, synchronize thresholding P_syn, minimum intermediate value field intensity E_medWith radio-frequency protection ratio PR； Select or weighting function w (Δ τ is set_i), selecting non-receiver to be received and situational variables, it is dry than CNIR value that situational variables is that load is made an uproar Or covering spot probability；

Reference time t₀There are two kinds of selection standards: the first is to arrive the earliest and power is more than P_synTime of advent of signal be Benchmark, the second is on the basis of the time of advent of level peak signal；Synchronize thresholding P_syn=N₀+Margn；

Described weighting function w (Δ τ_i) it is: $w\left({\mathrm{\&Delta;\&tau;}}_i\right)=\left\{\begin{array}{c}0,\mathrm{\&Delta;}{\mathrm{\&tau;}}_i<-T_{GI}\\ 1,-T_{GI}\&le;{\mathrm{\&Delta;\&tau;}}_i\&le;T_{GI}\\ 0,\mathrm{\&Delta;}{\mathrm{\&tau;}}_i>T_{GI}\end{array}\right.,$ Wherein, T_GIRepresent frame head length；

Step G: determine reference time t₀, and each transmitter time of advent is relative to reference time t₀Delay inequality Δ τ_i, and obtain Obtain each grid available power and weighting function w (the Δ τ of jamming power_i) value, signal level is set and is synchronizing more than thresholding Signal is synchronized by receiver, according to weighting function w (Δ τ_i) calculate each transmitter available power at each grid central point with dry Disturb power, obtain synthesis available power U and the lognormal distribution parameter of cumulative interference power I of each grid central point；If point Analysis variable is CNIR, calculates the CNIR value covering field intensity more than or equal to the region of minimum intermediate value field intensity, the CNIR value in other regions Represent with 255, the CNIR value of calculating is sent to GIS module；If situational variables is for covering spot probability, calculate covering field intensity big The synthesis available power of each grid and the correlation coefficient of cumulative interference power and covering in the region equal to minimum intermediate value field intensity Spot probability, in other regions, the covering spot probability of grid is 0, and the spot probability of grid is sent to GIS module；

Synthesis available power U of grid central point: $U=\underset{i=1}{\overset{N_1}{\mathrm{\&Sigma;}}}{U}_i=\underset{i=1}{\overset{N_1}{\mathrm{\&Sigma;}}}{P}_{ri}w\left({\mathrm{\&Delta;\&tau;}}_i\right);$

The cumulative interference power I of grid central point: $I=\underset{i=1}{\overset{N_1}{\mathrm{\&Sigma;}}}{I}_i+\underset{j=1}{\overset{N_2}{\mathrm{\&Sigma;}}}{I}_j+\underset{k=1}{\overset{N^{\&prime;\&prime;}}{\mathrm{\&Sigma;}}}{I}_k=\underset{i=1}{\overset{N_1}{\mathrm{\&Sigma;}}}{P}_{ri}\&lsqb;1-w\left({\mathrm{\&Delta;\&tau;}}_i\right)\&rsqb;+\underset{j=1}{\overset{N_2}{\mathrm{\&Sigma;}}}{P}_{rj}+\underset{k=1}{\overset{N^{\&prime;\&prime;}}{\mathrm{\&Sigma;}}}{P}_{rk};$

Wherein, N₁For receiving power in SFN more than P_synTransmitting number of stations, N₂For receiving power in SFN less than P_syn Transmitting number of stations, N " be the outside co-channel transmitter number of SFN, U_iHaving of place generation is being received for i-th transmitter With signal power, P_riThe local power received for i-th transmitter, I_iThe interference that place produces is being received for i-th transmitter Signal power；

P_riStatistical property obeys logarithm normal distribution, and the dB value changed is W_ri=10 log₁₀(P_ri)；

W_riAverageStandard varianceWherein,It is the standard variance of shadow fading, Ground digital television broadcast takes 5.5dB；Represent the local average power that i-th transmitter receives；

U_iDB value X converted_i=10 log₁₀(U_i) averageAnd standard varianceIt is respectively as follows:

$m_{X_i}=m_{W_{ri}}+10\&CenterDot;{\log }_{10}\left(w\right({\mathrm{\&Delta;\&tau;}}_i\left)\right)=10\&CenterDot;{\log }_{10}\left(\stackrel{\&OverBar;}{P_{ri}}\right)+10\&CenterDot;{\log }_{10}\left(w\right({\mathrm{\&Delta;\&tau;}}_i\left)\right),{\mathrm{\&sigma;}}_{X_i}={\mathrm{\&sigma;}}_{W_{ri}}=5.5dB;$

I_iDB value Y converted_i=10 log₁₀(I_i) averageAnd standard varianceIt is respectively as follows:

$m_{Y_i}=m_{W_{ri}}+10\&CenterDot;{\log }_{10}(1-w({\mathrm{\&Delta;\&tau;}}_i\left)\right)=10\&CenterDot;{\log }_{10}\left(\stackrel{\&OverBar;}{P_{ri}}\right)+10\&CenterDot;{\log }_{10}(1-w({\mathrm{\&Delta;\&tau;}}_i\left)\right),{\mathrm{\&sigma;}}_{Y_i}={\mathrm{\&sigma;}}_{W_{ri}}=5.5dB;$

DB value X and Y that synthesis available power U and cumulative interference power I convert are respectively as follows:

X=10log₁₀(U), Y=10log₁₀(I)；

K factor lognormal method is utilized just to ask for the logarithm of synthesis available power U of each grid central point and cumulative interference power I State distributed constant, specifically:

First, willWithValue be converted into napier's value；

Then, napier is utilized to be worth average m calculating X_XAnd variance

$m_X=ln\&lsqb;\underset{i=1}{\overset{N_1}{\&Sigma;}}\exp (m_{X_i}+\frac{{{\mathrm{\&sigma;}}_{X_i}}^2}{2})\&rsqb;-\frac{{{\mathrm{\&sigma;}}_X}^2}{2},{{\mathrm{\&sigma;}}_X}^2=ln\&lsqb;1+k\frac{\underset{i=1}{\overset{N_1}{\mathrm{\&Sigma;}}}\{\exp \left({{\mathrm{\&sigma;}}_{X_i}}^2\right)-1\}\exp (2m_{X_i}+{{\mathrm{\&sigma;}}_{X_i}}^2)}{{\left(\underset{i=1}{\overset{N_1}{\mathrm{\&Sigma;}}}\exp \right(m_{X_i}+\frac{{{\mathrm{\&sigma;}}_{X_i}}^2}{2}\left)\right)}^2}\&rsqb;;$

Napier is utilized to be worth average m calculating Y_YAnd variances sigma_Y ²For:

$m_Y=\ln \&lsqb;\underset{i=1}{\overset{N_1+N_2+N^{\&prime;\&prime;}}{\mathrm{\&Sigma;}}}\exp (m_{Y_i}+\frac{{{\mathrm{\&sigma;}}_{Y_i}}^2}{2})\&rsqb;-\frac{{{\mathrm{\&sigma;}}_Y}^2}{2},{{\mathrm{\&sigma;}}_Y}^2=ln\&lsqb;1+k\frac{\underset{i=1}{\overset{N_1+N_2+N^{\&prime;\&prime;}}{\mathrm{\&Sigma;}}}\{\exp \left({{\mathrm{\&sigma;}}_{Y_i}}^2\right)-1\}\exp (2m_{Y_i}+{{\mathrm{\&sigma;}}_{Y_i}}^2)}{{\left(\underset{i=1}{\overset{N_1+N_2+N^{\&prime;\&prime;}}{\mathrm{\&Sigma;}}}\exp \right(m_{Y_i}+\frac{{{\mathrm{\&sigma;}}_{Y_i}}^2}{2}\left)\right)}^2}\&rsqb;;$

K is correction factor, depends on the standard variance receiving signal, when standard variance is less than 6dB, takes 0.7；

Finally, the m that will obtain_X、σ_X、m_YAnd σ_YNapier's value be converted into dB value；

The relevant moments method of coupling is utilized to calculate synthesis available power and the correlation coefficient of cumulative interference power of each grid, specifically:

First, X is made_i'=ln (U_i), then X_i'=λ X_i, parameterThen X_i' averageWith standard side DifferenceIt is respectively as follows:In like manner, Y is made_i'=ln (I_i), X'=ln (U), Y'=ln (I), then Y_i' With Y_i, X' Yu X, Y' Yu Y be respectively provided with X_i' and X_iBetween identical relation, and X_i' and Y_i' correlation coefficientEqual to X_iAnd Y_i's Correlation coefficientThe correlation coefficient r of X' Yu Y'_X'Y'Correlation coefficient r equal to X and Y_XY；

Make variable φ=X'+Y', then average m of φ_φ=m_X'+m_Y', variances sigma_φ ²=σ_X' ²+2r_X'Y'σ_X'σ_Y'+σ_Y' ²；Coupling U's Yu I Correlation Moment $E\&lsqb;UI\&rsqb;=E\&lsqb;e^{X^{\&prime;}}{e}^{Y^{\&prime;}}\&rsqb;=E\&lsqb;e^{X^{\&prime;}+Y^{\&prime;}}\&rsqb;=E\&lsqb;e^{\mathrm{\&phi;}}\&rsqb;=e^{{}_{m_{\mathrm{\&phi;}}+{{\mathrm{\&sigma;}}_{\mathrm{\&phi;}}}^2\mathrm{/2}}},$ Correlation Moment E [UI] can also be expressed as:

$\begin{array}{c}E\&lsqb;UI\&rsqb;=E\&lsqb;e^{X^{\&prime;}}{e}^{Y^{\&prime;}}\&rsqb;=E\&lsqb;e^{X^{\&prime;}+Y^{\&prime;}}\&rsqb;\\ =E\&lsqb;(e^{{X_1}^{\&prime;}}+e^{{X_2}^{\&prime;}}+...+e^{{X_{N_1}}^{\&prime;}})(e^{{Y_1}^{\&prime;}}+e^{{Y_2}^{\&prime;}}+...+e^{{Y_{N_1+N_2+N^{\&prime;\&prime;}}}^{\&prime;}})\&rsqb;\\ =\underset{i=1}{\overset{N_1}{\mathrm{\&Sigma;}}}\underset{i=1}{\overset{N_1+N_2+N^{\&prime;\&prime;}}{\mathrm{\&Sigma;}}}E\&lsqb;\left(e^{{X_i}^{\&prime;}+{Y_i}^{\&prime;}}\right)\&rsqb;\\ =\underset{i=1}{\overset{N_1}{\mathrm{\&Sigma;}}}\underset{i=1}{\overset{N_1+N_2+N^{\&prime;\&prime;}}{\mathrm{\&Sigma;}}}{e}^{m_{{X_i}^{\&prime;}}+m_{{Y_i}^{\&prime;}}+1/2\&CenterDot;({\mathrm{\&sigma;}}_{X_i^{\&prime;}}^2+{\mathrm{\&sigma;}}_{Y_i^{\&prime;}}^2+2r_{X_i^{\&prime;}{Y}_i^{\&prime;}}{\mathrm{\&sigma;}}_{X_i^{\&prime;}}{\mathrm{\&sigma;}}_{Y_i^{\&prime;}})}\\ =v\end{array}$

Thus obtain: $r_{XY}=r_{X^{\&prime;}{Y}^{\&prime;}}=\frac{2\&lsqb;\ln \left(v\right)-(m_{X^{\&prime;}}+m_{Y^{\&prime;}})\&rsqb;-({{\mathrm{\&sigma;}}_{X^{\&prime;}}}^2+{{\mathrm{\&sigma;}}_{Y^{\&prime;}}}^2)}{2{\mathrm{\&sigma;}}_{X^{\&prime;}}{\mathrm{\&sigma;}}_{Y^{\&prime;}}},$ WithIt is respectively Y_i' average and standard Variance；

Wherein, complimentary cumulative function CCDF is utilized to calculate the covering spot probability of each grid, specifically:

The covering spot probability P of i-th grid_c(x_i,y_i) it is: P_c(x_i,y_i)=P{ γ ＞ PR}=P{ γ_dB＞ PR_dB}；(x_i, y_i) represent i-th grid central point coordinate；γ is CINR value,γ_dBAnd PR_dBRepresent the dB value taking γ Yu PR；

Make ψ=X-Y, then average m of ψ_ψ=m_X-m_Y, variances sigma_ψ ²=σ_X ²+σ_Y ²-2r_XYσ_Xσ_Y；If N₀, I separate, thenm_XAnd σ_XFor average and the standard variance of X, m_YAnd σ_YFor Y average and Standard variance；

For synthesizing the useful field intensity grid less than minimum intermediate value field intensity, the covering spot probability P of this grid_c=0；

Synthesis is useful by force more than or equal to the region of minimum intermediate value field intensity, if cumulative interference field intensity exists, then this grid Covering spot probability:

$P_c=Q\left(\frac{{\mathrm{PR}}_{dB}-(m_X-m_Y)}{\sqrt{{{\mathrm{\&sigma;}}_X}^2+{{\mathrm{\&sigma;}}_Y}^2-2r_{XY}{\mathrm{\&sigma;}}_X{\mathrm{\&sigma;}}_Y}}\right)\&times;Q\left(\frac{{\mathrm{PR}}_{dB}-(m_X-N_0)}{\sqrt{{{\mathrm{\&sigma;}}_X}^2}}\right)$

Otherwise, $P_c=Q\left(\frac{{(C/N)}_{dB}-(m_X-N_0)}{\sqrt{{{\mathrm{\&sigma;}}_X}^2}}\right);$

Wherein, function(C/N)_dBDB value for carrier-to-noise ratio thresholding；

The color comparator scope ratio that step H:GIS module defines with system according to covering spot probability value or the CNIR value of each grid Relatively, carrying out each grid color value and calculate and colour, output result covers on electronic chart and at terminal demonstration；

Step I: according to statistical regions needed for geographical feature selection, statistical condition is set；

Step J: meet the grid of statistical condition in filtering out selected areas, and the selection result is sent to GIS module, and calculate The percentage ratio of total grid number in meeting statistical regions shared by the grid number of statistical condition in selected areas；

The budget result meeting the grid that statistical condition requires is covered on electronic chart and shows at end by step K:GIS module End, is unsatisfactory for the budget result not display of the grid that statistical condition requires；

Step L: optimize region according to needed for geographical feature selection, arranges maximum genetic algebra G_maxWith population scale P_s, population is advised Mould P_sArrange according to transmitter number；

Step M: in the case of not changing transmitter parameter, calculates based on genetic algorithm and optimization areal coverage can be made to reach The emission delay optimum combination of each transmitter and coverage rate in the highest SFN；Step M implements step:

Step 401: initialize genetic algorithm parameter, including: population scale P_s, maximum genetic algebra G_max, be in SFN sharp The transmitter number N of the state of living_opt, coding figure place l, generation gap GGAP, mutation probability P_mWith crossover probability P_cr；

Wherein, l=20, GGAP=0.9, P are set_m=0.01, P_cr≥0.7；

Step 402: select the N that SFN is internal and is active_optIndividual transmitter emission delay, as genetic algorithm Decision variable, N_optIndividual decision variable composition time delay vector delay:

Step 403: determine span and the constraints of decision variable；

I-th transmitter emission delay delay is set_iSpan be: 0≤delay_i≤T_GI, and delay_i=n × 0.1 μ S, 1≤i≤N_opt, n is positive integer；

Step 404: use Gray code to delay_iEncode；

Step 405: stochastic generation P_sIndividual a length of l × N_optTime delay vector as initial population；

Step 406: the value of each time delay vector in contemporary population is sent to each transmitting station, calculates each transmitting station to excellent The time of advent of each grid central point in change region, selection state is penetrated as non-transmitting-receiving with nonactivated transmitter for activating Machine, covers the field intensity covering spot probability more than or equal to the grid of minimum intermediate value field intensity, finally calculates mesh in calculation optimization region Offer of tender numerical value f (delay):

$f\left(delay\right)=(1-\underset{j=1}{\overset{M^{\&prime;}}{\&Sigma;}}{\mathrm{Cov}}_j/M^{\&prime;})\&times;100\%$

Wherein, M' is grid number total in optimizing region；

Step 407: use fitness assignment algorithm based on linear ordering to distribute fitness value, selecting pressure reduction is 2, object function Being worth the least, fitness value is the biggest；Using time delay vector minimum for target function value in the present age as optimum time delay vector, and preserve optimum Coverage rate corresponding to time delay vector；

Step 408: judge whether to meet end condition, if it is, the coverage rate that the optimum time delay of output is vectorial and corresponding, optimum Time delay vector i.e. emission delay optimum combination；If it does not, continue executing with step 409；End condition refers to: reached maximum When genetic algebra or optimum time delay vector are during evolution without changing；

Step 409: use random ergodic sampling operator；

Step 410: use single-point crossover operator；

Step 411: use Discrete mutation operator；Afterwards, heavily insertion based on relevance grade is used to guarantee P_s× (1-GGAP) is individual the suitableeest The time delay vector answered always is traveled to the next generation continuously；Then 406 execution are gone to step；

Step N: the emission delay optimum combination of each transmitter and coverage rate in output SFN.

3. the method obtained and optimize DTMB SFN coverage rate as claimed in claim 2, it is characterised in that described step In D, use ITU-R P.1546, ITU-R P.370, ITU-R P.526 or Okumura-Hata radio waves propagation model and school thereof Positive model, the reception power of each transmitting station of calculating state of activation to each grid central point and field intensity value.

4. the method obtained and optimize DTMB SFN coverage rate as claimed in claim 2, it is characterised in that described step In G, X_i' and Y_i' correlation coefficientEqual to X_iAnd Y_iCorrelation coefficientEqual to 1.

5. the method obtained and optimize DTMB SFN coverage rate as claimed in claim 2, it is characterised in that described step In G, if situational variables is CNIR, for synthesizing the useful field intensity region less than minimum intermediate value field intensity, make CNIR=255；For Synthesis is useful by force more than or equal to the region of minimum intermediate value field intensity, if cumulative interference field intensity exists, then and CNIR=m_X- 10log10(10^m_Y/10+10^N₀/ 10), otherwise CNIR=m_X-N₀。

6. the method obtained and optimize DTMB SFN coverage rate as claimed in claim 2, it is characterised in that described step In J, when selecting situational variables for covering spot probability, under fixed reception mode, when the covering spot probability of jth grid When value is more than or equal to 70%, then it is assumed that this grid is capped, cover parameter Cov_j=1, otherwise, Cov_j=0, then statistical regions Coverage rateM is the grid number that statistical regions is total.