Summary of the invention
In actual scene, a LED lamp is usually to be made of multiple LED, therefore, for visible light interference channel net
Network can carry out more LED Wave beam forming designs, to reduce the interference between each cell.
It constructs a kind of transmitting signal with maximum entropy distribution first to improve transmission capacity, is then directed to different applications
Scene and network size, the present invention propose visible light communication multiple cell interference management method respectively, specifically include a kind of concentration
Formula visible light communication multiple cell interference management method and a kind of distributed visible light communication multiple cell interference management method.
Centralized visible light communication multiple cell interference management method is to establish master controller, collects the complete of each transmitting-receiving pair
Portion's information optimizes processing, and a kind of centralized visible light communication multiple cell interference management method includes the following steps:
Step 1, for i-th of user in visible light communication network, it is not added the transmitting signal X of direct current biasingi
It is configured;
Step 2, the ABG (Alpha-Beta-Gamma) for solving i-th of user is distributed corresponding probability density function;
Step 3, the transmitting signal X that direct current biasing is not added is calculatediDifferential entropy { h (Xi)};
Step 4, it is assumed that comprising K transmission pair in visible light communication network, in each transmission in, transmitting terminal includes N
A LED, receiving end only include a photoelectric detector PD (PhotoDetector), solve transmitting signal XiOptical power and electric work
Rate;
Step 5, channel parameter is set;
Step 6, the channel capacity lower bound based on ABG distribution of i-th of user is solved
Step 7, optimal beam forming designs;
Step 8: beamformer output forms matrix { wi}。
In step 1, the transmitting signal X of direct current biasing is not addediObey continuous probability distribution, the probability of this probability distribution
Density function is fi(Xi), and probability density function fi(Xi) meet following condition:
Wherein, AiFor the amplitude peak and A of source signali> 0, then the transmitting signal X of direct current biasing is not addediAmplitude model
Enclose, mean value and average electrical power difference it is as follows:
|Xi|≤Ai,
Wherein εiFor the average electrical power of target user i,To average.
In step 2, the ABG of i-th of user is distributed corresponding probability density function fi ABG(Xi) are as follows:
Wherein, K indicates total number of users, i.e., transmission is to sum, parameter alphai, βiAnd γiFor following solution of equations:
Wherein π is pi, and e is natural logrithm, functionX∈[-Ai,Ai],
Erf () is Gauss error function.
In step 3, the transmitting signal X that direct current biasing is not added is calculated by following formulaiDifferential entropy { h (Xi) }:
Wherein αiAnd γiFor ABG distribution parameter, log2() is the logarithmic function with 2 for the truth of a matter, and ln () is nature pair
Number.
In step 4, for i-th of transmission pair, w is enablediIndicate Wave beam forming vector,
wi,nIndicate the weighting coefficient of n-th of LED of i-th of transmission pair, n=1 ..., N,The real number space of expression N-dimensional, i-th
Transmitting terminal transmits information XiTo i-th of receiving end, wherein | Xi|≤Ai,AndK
Transmission is indicated to sum, the signal s that i-th of transmitting terminal is emittediAre as follows:
Wherein liFor the direct current biasing vector of i-th of user,biFor i-th user's
Direct current biasing size, and bi>=0, it is necessary for non-negative due to emitting signal in visible light communication, wiIt must satisfy:
Wherein | | it is modulus,
The resulting transmitting signal s of step 4iOptical powerAnd electrical powerIt is respectively as follows:
Wherein K indicates transmission to sum, and N indicates that LED sum in each transmitting terminal, i indicate the serial number of transmission pair,N indicates LED serial number in each transmitting terminal,| | | | to seek 2- norm.
Step 5 includes: because the channel gain of usual line-of-sight propagation (LOS) link is much larger than the scattering chain reflected to form
Road, it is possible to which the influence for only considering line-of-sight propagation link shines model according to lambert, n-th of LED to i-th on j-th of lamps and lanterns
The channel gain g of a useri,j,nIt indicates are as follows:
Wherein, cos () is cosine function, | | it is modulus, m indicates Lambert emission grade,
Log () is logarithmic function, and φ indicates the angle of departure of LED, φ1/2Indicate the half of half-power angle, ηcIndicate electro-optic conversion effect
Rate, ηlIndicate photoelectric conversion efficiency, G indicates the gain of trsanscondutance amplifier, ARIndicate receiving end physical area size,Wherein nrIndicate receiving end collector lens refractive index, APDIndicate the area of photoelectric detector, sin
() is SIN function, di,j,nIndicate n-th of LED at a distance from i-th of user, ψnIndicate the incidence angle of receiving end, ψFoVIt indicates
The visual field (FoV) width of user.
In step 6, pass through the channel capacity lower bound based on ABG distribution of i-th of user of following equations
Wherein, σ2For noise power, gi,jFor j-th of transmitting terminal to the channel gain of i-th of user, wjEmit for j-th
The Wave beam forming vector at end.
In step 7, according to convex optimum theory, under each ownership goal rate constraint and each LED power constraint condition, with
It minimizes the visible light beam that transmission power is target and forms frame, obtain following formula:
Wherein, s.t. expression suchthat, i.e., " so that ", RiIt is the targeted rate of i-th of receiving end, enFor in addition to i-th
Element is 1 and vector that other elements are 0, and min expression is minimized, and s.t. is such that, i.e., " so that ";
But first of above-mentioned optimization problem is constrained to a quadratic constraints, so that this problem is one non-convex secondary
It constrains double optimization problem (QCQP), is NP- difficulty problem according to this kind of optimization problems of existing document.
Following optimization problem is converted by above-mentioned formula:
Wherein gi,jFor j-th of transmitting terminal to the channel gain of i-th of user, K indicates transmission to sum, and N indicates each transmitting
LED sum in end,(parameter alphaj、γjUsing solving parameter alpha in step 2i, βiAnd γiMethod be calculated,
The subscript of parameter in equation group is changed to j),And
It is handled using semi definite programming technology SDR (Semi-DefiniteRelaxation), definitionSo thatAnd
And rank (Wi)=1, above-mentioned optimization problem relaxation are as follows:
Wherein Tr () is the mark for seeking matrix, this problem is convex semi definite programming problem, by up to interior-point algohnhm solution
Optimal solution out, i.e.,Minimum value and corresponding wi。
Centralized visible light communication multiple cell interference management method, it is desirable that carry out data between different LED light and information is total
It enjoys, this will lead to a large amount of signaling overheads.In network passback capacity by limited time, this problem will will limit visible light network
Using.In order to reduce the expense of data exchange, the invention also discloses a kind of distributed visible light communication multiple cell interference managements
Method includes the following steps:
Step 1, for i-th of user in visible light communication network, it is not added the transmitting signal X of direct current biasingi
It is configured;The transmitting signal X of direct current biasing is not addediObey continuous probability distribution, the probability density letter of this probability distribution
Number is fi(Xi), and probability density function fi(Xi) meet following condition:
Wherein, Ai> 0 is the amplitude peak of source signal, then the transmitting signal X of direct current biasing is not addediAmplitude range,
Value and average electrical power difference are as follows:
|Xi|≤Ai,
Wherein εiFor the average electrical power of target i,To average;
Step 2, the ABG (Alpha-Beta-Gamma) for solving i-th of user is distributed corresponding probability density function:
Wherein, K indicates the quantity of user, parameter alphai, βiAnd γiFor following solution of equations:
Wherein π is pi, and e is natural logrithm, functionX∈[-Ai,Ai],
Erf () is Gauss error function;
Step 3, the transmitting signal X that direct current biasing is not added is calculatediDifferential entropy { h (Xi) }:
Wherein αiAnd γiThe ABG distribution parameter solved by equation in step 2, log2() is the logarithm with 2 for the truth of a matter
Function, ln () are natural logrithm;
Step 4, it is assumed that comprising K transmission pair in visible light communication network, in each transmission in, transmitting terminal includes N
A LED, receiving end only include a photoelectric detector PD (PhotoDetector), solve transmitting signal XiOptical power and electric work
Rate: for i-th of transmission pair, w is enablediIndicate Wave beam forming vector,wi,nIndicate i-th
The weighting coefficient of n-th of LED of a transmission pair, n=1 ..., N,Indicate the real number space of N-dimensional, i-th of transmitting terminal transmission
Information XiTo i-th of receiving end, wherein | Xi|≤Ai,AndI-th of transmitting terminal
The signal s emittediAre as follows:
Wherein liFor the direct current biasing vector of i-th of user,biFor i-th user's
Direct current biasing size, and bi>=0, it is necessary for non-negative due to emitting signal in visible light communication, wiIt must satisfy:
Wherein | | it is modulus,
Emit signal siOptical powerAnd electrical powerIt is respectively as follows:
Wherein K indicates transmission to sum, and N indicates that LED sum in each transmitting terminal, i indicate the serial number of transmission pairN indicates LED serial number in each transmitting terminal,| | | | to seek 2- norm;
Step 5, channel parameter is set: being shone model according to lambert, n-th of LED to i-th user on j-th of lamps and lanterns
Channel gain gi,j,nIt indicates are as follows:
Wherein, cos () is cosine function, | | it is modulus, m indicates Lambert emission grade,
Log () is logarithmic function, and φ indicates the angle of departure of LED, φ1/2Indicate the half of half-power angle, ηcIndicate electro-optic conversion effect
Rate, ηlIndicate photoelectric conversion efficiency, G indicates the gain of trsanscondutance amplifier, ARIndicate receiving end physical area size,Wherein nrIndicate receiving end collector lens refractive index, APDIndicate the area of photoelectric detector, sin
() is SIN function, di,j,nIndicate n-th of LED at a distance from i-th of user, ψnIndicate the incidence angle of receiving end, ψFoVIt indicates
The visual field width of user;
Step 6, centralized optimization problem K distributed subproblem is decomposed into solve;According to change of direction multiplier
ADMM (Alternating Direction Method of Multipliers) optimization algorithm, by centralized interference management frame
The optimization problem of frame converts are as follows:
Wherein gi,jFor j-th of transmitting terminal to the channel gain of i-th of user, AndRiIt is the targeted rate of i-th of user, enFor
In addition to i-th of element is 1 and vector that other elements are 0, σ2For noise power, wjFor j-th of transmitting terminal Wave beam forming to
Amount, local variable is that local interference information is vi, vi=[vi,1,...,vi,j,...vi,K,ηi]T, j ≠ i, andWherein variable vi,jTable
Show the interference of j-th of LED light to i-th of user;Variable ηiIndicate total interference that i-th of LED light receives, Tr () is to seek square
The mark of battle array;
Define global variableThe relationship of local variable and global variable are as follows:
vi=Ziv
WhereinReal number space is tieed up for (K+1) × (K+1), the problem after conversion is write a Chinese character in simplified form
Are as follows:
Wherein subproblem fi(vi) is defined as:
s.t.uiTr(Wigi,igi,i T)≥ηi+ci
Simplified problem is a globally consistent problem, i.e., according to constraint condition, local variable viIt finally should be with the overall situation
Variable v is consistent, is solved using alternating direction multiplier ADMM algorithm, corresponding LagrangianLζ({vi},τi, v) are as follows:
Wherein τiIt for the Lagrange multiplier with optimization problem constraint link, or is dual variable, ζ is penalty factor;
Step 7, the number of iterations t=0, initializing variable v (0), τ (0) and penalty factor ζ are enabled, ξ > 0 is given precision;
Step 8, local variable { v when t+1 iteration is updatedi(t+1) } subproblem f, is solvedi(vi) and following K distribution
Formula optimization problem:
WhereinExpression makes the smallest value of B, τ in Ai() is dual variable, vi() is that part becomes
Amount;
Step 9, local variable { v when t+1 iteration is exchanged with other LED lighti(t+1)};
Step 10, the global variable v (t+1) when t+1 iteration is updated, it may be assumed that
WhereinExpression makes the smallest value of B in A;
Step 11, dual variable τ when t+1 iteration is updatedi(t+1), it may be assumed that
τi(t+1)=τi(t)+ζ(vi(t+1)-Ziv(t+1)).
Step 12, whenWhen, terminate iteration;Otherwise, t is updated to t+1, and returned
Step 8;
Step 13, beamformer output forms matrix { wi}。
The present invention relates to visible light communication field interference channel MISO multiple cell visible light interference management frames.Benefit of the invention
With information theory, signal processing and convex optimization method, centralized and distributed two kinds of interference management frames are proposed.
The utility model has the advantages that interference management frame proposed by the invention can effectively reduce the interference of minizone, guarantee to use
Family targeted rate, while transmission power is reduced, improve efficiency.The present invention proposes to propose targetedly side for different scenes
Method, while method computational accuracy is high, solving speed is fast, and distributed interference management method can effectively reduce information in system
Exchange, mitigates network load, these features decrease the energy consumption of system entirety, improve system energy efficiency.
Specific embodiment
The present invention will be further described with reference to the accompanying drawings and embodiments.
The present invention proposes two kinds of new frames for visible light interference management frame, using information theory, signal processing and
Convex optimization method proposes and meets visible light communication constraint requirements, possesses closed expression and tighter channel capacity calculates
Method, the design and calculating of completion visible light interference management and Wave beam forming that can be simple and fast.
Fig. 2 shows the basic flow of the centralized visible light communication multiple cell interference management using technical solution of the present invention
Journey, it is seen that light interference channel system model is as shown in Figure 1, the specific steps are as follows:
Step 1: system relevant parameter and input signal are arranged according to design requirement.
For i-th of user, the transmitting signal that direct current biasing is not added is Xi, parameter is as follows:
Peak value: | Xi|≤Ai;
Mean value:
Mean-square value:
Step 2: seek ABG (Alpha-Beta-Gamma) Parameter Relation of each user:
Wherein π is pi, and e is natural logrithm,Erf () is Gauss mistake
Difference function.
Obtain corresponding parameter alphai, βiAnd γi, calculate the probability density function of the ABG probability distribution of each user:
Step 3: calculating information source differential entropy { h (Xi), i.e.,Wherein ln () is natural logrithm.
Step 4: setting emission parameter.
Transmission is to quantity: K;
LED quantity in each transmitting terminal: N
For i-th,A transmitting terminal,
Input reference signal: | si|≤Ai;
Input signal mean value:
Input signal mean-square value:
Direct current biasing:
Emit signal: xi=wisi+bi,
Step 5: setting channel parameter.The channel gain of n-th of LED to i-th user can be expressed as on j-th of lamps and lanterns
WhereinIndicate Lambert emission grade;φ indicates the angle of departure of LED, φ1/2Indicate half-power
The half at angle;ηcIndicate electro-optical efficiency;ηlIndicate photoelectric conversion efficiency;The gain of G expression trsanscondutance amplifier;ARExpression connects
Receiving end physical area size;di,nIndicate n-th of LED at a distance from i-th of user;ψnIndicate the incidence angle of receiving end;ψFoVTable
Show the visual field (FoV) width of user.The definition of receiving end physical area size isWherein nrExpression connects
Receiving end collector lens refractive index, APDIndicate the area of photoelectric detector.
Step 6: solve the channel capacity lower bound based on ABG distribution for i-th of user:
Wherein RiIt is the targeted rate of i-th of receiving end, enTo be 1 in addition to i-th of element and vector that other elements are 0.
π is pi, σ2For noise power, | | it is modulus, | | | | to seek 2- norm, min is to minimize, αiAnd γiFor ABG
Distribution parameter, εiFor the average electrical power of corresponding transmitting terminal, gi,jFor j-th of transmitting terminal to the channel gain of i-th of receiving end, wj
For the Wave beam forming vector of j-th of transmitting terminal, log2() is the logarithmic function with 2 for the truth of a matter, εiFor average electrical power, biFor
Direct current biasing, AiFor the amplitude peak of source signal.
Step 7: Wave beam forming optimization problem is solved, wave beam formed matrix is obtained:
WhereinAndπ is pi,
σ2For noise power, | | it is modulus, min is to minimize, εiFor the average electrical power of corresponding transmitting terminal, gi,jIt is sent out for j-th
Penetrate channel gain of the end to i-th of receiving end, wjFor the Wave beam forming vector of j-th of transmitting terminal, log2It is the truth of a matter that (), which is with 2,
Logarithmic function, εiFor average electrical power, biFor direct current biasing, AiFor the amplitude peak of source signal.
Fig. 3 shows the basic flow of the distributed visible light communication multiple cell interference management using technical solution of the present invention
Journey, it is seen that light interference channel system model is as shown in Figure 1, the specific steps are as follows:
Step 1: system relevant parameter and input signal are arranged according to design requirement.
For i-th of user, the transmitting signal that direct current biasing is not added is Xi, parameter is as follows:
Peak value: | Xi|≤Ai;
Mean value:
Mean-square value:
Step 2: seeking the solving equations of each user
Wherein π is pi, and e is natural logrithm,Erf () is Gauss mistake
Difference function.
Obtain corresponding parameter alphai, βiAnd γi, calculate the probability density function of the ABG probability distribution of each user:
Step 3: calculating information source differential entropy { h (Xi), i.e.,Wherein ln () is natural logrithm.
Step 4: setting emission parameter.
Transmission is to quantity: K;
LED quantity in each transmitting terminal: N
For i-th,A transmitting terminal,
Input reference signal: | si|≤Ai;
Input signal mean value:
Input signal mean-square value:
Direct current biasing:
Emit signal: xi=wisi+bi,
Step 5: setting channel parameter.The channel gain of n-th of LED to i-th user can be expressed as on j-th of lamps and lanterns
WhereinIndicate Lambert emission grade;φ indicates the angle of departure of LED, φ1/2Indicate half-power
The half at angle;ηcIndicate electro-optical efficiency;ηlIndicate photoelectric conversion efficiency;The gain of G expression trsanscondutance amplifier;ARExpression connects
Receiving end physical area size;di,nIndicate n-th of LED at a distance from i-th of user;ψnIndicate the incidence angle of receiving end;ψFoVTable
Show the visual field (FoV) width of user.The definition of receiving end physical area size isWherein nrExpression connects
Receiving end collector lens refractive index, APDIndicate the area of photoelectric detector.
Step 6: distributed optimization model of the construction based on ADMM algorithm;
Step 7: iterative calculation solves the distributed optimization model of ADMM algorithm, obtains beam forming matrix.
In order to verify effect of the present invention, simulation comparison experiment has been carried out.
For basic visible light interference channel communication system, the calculation formulae for related parameters is as follows:
System average electrical power:
System average light power:
System Normalized Signal/Noise Ratio (dB):
System peak-to-average power ratio:
Given system Signal to Noise Ratio (SNR) and system peak-to-average power ratioIn the case where,
System average electrical power:
System average light power:
For i-th of LED light,
Average electrical power distribution coefficient: vi,vi∈[0,1],
Average light power distribution coefficient: μi,μi∈[0,1],
Average electrical power: εi=viε;
Average light power: Ai=μiA;
Peak-to-average power ratio:
For distributed visible light communication interference management frame, LED light and receiving end position coordinates are as shown in table 1, distribution
Formula visible light communication interference management frame emulation experiment experiment 1 illustrates this hair according to the emulation experiment parameter given shown in table 2
Bright technical solution is in varied situations with the general power situation of change that targeted rate changes, distributed visible light communication interference pipe
Frame emulation experiment experiment 2 is managed according to the emulation experiment parameter given shown in table 3, technical solution of the present invention is illustrated and is not sympathizing with
With the relative error situation of change that the number of iterations changes under condition.Simulation result according to Fig.4, can be seen that the present invention is dry
Disturb management method be it is feasible, optimal general power can be solved.Algorithmic statement speed proposed by the present invention as can be seen from Figure 5
It spends quickly, and error can be less than 5%.(figure Chinese and English is explained as follows: ElectricalPower: electrical power,
Ratethreshold: targeted rate, RelativePowerError: relative power error, IterationIndex: iteration time
Number, Bits/sec/Hz: bits per second per Hertz, dBm: decibel milliwatt)
The calculation formulae for related parameters is as follows:
For basic visible light interference channel communication system, the calculation formulae for related parameters is as follows:
System average electrical power:
System average light power:
System Normalized Signal/Noise Ratio (dB):
System peak-to-average power ratio:
Given system Signal to Noise Ratio (SNR) and system peak-to-average power ratioIn the case where,
System average electrical power:
System average light power:
System gross electric capacity:
For i-th of LED light,
Average electrical power distribution coefficient: vi,vi∈[0,1],
Average light power distribution coefficient: μi,μi∈[0,1],
Average electrical power: εi=viε;
Average light power: Ai=μiA;
Peak-to-average power ratio:
Wherein assume A1=A2=A3=A4, ε1=ε2=ε3=ε4And
Assuming that in (20 × 20 × 5m3) visible light communication network network, wherein one, room corner conduct are disposed in room
The origin (0,0,0) of rectangular coordinate system in space includes (K=4) a transmission pair in network.In other words, 4 LED are disposed in room
Lamp, every lamp is the light emitting array being made of multiple LED, and the center of array is as shown in table 3, and parasang is rice in table.
Table 1:LED lamp and receiving end position coordinates
Table 2: visible light communication interference management frame emulation experiment tests 1 parameter
Table 3: visible light communication interference management frame emulation experiment tests 2 parameters
The present invention provides visible light communication multiple cell interference management methods, implement method and the way of the technical solution
There are many diameter, the above is only a preferred embodiment of the present invention, it is noted that for those skilled in the art
For, various improvements and modifications may be made without departing from the principle of the present invention, these improvements and modifications should also regard
For protection scope of the present invention.All undefined components in this embodiment can be implemented in the prior art.