CN103228035A - Adaptive modulation method based on average BER constraint in distributed antenna system - Google Patents

Abstract

The invention discloses an adaptive modulation method based on average BER (Bit Error Rate) constraint in a wireless distributed multi-antenna system, and provides an optimization design scheme of an optimal adaptive switching threshold. Generally, an AM (Amplitude Modulation) scheme based on instant BER constraint can allow an average BER of the system to meet a service quality requirement, and can strictly control the average BER to be much less than a target BER, but the strict constraint results in failed adaptive optimization of the switching threshold according to a channel condition, thereby resulting in loss of SE (Spectrum Efficiency). In order to avoid the loss of the SE caused by the strict constraint, an adaptive threshold optimization scheme combining a KKT (Karush-Kuhn-Tucker) condition and a Newton method is provided by maximizing the SE of the system and designing the adaptive modulation method based on the average BER constraint under the target BER constraint condition. A Matlab simulation platform indicates that the provided AM method and the adaptive switching threshold can ensure that the SE of the system can be improved greatly under the condition that the average BER of the system meets a requirement of the target BER.

Description

The self-adaptive modulation method that retrains based on average BER in the distributing antenna system

Technical field

The invention belongs to wireless communication field, relate to the adaptive modulation scheme design of radio communication, relate to Adaptive Modulation handoff threshold optimized Algorithm and corresponding dispersion ratio adaptive modulation scheme design in the distributing antenna system in particular.

Background technology

The rapid emergence of mobile internet and compunication, broadband multimedia services are that the data service of representative rises to core business gradually; The data service of this fast development needs high data rate to insert, and makes the radio spectrum resources growing tension, so the transmission technology of spectral efficient is more just constantly sought to alleviate limited frequency spectrum resources by research institution both at home and abroad.Following wireless communication technology need be supported the transfer of data of high-quality and two-forty, in recent years, along with the growing tension of frequency spectrum resource and the sustainable growth of user's request, though the transmission technology that high frequency such as MIMO and OFDM is imitated has improved power system capacity greatly, it still can't satisfy the demand of people to the service of broadband high network quality.The wideband transmit ability that the MIMO-OFDM technology provides can improve point-to-point transmission capacity, but owing to adopted higher frequency range, causes system ovelay range and coverage rate to be restricted.Reach the desired transmission rate of 4G system,, have only by dwindling radius of society or the sub-district cracking solves, but this has increased the operation cost of 4G system undoubtedly if continue to adopt existing cellular system.In addition, because high-lager building is more and more, cause cellular cell to have a large amount of coverage holes, if adopt the method for simple raising transmitting power, not only the construction to the base station proposes higher requirement, also can bring the minizone Communication Jamming simultaneously, causes the utilance of frequency spectrum resource to reduce.So, conventional cellular systems occur these switch frequently, disturbing increases, problems such as cost raising can't adapt to the demand for development of future communications system.

The spaced antenna technology provides the space resources more efficient use as the open architecture of following radio communication.At present, there are some researches show with centralized antenna system and compared that distributing antenna system has the ability of higher capacity, sub-district average coverage rate, lower transmitted power, anti-shadow fading and increases demands such as multiple wireless traffic.

Technology for self-adaptively transmitting can be according to the requirement of wireless communications environment and service quality (QoS), the various parameters that dynamically change transmitting terminal improve resource utilization ratio, to obtain higher system throughput and capacity, be applied in the wireless communication system as a kind of strong technology that improves the transmission system spectrum efficiency.

Since when the distribution character of antenna provides the space resources more efficient use also the resource allocation techniques to the advanced person stern challenge has more been proposed.At present, existing a lot of documents carry out power allocation scheme research at the distributed MIMO system; Yet AM is less relatively as the research of key technology in distributing antenna system that improves system spectral efficiency.

In the cordless communication network in future, Radio Link is under the prerequisite that guarantees QoS of survice, and higher transmission rate is the target that people pursue always.For strict guarantee QoS, common Adaptive Modulation (AM) scheme generally is based on instantaneous bit error rate (BER) constraint and designs, like this can simplified design, and can guarantee that the BER all the time of system can both satisfy target BER requirement, but, the requirement of this strictness can make the handoff threshold value not carry out adaptive optimization (being that threshold value is to fix) according to channel information, will unavoidably bring the loss of system spectral efficiency (SE).

For this reason, this patent provides based on the adaptive threshold value optimized Algorithm of average BER constraints and corresponding AM conceptual design, greatly improves in the hope of satisfying the SE that obtains to unite when system service quality (target BER) requires.

At present, existing document is studied at self adaptation handoff threshold optimization problem.In these AM conceptual designs, people such as Torrance propose to make average BER and ASE simultaneously with the method for the difference minimum of desired value, set up corresponding cost equation and search for optimum AM threshold value.Lai has provided a kind of Two-mode method and has optimized threshold value, and this method still needs to adopt the cost equation of Torrance proposition, comes down to a kind of modification of Torrance method, but only considers adjacent two kinds of modulation systems at every turn.It is that the cost equation is tried to achieve optimum adaptive threshold to maximize average SE that scholar Chung and Choi have proposed employing Lagrange optimization method, and it finds the solution same needs by a large amount of computer numerical search.In addition, the member of seminar is at Space Time Coding, adopt the Lagrange method to ask for optimum adaptive threshold value with maximization ASE, it is when separating Nonlinear System of Equations, adopted a kind of simplification Newton solution by iterative method, though can avoid original document to find the solution threshold value by a large amount of numerical searches, its iteration Jacobi matrix is only relevant with initial value, do not replace in real time according to the iteration intermediate quantity, the threshold value that finally causes asking for is one group of suboptimal solution.In these optimization methods, Torrance method and Lagrange method are similar, just represent average BER constraints with different forms in cost function.Based on above problem, the present invention will design its adaptive modulation scheme under average BER constraints, the maximizing system spectral efficiency at wireless distributed antenna system characteristics; And a kind of adaptive threshold optimized Algorithm of uniting Karush-Kuhn-Tucker (KKT) optimal conditions and Newton method is proposed, and can be when trying to achieve optimum thresholding, reducing the Lagrange scheduling algorithm needs a large amount of numerical searches find the solution the high complexity that causes.

Below will be described in detail purpose of the present invention and characteristic in conjunction with the accompanying drawings by instantiation, these specific embodiments are illustrative, do not have restricted.

Summary of the invention

The present invention be directed to distributing antenna system, proposed a kind of Adaptive Modulation thresholding optimization method, provide corresponding adaptive modulation scheme design based on average BER constraint.Purpose is to make that the system self-adaption switching threshold can be according to the channel condition adaptive optimization, maximizes the SE of system simultaneously satisfying target BER.Following steps have been adopted in distributing antenna system AM method that the present invention proposes and the design of corresponding thresholding prioritization scheme:

(1) distributing antenna system model and channel model have been provided.

The distributing antenna system model as shown in Figure 1.

Accompanying drawing 1 has provided the physical model of distributing antenna system, the border circular areas that it is D that cell unit is modeled as a radius, and each remote antenna (RA) at random is placed in the diverse location of sub-district, is designated as RA _i(i=1,2 ..., N), be connected to a CPU (MCU) by specific transmission channel (optical fiber).Be without loss of generality, suppose that each portable terminal (MT) all has L antenna.

By pertinent literature knowledge, through performance analysis and calculating can be in the hope of system's received signal to noise ratio (SNR) γ of i root transmitting antenna under this model _iProbability density function (PDF) expression formula is

$f_{{\mathrm{\γ}}_i}\left(\mathrm{\γ}\right)\≅\frac{\mathrm{\ξ}}{\sqrt{2\mathrm{\π}}{\widehat{\mathrm{\σ}}}_i\mathrm{\γ}}\exp (-\frac{{(101\mathrm{g\γ}-{\widehat{\mathrm{\μ}}}_i)}^2}{2{\widehat{\mathrm{\σ}}}_i^2})$ Expression formula 1

Wherein:

The average that is this APPROXIMATE DISTRIBUTION can be in order to the following formula subrepresentation:

Here ψ () is an Euler psi function;

The variance that is this APPROXIMATE DISTRIBUTION is

ζ () is a Reimann zeta function.

(2) provide the dispersion ratio AM scheme and the thresholding prioritization scheme of system according to the PDF of the system model of described step (1) and received signal to noise ratio.

The discrete adaptive modulation scheme of distributing antenna system is as follows, is BER by target setting BER ₀, instantaneous SNR is divided into several intervals, [γ _n, γ _N+1), n=0,1 ... .., N-1 is γ wherein ₀=0, γ _N+1=+∞, [γ ₀, γ ₁) situation that representation signal transmission is interrupted, N is the total number of system modulation mode.Planisphere size M for discrete M-QAM _nCan be defined as { M ₀=0 ...., M _n=2 ⁿ, n=1,2 ...., N-1}, when instantaneous SNR between [γ _n, γ _N+1) when interval, system has planisphere with selection and is of a size of M _nModulation system, corresponding message transmission rate b _n=log ₂M _n

Under instantaneous BER constraints, equaling target BER by the instantaneous BER that sets M-QAM can be in the hope of fixing self adaptation switching threshold.It is many that the threshold value of trying to achieve this moment can make that the average BER of system is strict controlled under the target BER, but this strict constraint will bring the decline of the SE of system inevitably.

Therefore, the present invention will average AM conceptual design and the optimization of handoff threshold value under the BER constraints, can greatly improve so that system obtains the SE of system when satisfying target BER.

Consider following nonlinear optimization problem

Maximizef (X), expression formula 2

S.t.g _i(X) 〉=0, expression formula 3

h _j(X)=0 expression formula 4

The Mathematical Modeling of KKT optimal conditions is expressed as following canonical form:

$\▿f\left(X^{\′}\right)+\mathrm{\Σ}{\mathrm{\λ}}_i\▿g_i\left(X^{\′}\right)+\mathrm{\Σ}{\mathrm{\μ}}_j\▿h_j\left(X^{\′}\right)=0,$ Expression formula 5

g _i(X ') 〉=0 expression formula 6

h _j(X ')=0, expression formula 7

λ _ig _i(X ')=0, expression formula 8

λ _i≥0。Expression formula 9

At distributing antenna system, based on the optimizing process of the self adaptation handoff threshold value of average BER constraint be: satisfy target BER (BER at average BER _Obj) condition under, optimize adaptive threshold by the average SE of maximization system.Corresponding optimization aim problem can be expressed as

$\max \left(\stackrel{\‾}{\mathrm{SE}}\right),$ Expression formula 10

$s.t.\stackrel{\‾}{\mathrm{BER}}\≤{\mathrm{BER}}_{\mathrm{obj}},$ Expression formula 11

γ _i≥0。Expression formula 12

We can obtain by said system optimization aim, constraints and KKT condition

$f\left(\mathrm{\γ}\right)=\stackrel{\‾}{\mathrm{SE}}=\underset{n=1}{\overset{N}{\mathrm{\Σ}}}{b}_n{\∫}_{{\mathrm{\γ}}_n}^{{\mathrm{\γ}}_{n+1}}{f}_{\mathrm{\γ}}\left(\mathrm{\γ}\right)\mathrm{d\γ},$ Expression formula 13

g _i(γ)=γ _i〉=0, expression formula 14

$h_j\left(\mathrm{\γ}\right)=\stackrel{\‾}{\mathrm{SE}}\·{\mathrm{BER}}_{\mathrm{obj}}-\underset{n=1}{\overset{N}{\mathrm{\Σ}}}{b}_n\·{\mathrm{BER}}_{M_n}=0,$ Expression formula 15

Wherein: γ=[γ ₁, γ ₂...., γ _N] be the threshold value vector.

So utilize the KKT condition can obtain following equation:

$F\left(\mathrm{\γ}\right)=\▿\stackrel{\‾}{\mathrm{SE}}\left({\mathrm{\γ}}^{\′}\right)+\▿[\mathrm{\μ}\·(\stackrel{\‾}{\mathrm{SE}}\left(\mathrm{\γ}\right)\·{\mathrm{BER}}_{\mathrm{obj}}-\underset{n=1}{\overset{N}{\mathrm{\Σ}}}{b}_n\·{\mathrm{BER}}_{M_n})]=0,$ Expression formula 16

(3) according to the optimization aim function in the described step (2), utilize the Newton method that it is found the solution.

The Newton method is found the solution nonlinear equation F (x)=0, is the linearizing a kind of approximation method of nonlinear equation.Utilize f (x) at x ₀Carry out Taylor series expansion near the point, get the approximate equation of its linear segment, then have as Equation f (x)=0

F (x ₀)+f ' (x ₀) (x-x ₀)=0 expression formula 17

If f ' is (x ₀) ≠ 0, then its separate into

x ₁=x ₀-f (x ₀)/f ' is expression formula 18 (x)

The present invention is further illustrated below in conjunction with drawings and Examples.

Description of drawings

Fig. 1 is the distributing antenna system structural representation

Fig. 2 is a distributing antenna system Adaptive Modulation schematic diagram

Fig. 3 is the AM spectrum efficiency of distributing antenna system based on average BER constraints

Fig. 4 is that distributing antenna system is based on the average BER of the AM of average BER constraints

Embodiment

Distributing antenna system AM method and corresponding adaptive threshold optimized Algorithm that the present invention proposes verify by the Matlab platform, from simulation result as can be seen this scheme can obtain SE and greatly improve.Provide concrete technical application scheme below:

(1) the native system center cell selects a remote antenna to come transmission signals, promptly selects maximum that root RA of output signal-to-noise ratio (SNR) _iCome transmission signals.Sending criterion can be by following model representation:

γ=max{ γ ₁..., γ _NExpression formula 19

Because the distance between each RA is bigger, so can think γ ₁..., γ _NBetween separate.Utilize expression formula 1 can obtain the cumulative distribution function of γ:

Expression formula 20

By to the following formula differentiate, the PDF that we can obtain γ is:

$f_{\mathrm{\γ}}\left(\mathrm{\γ}\right)=\underset{i=1}{\overset{N}{\mathrm{\Σ}}}\left[f_{{\mathrm{\γ}}_i}\left(\mathrm{\γ}\right)\underset{k=1,k\≠i}{\overset{N}{\mathrm{\Σ}}}{F}_{{\mathrm{\γ}}_k}\left(\mathrm{\γ}\right)\right]$ Expression formula 21

(2) average SE and average BER and thresholding optimal design in the distributing antenna system

According to the analysis of front, by mathematical derivation, the spectrum efficiency expression formula in the time of can obtaining distributing antenna system employing dispersion ratio adaptive modulation scheme:

$\stackrel{\‾}{\mathrm{SE}}=\underset{n=1}{\overset{N-1}{\mathrm{\Σ}}}{b}_n\·{\∫}_{{\mathrm{\γ}}_n}^{{\mathrm{\γ}}_{n+1}}{f}_{\mathrm{\γ}}\left(\mathrm{\γ}\right)\mathrm{d\γ}=\underset{n=0}{\overset{N-1}{\mathrm{\Σ}}}{b}_n\·[F_{\mathrm{\γ}}\left({\mathrm{\γ}}_{n+1}\right)-F_{\mathrm{\γ}}\left({\mathrm{\γ}}_n\right)]$ Expression formula 22

The average BER of system can accurately be calculated by the ratio of average errored bit number and overall average transmitted bit number, promptly by following formula definition:

$\stackrel{\‾}{\mathrm{BER}}=\underset{n=0}{\overset{N-1}{\mathrm{\Σ}}}{b}_n\·{\mathrm{BER}}_{M_n}/\stackrel{\‾}{\mathrm{SE}}$ Expression formula 23

BER wherein _nFor system is in n interval average BER:

${\mathrm{BER}}_{M_n}=\underset{m}{\mathrm{\Σ}}{\mathrm{\α}}_m(\mathrm{erfc}\left(\sqrt{{\mathrm{\β}}_m\mathrm{\γ}}\right)F_r\left(\mathrm{\γ}\right){|}_{{\mathrm{\γ}}_n}^{{\mathrm{\γ}}_{n=1}}+\sqrt{\frac{{\mathrm{\β}}_m}{\mathrm{\π}}}{\∫}_{{\mathrm{\γ}}_n}^{{\mathrm{\γ}}_{n+1}}\frac{\exp (-{\mathrm{\β}}_m\mathrm{\γ})}{\sqrt{\mathrm{\γ}}}{F}_r\left(\mathrm{\γ}\right)\mathrm{d\γ})$ Expression formula 24

Because there is integration in above-mentioned expression formula, so fail to provide the closed expression of BER.For this reason, the present invention adopts the Newton-Cotes numerical integration to simplify the calculating of above-mentioned expression formula.The Newton-Cotes numerical integration is that definite integral is uniformly-spaced got basic point and utilized weight coefficient to sue for peace and represent in integral domain, can obtain following closed expression thus:

${\mathrm{BER}}_{M_n}\≅\underset{m}{\mathrm{\Σ}}{\mathrm{\α}}_m(\mathrm{erfc}\left(\sqrt{{\mathrm{\β}}_m\mathrm{\γ}}\right)F_r\left(\mathrm{\γ}\right){|}_{{\mathrm{\γ}}_n}^{{\mathrm{\γ}}_{n=1}}+\sqrt{\frac{{\mathrm{\β}}_m}{\mathrm{\π}}}\·\mathrm{Ah}\underset{j=0}{\overset{U}{\mathrm{\Σ}}}{W}_j\frac{\exp (-{\mathrm{\β}}_m{\mathrm{\γ}}_j)}{\sqrt{\mathrm{\γ}}}{F}_r\left({\mathrm{\γ}}_j\right))$ Expression formula 25

A and W in the formula _jBe respectively constant and the weight coefficient corresponding, and h is the equally spaced length of integral domain with U

Above-mentioned expression formula 22 and 25 SE and BER Performance Evaluations for system provide effective calculation.

(3) the adaptive threshold optimal design that retrains based on average BER

Expression formula 16 is launched, to variable γ _nCan obtain following general formula

$F\left({\mathrm{\γ}}_n\right)=(1+\mathrm{\μ}\·{\mathrm{BER}}_{\mathrm{obj}})(b_{n-1}-b_n)f_{\mathrm{\γ}}\left({\mathrm{\γ}}_n\right)$

Expression formula 26

$+\mathrm{\μ}[b_n{\mathrm{BER}}_{M_n}\left({\mathrm{\γ}}_n\right)-b_{n-1}{\mathrm{BER}}_{M_{n-1}}\left({\mathrm{\γ}}_n\right)]f_{\mathrm{\γ}}\left({\mathrm{\γ}}_n\right)$

Utilize Newton ' the s method shown in the expression formula 18 to come expression formula 26 is found the solution, can obtain following relational expression

$\left\{\begin{array}{c}{x}^{(k+1)}=x^{\left(k\right)}+{\mathrm{\β}}^{\left(k\right)}\\ \mathrm{DF}\left(x^{\left(k\right)}\right)\·{\mathrm{\β}}^{\left(k\right)}=-F\left(x^{\left(k\right)}\right)\end{array}\right.,$ Expression formula 27

β in the formula ^(k)And x ^(k)=[γ ₁, γ ₂..., γ _N, μ] and be respectively middle vector and the value of finding the solution vector that the k time iteration obtains, F (x ^(k)) and its Jacobi matrix D F (x ^(k)) be expressed as respectively

$F\left(x^{\left(k\right)}\right)=\left[\begin{array}{c}F\left({\mathrm{\γ}}_1\right)\\ F\left({\mathrm{\γ}}_2\right)\\ M\\ F\left({\mathrm{\γ}}_N\right)\\ F\left(\mathrm{\μ}\right)\end{array}\right]=\left[\begin{array}{c}-(1+\mathrm{\μ}\·{\mathrm{BER}}_{\mathrm{obj}})f_{\mathrm{\γ}}\left({\mathrm{\γ}}_1\right)+\mathrm{\μ}\left[{\mathrm{BER}}_{M_1}\left({\mathrm{\γ}}_1\right)\right]f_{\mathrm{\γ}}\left({\mathrm{\γ}}_1\right)\\ -(1+\mathrm{\μ}\·{\mathrm{BER}}_{\mathrm{obj}})f_{\mathrm{\γ}}\left({\mathrm{\γ}}_2\right)\\ +\mathrm{\μ}[2{\mathrm{BER}}_{M_2}\left({\mathrm{\γ}}_2\right)-{\mathrm{BER}}_{M_1}\left({\mathrm{\γ}}_2\right)]f_{\mathrm{\γ}}\left({\mathrm{\γ}}_2\right)\\ M\\ (1+\mathrm{\μ}\·{\mathrm{BER}}_{\mathrm{obj}})(b_{N-1}-b_N)f_{\mathrm{\γ}}\left({\mathrm{\γ}}_N\right)\\ +\mathrm{\μ}[b_N{\mathrm{BER}}_{M_N}\left({\mathrm{\γ}}_N\right)-b_{N-1}{\mathrm{BER}}_{M_{N-1}}\left({\mathrm{\γ}}_N\right)]f_{\mathrm{\γ}}\left({\mathrm{\γ}}_N\right)\\ \stackrel{\‾}{\mathrm{SE}}\left(\mathrm{\γ}\right)\·{\mathrm{BER}}_{\mathrm{obj}}-\underset{n=1}{\overset{N}{\mathrm{\Σ}}}{b}_n{\mathrm{BER}}_n\end{array}\right],$ Expression formula 28

$\mathrm{DF}\left({\mathrm{\γ}}^{\left(k\right)}\right)=\left[\begin{array}{c}{\▿}^T{F}_1\left(x^{\left(k\right)}\right)\\ {\▿}^T{F}_2\left(x^{\left(k\right)}\right)\\ M\\ {\▿}^T{F}_{N-1}\left(x^{\left(k\right)}\right)\\ {\▿}^T{F}_N\left(x^{\left(k\right)}\right)\\ {\▿}^T{F}_{N+1}\left(x^{\left(k\right)}\right)\end{array}\right]=\left[\begin{array}{ccccc}{\∂F}_1/{\∂\mathrm{\γ}}_1^{\left(k\right)}& {\∂F}_1/{\∂\mathrm{\γ}}_2^{\left(k\right)}& K& \∂F_1/{\∂\mathrm{\γ}}_N^{\left(k\right)}& \∂F_1/\∂{\mathrm{\μ}}^{\left(k\right)}\\ \∂F_2/{\∂\mathrm{\γ}}_1^{\left(k\right)}& \∂F_2/{\∂\mathrm{\γ}}_2^{\left(k\right)}& K& \∂F_2/\∂{\mathrm{\γ}}_N^{\left(k\right)}& \∂F_2/\∂{\mathrm{\μ}}^{\left(k\right)}\\ M& M& O& M& M\\ \∂F_{N-1}/\∂{\mathrm{\γ}}_1^{\left(k\right)}& \∂F_{N-1}/\∂{\mathrm{\γ}}_2^{\left(k\right)}& L& \∂F_{N-1}/\∂{\mathrm{\γ}}_N^{\left(k\right)}& \∂F_{N-1}/\∂{\mathrm{\μ}}^{\left(k\right)}\\ \∂F_N/\∂{\mathrm{\γ}}_1^{\left(k\right)}& \∂F_N/\∂{\mathrm{\γ}}_2^{\left(k\right)}& L& \∂F_N/\∂{\mathrm{\γ}}_N^{\left(k\right)}& \∂F_N/\∂{\mathrm{\μ}}^{\left(k\right)}\\ \∂F_{N+1}/\∂{\mathrm{\γ}}_1^{\left(k\right)}& \∂F_{N+1}/\∂{\mathrm{\γ}}_2^{\left(k\right)}& K& \∂F_{N+1}/\∂{\mathrm{\γ}}_N^{\left(k\right)}& \∂F_{N+1}/\∂{\mathrm{\μ}}^{\left(k\right)}\end{array}\right].$ Expression formula 29

When expression formula 27 is found the solution, need choose signification starter; Through theory analysis, the adaptive threshold value after this is optimized should be less than the fixed gate limit value, so the iteration initial value outline of choosing here is less than fixing AM threshold value.

By above-mentioned analysis, based on threshold value and the average signal-to-noise ratio under the average BER constraints

Relevant, along with

Increase, threshold value will reduce accordingly.When

When being increased to a certain critical value, all threshold values will equal zero, and promptly can have the threshold value of a forfeiture Adaptive Modulation

When

The time, system only adopts high-order modulating transmission data, and

Can be by following formula

${\∫}_0^{\∞}\mathrm{BE}{R}_{M_N}\left(\mathrm{\γ}\right)f_{\mathrm{\γ}}\left(\mathrm{\γ}\right)\mathrm{d\γ}={\mathrm{BER}}_{\mathrm{obj}}$ Expression formula 30

Calculate.

The present invention proposes in the distributing antenna system self-adaptive modulation method and the corresponding optimum handoff threshold optimized Algorithm based on average BER constraint.The advantage of extracting method and algorithm in order to embody utilizes the Matlab platform to carry out emulation testing.Adopt 4 kinds of modulation systems such as BPSK, 4QAM, 16QAM and 64QAM for dispersion ratio AM.Accompanying drawing 3, accompanying drawing 4 have provided based on the adaptive modulation scheme spectrum efficiency of optimum thresholding and average BER performance.

Accompanying drawing 3 has provided distributing antenna system AM scheme respectively based on " KKT condition optimizing thresholding " and " fixed threshold " the spectrum efficiency figure of system, wherein receiving terminal antenna number N down _r=2.As can be seen from the figure, in the effective signal-to-noise ratio scope

Most of zone in, improve significantly with respect to the system ASE of instantaneous BER constraints (fixed threshold) based on the system ASE of average BER constraints (optimization thresholding); But, under the relatively poor and good situation of channel condition, improve not obvious, this is because when low signal-to-noise ratio, can only adopt the low-order-modulated mode, and then cause the transmission rate of two kinds of systems under the constraints all very low, and the both is tending towards adopting high-order modulating when high s/n ratio, and promptly transmission rate is tending towards maximum.

Accompanying drawing 4 has provided distributing antenna system AM scheme based on average BER under the different B ER constraints, receiving terminal antenna number N _r=2; As can be seen from the figure, can satisfy quality of service requirement (BER based on the average BER of the system of average BER constraints _Obj=10 ^-3), and the fixed threshold method is because real-time strict demand makes the BER performance of system be better than the BER performance of optimizing gate method.But by accompanying drawing 3 as can be known, the system self-adaption modulation scheme is to exchange the lifting of BER performance for to sacrifice certain SE under this fixed threshold, and this part the BER performance that promotes is a kind of waste when satisfying the condition of system QoS, thereby causes the loss of spectrum efficiency; When

Reach

After, it is maximum that the SE performance reaches, and average BER curve begins to descend and overlaps with the BER curve of 64QAM, shows that thresholding optimization reaches optimum, has verified the threshold value that has forfeiture self adaptation handoff threshold in the expression formula 30 really, promptly

The time, Adaptive Modulation is no longer valid, and system will adopt flank speed 64QAM to transmit.

This shows the superiority of this programme, it has guaranteed the accuracy of the average BER of system under the situation of having avoided loss of spectral efficiency, and has effectively reduced amount of calculation.

The content that is not described in detail in the present patent application book belongs to this area professional and technical personnel's known prior art.

Claims (4)

1. distributed multi-antenna system (DAS) is characterized in that comprising that based on the adaptive modulation scheme design of average BER constraint step is as follows:

(1) sets up distributed multi-antenna service system model and channel model.

(2) utilize Karush-Kuhn-Tucker (KKT) optimal conditions, be subject to target BER according to the average BER of system, maximization system's average spectral efficiency (ase) (SE) is optimized the self adaptation handoff threshold, sets up corresponding optimization aim function.

(3) utilize the Newton interative computation that above-mentioned optimization aim function is found the solution, find out system optimal self adaptation switching threshold

(4) in conjunction with DAS system channel characteristic, the optimum gate limit value of instantaneous BER formula and acquisition provides Adaptive Modulation SE and average BER expression formula based on average BER constraint, for the systematic function assessment provides computational methods.

2. distributed MIMO Adaptive Modulation algorithm design scheme according to claim 1 is characterized in that described step (1) comprising:

(1a) among the DAS because each spaced antenna is different apart from the distance of receiving terminal, so must consider the different large scale slow fading that each remote antenna is experienced to the receiving terminal antenna, be this Channel Modeling a channel that comprises shadow fading, path loss and Rayleigh rapid fading here.

(1b) according to described wireless distributed multiaerial system of step (1a) and channel model, the probability density function (PDF) of the instantaneous received signal to noise ratio γ of the system that solves.

3. the foundation of optimization aim function according to claim 2 is characterized in that described step (2) comprising:

(2a) according to the PDF of the received signal to noise ratio γ of system that obtains in the step (1b), provide spectrum efficiency and average Bit Error Rate Computation formula that the DAS adaptive modulation system contains handoff threshold.

(2b) set up the optimization aim function according to optimization aim, constraints and the KKT condition of system, this function is relevant with average BER expression formula with the SE of system, and average SE and average BER are the functions about adaptive threshold.

4. according to claim 3 and 4 described DAS self-adaptive modulation method designs, it is characterized in that described step (3) comprising:

(3a) after the optimization problem modeling described in the step (2b), need find the solution this non-linear objective function, the self adaptation handoff threshold after just being optimized by iterative search.Therefore, propose the Newton method and find the solution this optimization switching threshold.

Need choose effective initial value when (3b) this Newton method is found the solution, through theory analysis, the adaptive threshold value after this is optimized is less than the fixed gate limit value under the instantaneous BER constraints, so the initial value of interative computation is less than fixedly AM threshold value.

(3c) based on the optimum thresholding that obtains, adopt newton-Ke Cifa, provide SE and the average BER closed expression of distributing antenna system AM, for the SE of system and BER Performance Evaluation provide direct calculation method.

(3d) utilize the designed self-adaptive modulation method performance of Matlab emulation platform test, its result verification the validity of institute's extracting method.Promptly put forward the AM method and can satisfy target BER requirement, have significant SE to improve than the AM method that retrains based on instantaneous BER usually.And the SE that is given and BER closed expression also provide effective ways for the Systems Theory assessment.

Images