CN103812629A - Resource allocation method under high-speed rail train-mounted base station communication architecture - Google Patents

Abstract

The invention relates to a resource allocation method under high-speed rail train-mounted base station communication architecture. The resource allocation method aims at the train-mounted base station communication architecture, and is characterized by maximizing capacity of a communication system through resource allocation under an OFDMA (orthogonal frequency division multiple access) modulation mode. The resource allocation method under the high-speed rail train-mounted base station communication architecture mainly includes: establishing a capacity function under the high-speed rail train-mounted base station communication architecture; allocating subcarriers to a plurality of users; achieving subcarrier pairing for a base station end and a train-mounted mobile base station end; respectively allocating subcarrier power to the base station end and the train-mounted base station end. The resource allocation method under the high-speed rail train-mounted base station communication architecture is of great significance in system performance optimization under high-speed rail scene communication architecture.

Description

Resource allocation methods under a kind of high ferro cell on wheels communication construction

Technical field:

The present invention relates to wireless mobile telecommunication technology field, particularly relate to the resource allocation methods under a kind of high ferro cell on wheels communication construction

Background technology:

As a kind of quick and easy, green public transport, speed is worldwide fast-developing and obtained achievement highly visible at the high-speed railway of 300km.Along with the development of high-speed railway, the Wireless Data Transmission demand on high-speed railway is also more and more higher.On the one hand, ever-increasing Railway safety control informational needs transmits between car ground; On the other hand, passenger wishes to enjoy on the road real-time information service, for example, check and accept mail, download file, video conference etc.According to statistics, the message transmission rate demand on train is 37.5Mbps, and will reach 0.5-5Gpbs future.

Meanwhile, the 4th third-generation mobile communication (LTE-A) is by standardization.Wherein, OFDM access (Orthogonal frequency division multiple access---OFDMA) is owing to there being good anti-frequency selective fading ability, and be easy to carry out resource distribution, be chosen as the downlink transmission technology of LTE-A.But OFDM is very sensitive for time varying channel, the Doppler being produced by high-speed mobile can destroy the orthogonality of intercarrier, causes producing inter-carrier interference (inter-carrier interference---ICI).Along with the increase of translational speed, inter-carrier interference is more and more obvious, and can reduce systematic function.

Under high ferro communication scenes, people have proposed various communication constructions.Wherein, adopt the network architecture of vehicle-mounted mobile base station to be admitted widely, as shown in Figure 1.Be positioned at the mobile base station (mobile base station---MBS) of roof will be from ground the data retransmission that receives of base station (base station---BS) to the subscriber equipment of interior (user equipment---UE).Compared with traditional communication construction, the communication construction of this " double bounce " has the following advantages: first can avoid the penetration loss that brought by compartment.In addition, ground base station only need with train on a communication of mobile terminal, rather than all users of interior, reduce the communication cost between car ground, and have reduced the complexity that switch community.Finally, MBS can for example, carry out improved system performance by signal processing (: power division, allocation of carriers, coding etc.).

By above-mentioned analysis, the bottleneck that can see vehicle-mounted base communication construction under high ferro scene is BS-MBS link, and the ICI being produced by Doppler in this link will reduce systematic function.To distribute to come improved system performance by resource in the present invention.

Summary of the invention

For above the deficiencies in the prior art, the present invention, under the communication construction of high ferro cell on wheels, take maximized system capacity as target function, proposes the resource allocation methods under a kind of high ferro cell on wheels communication construction.

Technical scheme of the present invention is as follows:

A resource allocation methods under high ferro cell on wheels communication construction, said method comprising the steps of:

Step 201: set up the system model under high ferro cell on wheels communication construction, calculating target function.

Step 202: the subcarrier in OFDM is carried out to resource blocking.The length of Resource Block is relevant with the relativity of time domain of channel, meets Ε [h (l, w1) h ^h(l, w2)]>=ε.

Step 203: computational resource block length Ns.

Step 204: the length T c that the poor search of user resources is set.

Step 205: enter the search that user resources piece is distributed, determine the channel response of MBS-UE.

Step 206: the power of initialization BS and MBS end.

Step 207: the initial value l that loop optimization is set _p=1.

Step 208: calculate the SINR of BS end and the SNR of MBS end.

Step 209: Resource Block pairing.

Step 210: the internal subcarrier of Resource Block matches.

Step 211: optimize BS end and MBS end power.

Step 212: judge whether optimizing power restrains, if not, return to step 208, if so, enter step 213.

Step 213: if step 212 judges BS end and MBS end power convergence, power output, and return to step 204.

Step 214: the relatively power system capacity under different user resource allocation condition, power system capacity maximum be the resource optimization condition obtaining.

Described step 201, BS obeys respectively different power constraint P from the power at MBS place _band P _r.

Described step 201, MBS only carries out the redistribution of power on each subcarrier, there is no decoding and at encoding function.

The beneficial effect that the present invention produces: under high ferro scene communication construction, by the subcarrier of BS end and MBS end is distributed to multiple users, realize subcarrier pairing and BS end and the MBS end allocation of subcarriers power of base station end and vehicle-mounted mobile base station end, the Doppler of antagonism BS-MBS link disturbs, improve power system capacity, optimization system performance.

Accompanying drawing explanation

Fig. 1 illustrates vehicle-mounted mobile base station communication construction figure under high ferro scene;

Fig. 2 is the resource allocation flow figure illustrating under high ferro vehicle-mounted mobile base station communication framework;

Fig. 3 illustrates in the embodiment of the present invention, under different normalization Doppler conditions, adopts resource allocation algorithm systematic function of the present invention to improve comparison diagram;

Fig. 4 illustrates in the embodiment of the present invention, under different average SNR conditions, and while adopting resource allocation algorithm of the present invention, the impact of different resource block length on systematic function.

Embodiment:

For above-mentioned purpose of the present invention, feature and advantage can be become apparent more, below in conjunction with the drawings and specific embodiments, the application is described in further detail.

With reference to accompanying drawing 2, show the resource allocation methods schematic flow sheet under high ferro cell on wheels communication construction of the present invention, comprising:

Step 201: set up system model, calculating target function.

OFDMA system under a high ferro cell on wheels communication construction, contains N subcarrier, K user.Ground BS is by the MBS of car body top and several telex networks of interior.Suppose in continuous double bounce, the instantaneous gain of channel remains unchanged.In the first jumping, BS sends signal to MBS, considers allocation of carriers and power division in this process.In the second jumping, MBS redistributes carrier wave and power, and signal is transmitted to K user.With SP (k, m, n) represent the first subcarrier m in jumping (m=1,2 ..., N) subcarrier n in jumping with second (n=1,2 ..., N) pairing, simultaneously subcarrier pair (m, n) is distributed to user.The power of subcarrier pair SP (k, m, n) comprises

with

represent that respectively the BS that distributes to k user holds the power of m subcarrier and MBS to hold the power of n subcarrier.We further define

with

the channel coefficients of subcarrier pair SP (k, m, n) at BS-MBS link and MBS-UE link.The ratio of corresponding channel coefficients and interference noise is respectively

wherein with

respectively the variance of additive white Gaussian noise in two-hop link,

it is the interference power that ICI produces.

Consider

accurate expression formula comparatively complicated and not directly perceived, it is as follows that we utilize the statistical average performance number of ICI to calculate (list of references " Wang; Tiejun; et al. " Performance degradation of OFDM systems due to Doppler spreading. " Wireless Communications, IEEE Transactions on5.6 (2006): 1422-1432. ") expression formula

${\mathrm{\&sigma;}}_{\mathrm{<figure-callout\; id="2"\; label="ICI"\; filenames="HDA0000467381200000032.png"\; state="\{\{state\}\}">ICI</figure-callout>}}^2=E\left[{\left|\underset{j\&NotEqual;m}{\overset{N}{\underset{j=1}{\mathrm{\&Sigma;}}}}{\mathrm{\&alpha;}}_j\sqrt{p_j^{\left(1\right)}}\right|}^2\right]=\frac{{\left({\mathrm{NT}}_s{f}_d\right)}^2}{2}\underset{j\&NotEqual;m}{\overset{N}{\underset{j=1}{\mathrm{\&Sigma;}}}}\frac{p_j^{\left(1\right)}}{{(j-m)}^2}---\left(1\right)$

Wherein f _dfor maximum Doppler, T _sfor the systematic sampling cycle.The OFDMA power system capacity that arrives user k from BS by MBS by subcarrier pair can be expressed as

$\begin{array}{c}{C}_{m,n}^k=\frac{1}{2}{\log }_2(1+\frac{r_{m,n,k}^{\left(1\right)}{p}_{m,n,k}^{\left(1\right)}{r}_{m,n,k}^{\left(2\right)}{p}_{m,n,k}^{\left(2\right)}}{1+r_{m,n,k}^{\left(1\right)}{p}_{m,n,k}^{\left(1\right)}+r_{m,n,k}^{\left(2\right)}{p}_{m,n,k}^{\left(2\right)}})\\ =\frac{1}{2}{\log }_2(1+\frac{p_{m,n,k}^{\left(1\right)}{p}_{m,n,k}^{\left(2\right)}}{(a_{m,n,k}+{\mathrm{\&rho;}}_{m,n,k}\underset{j\&NotEqual;m}{\overset{N}{\underset{j=1}{\mathrm{\&Sigma;}}}}\frac{p_{j,i,u}^{\left(1\right)}}{{(j-m)}^2})\&CenterDot;b_{m,n,k}+b_{m,n,k}{p}_{m,n,k}^{\left(1\right)}+(a_{m,n,k}+{\mathrm{\&rho;}}_{m,n,k}\underset{j\&NotEqual;m}{\overset{N}{\underset{j=1}{\mathrm{\&Sigma;}}}}\frac{p_{j,i,u}^{\left(1\right)}}{{(j-m)}^2})p_{m,n,k}^{\left(2\right)}})\end{array}---\left(2\right)$

Wherein $a_{m,n,k}=\frac{{\mathrm{\&sigma;}}_{\mathrm{<figure-callout\; id="2"\; label="B"\; filenames="HDA0000467381200000032.png"\; state="\{\{state\}\}">B</figure-callout>}}^2}{{\left|h_{m,n,\mathrm{<figure-callout\; id="1"\; label="k"\; filenames="HDA0000467381200000021.png,HDA0000467381200000032.png"\; state="\{\{state\}\}">k</figure-callout>}}^1\right|}^2},b_{m,n,k}=\frac{{\mathrm{\&sigma;}}_R^2}{{\left|h_{m,n,k}^{\left(2\right)}\right|}^2}$ And ${\mathrm{\&rho;}}_{m,n,k}=\frac{{\left({\mathrm{NT}}_s{f}_d\right)}^2}{2{\left|h_{m,n,\mathrm{<figure-callout\; id="1"\; label="k"\; filenames="HDA0000467381200000021.png,HDA0000467381200000032.png"\; state="\{\{state\}\}">k</figure-callout>}}^1\right|}^2}.$

Definition subcarrier pairing ginseng φ _m,n{ 0,1} works as φ to ∈ _m,nbe 1 to be to represent that the subcarrier m of BS end and the subcarrier n of MBS end match, otherwise φ _m,nbe 0.Further definition ω _{k, mn}for user assignment parameter, wherein ω _{k, mn}be that 1 expression subcarrier pair (m, n) is distributed to user k, otherwise be 0.So total power system capacity optimization problem can be expressed as

$C=\underset{p_{m,n,k}^{\left(1\right)}{p}_{m,n,k}^{\left(2\right)},\mathrm{SP}(m,n,k)}{\max }\left(\underset{k=1}{\overset{k}{\mathrm{\&Sigma;}}}{\mathrm{\&omega;}}_{k,\mathrm{mn}}\underset{m=1}{\overset{N}{\mathrm{\&Sigma;}}}\underset{n=1}{\overset{N}{\mathrm{\&Sigma;}}}{\mathrm{\&phi;}}_{m,n}{C}_{m,n}^k\right)---\left(3\right)$

Obey

$C1:\underset{m=1}{\overset{N}{\mathrm{\&Sigma;}}}{\mathrm{\&phi;}}_{m,n}=1,\&ForAll;n$

$C2:\underset{m=1}{\overset{N}{\mathrm{\&Sigma;}}}{\mathrm{\&phi;}}_{m,n}=1,\&ForAll;m$

$C3:\underset{k=1}{\overset{K}{\mathrm{\&Sigma;}}}{\mathrm{\&omega;}}_{k,\mathrm{mn}}=1,\&ForAll;m,n,\mathrm{\&omega;}$

$C4:\underset{k=1}{\overset{K}{\mathrm{\&Sigma;}}}\underset{m=1}{\overset{N}{\mathrm{\&Sigma;}}}\underset{N=1}{\overset{N}{\mathrm{\&Sigma;}}}{p}_{m,n,k}^{\left(1\right)}\&le;P_B$

$C5:\underset{k=1}{\overset{K}{\mathrm{\&Sigma;}}}\underset{m=1}{\overset{N}{\mathrm{\&Sigma;}}}\underset{N=1}{\overset{N}{\mathrm{\&Sigma;}}}{p}_{m,n,k}^{\left(2\right)}\&le;P_R$

$C6:p_{m,n,k}^{\left(1\right)}\&GreaterEqual;0,p_{m,n,k}^{\left(2\right)}\&GreaterEqual;0$

C7:φ _m,n∈[0,1],ω _k∈[0,1]

Wherein P _band P _rrepresent respectively the total transmitting power in BS place and MBS place.Constraints C1 and C2 guarantee that m subcarrier in the first jumping can only have n subcarrier pairing in the second jumping.Constraints C3 guarantees that each subcarrier pair can only distribute to a user.C4 and C5 are respectively at the total power constraint in BS place and MBS place.

Step 202: the subcarrier in OFDM is carried out to resource blocking.

In order to obtain maximum power system capacity, we adopt the method for poor search that subcarrier is distributed to user.In the time that N is larger, for example N=16, K=4, the value of TK is 63063000, therefore allocation of carriers becomes very complicated.In order to reduce system redundancy and the computation complexity of single sub-carrier distribution, can consider the correlation of adjacent sub-carrier in ofdm system.Therefore, using continuous subcarrier as a Resource Block, and carry out allocation of carriers take Resource Block as unit and can simplify system complexity.The length of Resource Block is relevant with the relativity of time domain of channel, meets

Ε[h(l,w1)h ^H(l,w2)]≥ε （4）

Wherein h (l, w1) and h (l, w2) represent that respectively channel tap l is respectively at the coefficient of moment w1 and w2, and ε is the degree of correlation parameter in 0 to 1 scope.In the present embodiment, the value that ε is set is 0.99.

Step 203: computational resource block length N _s.

In the embodiment of the present invention, preferential, channel is obeyed Jake ' s model, and the time domain of channel tap coefficient is relevant so can be expressed as:

Ε[h(l,w1)h ^H(l,w2)]=J ₀(2πf _d(w2-w1)T _s) （5）

Wherein

represent first-order bessel function.In the present invention, determine the size of Resource Block according to the relativity of time domain of channel tap, resource block length is N _smeet

Step 204: the poor search length Tc of user resources is set.

In the time having N sub-allocation of carriers to K user, each user assignment

individual subcarrier.The possibility of this method of salary distribution has:

$T_s=W_{N_c}^{N_{\mathrm{cperu}}}{W}_{N_c-N_{\mathrm{cperu}}}^{N_{\mathrm{cperu}}}...W_{N_{\mathrm{cperu}}}^{N_{\mathrm{cperu}}}---\left(7\right)$

Wherein represent to get the combination possibility of r number from R number.The resource block length of supposing continuous subcarrier composition is Ns, and an OFDM symbol can be divided into

individual Resource Block, each CU

individual Resource Block, the possibility of the mode of distributing take Resource Block as unit has.

$T_c=W_{N_c}^{N_{\mathrm{cperu}}}{W}_{N_c-N_{\mathrm{cperu}}}^{N_{\mathrm{cperu}}}...W_{N_{\mathrm{cperu}}}^{N_{\mathrm{cperu}}}---\left(8\right)$

If 16 subcarriers are divided into 4 Resource Block, and distribute to 4 users, always have 24 kinds and distribute possibility.

Step 205: enter the search that user resources piece is distributed, determine the channel response of MBS-UE.

The channel response matrix of supposing K user is $H_k=\left[\begin{array}{c}h\left(1\right)\\ h\left(2\right)\\ \&CenterDot;\\ \&CenterDot;\\ \&CenterDot;\\ h\left(K\right)\end{array}\right],$ Wherein h (k) is k user's channel coefficient matrix, is expressed as h (k)=[h ₁(k), h ₂(k) ..., h _n(k)], h wherein _n(k) represent that MBS is in the channel of user k, the channel coefficients of n subcarrier.

If subcarrier is distributed take Resource Block as unit, K user's channel response matrix is $H_K^c=\left[\begin{array}{c}{h}^c\left(1\right)\\ \&CenterDot;\\ \&CenterDot;\\ \&CenterDot;\\ h^c\left(k\right)\\ \&CenterDot;\\ \&CenterDot;\\ \&CenterDot;\\ h^c\left(K\right)\end{array}\right],$ Wherein $h^c\left(k\right)=\left[h_1^c\left(k\right),...,h_{n_c}^c\left(k\right),...,h_{N_c}^c\left(k\right)\right],$ And $h_{n_c}^c\left(k\right)=[h_{(n_c-1)N_s+1}\left(k\right),h_{(n_c-1)N_s+2}\left(k\right),...,h_{n_c{N}_s}\left(k\right)],$

If N _{c_ID}=[1:N _c], matrix $N_{C\_P}=\left[\begin{array}{c}{N}_1^c\\ N_2^c\\ \&CenterDot;\\ \&CenterDot;\\ \&CenterDot;\\ N_{T_c}^c\end{array}\right]$ A T _c× N _cmatrix, wherein $N_{l_k}^c=[B_1^{t_c},B_2^{t_c},...,B_{N_c}^{t_c}]({\mathrm{<figure-callout\; id="1"\; label="l\; k"\; filenames="HDA0000467381200000021.png,HDA0000467381200000032.png"\; state="\{\{state\}\}">\; <figure-callout\; id="2"\; label="l\; k"\; filenames="HDA0000467381200000032.png"\; state="\{\{state\}\}">l</figure-callout>\; </figure-callout>}}_k\mathrm{1,2},...,T_c)$ About a kind of possibility in (8) permutation and combination.The channel coefficient matrix after the distribution of user resources piece can be expressed as $H_{C\_\max \mathrm{trix}}=\left[\begin{array}{c}{h}^{\left(2\right)}\left(1\right)\\ h^{\left(2\right)}\left(2\right)\\ \&CenterDot;\\ \&CenterDot;\\ \&CenterDot;\\ h^{\left(2\right)}\left(T_c\right)\end{array}\right],$ Wherein $h^{\left(2\right)}\left(l_k\right)=[h_{B_1^{l_k}}^c\left(1\right),...,h_{B_{N_{c\_\mathrm{peru}}}^{l_k}}^c\left(1\right),h_{B_{N_{c\_\mathrm{peru}}+1}^{l_k}}^c\left(2\right),...,h_{B_{2N_{c\_\mathrm{peru}}}^{l_k}}^c\left(2\right),...],$

Step 206: the power of initialization BS and MBS end.

In the present embodiment, the initialization power of BS end is

the initialization power of MBS end is $p_{m,n,k}^{\left(2\right)}\left(0\right)=\frac{P_R}{N},(n=\mathrm{1,2},...,N).$

Step 207: the initial value l that loop optimization is set _p=1.

Step 208: calculate the SINR of BS end and the SNR of MBS end.

By what obtain in step 201

with

can draw, the SINR of BS end is:

$R_m^{\left(1\right)}=\frac{p_{m,n,k}^{\left(1\right)}{\left|h_{m,n,k}^{\left(1\right)}\right|}^2}{{\mathrm{\&sigma;}}_{\mathrm{<figure-callout\; id="2"\; label="B"\; filenames="HDA0000467381200000032.png"\; state="\{\{state\}\}">B</figure-callout>}}^2+{\mathrm{\&sigma;}}_{\mathrm{ICI}}^2}---\left(9\right)$

The SNR of MSB end is:

$R_n^{\left(2\right)}=\frac{p_{m,n,k}^{\left(2\right)}{\left|h_{m,n,k}^{\left(2\right)}\right|}^2}{{\mathrm{\&sigma;}}_R^2}---\left(10\right)$

Step 209: Resource Block pairing.

If user assignment is definite, formula (3) can be expressed as again

$C=\max \underset{m=1}{\overset{N}{\mathrm{\&Sigma;}}}\underset{n=1}{\overset{N}{\mathrm{\&Sigma;}}}{\mathrm{\&phi;}}_{m,n}{C}_{m,n}^k---\left(11\right)$

For shortcut calculation, first hypothesis only has the situation of two subcarriers, and the situation of BS terminal carrier wave and the pairing of MBS terminal carrier wave only has two kinds so, and one is

${\mathrm{<figure-callout\; id="1"\; label="C"\; filenames="HDA0000467381200000021.png,HDA0000467381200000032.png"\; state="\{\{state\}\}">C</figure-callout>}}_1=\frac{1}{2}{\log }_2(1+\frac{R_1^{\left(1\right)}{R}_1^{\left(2\right)}}{1+R_1^{\left(1\right)}+R_1^{\left(2\right)}})(1+\frac{R_2^{\left(1\right)}{R}_2^{\left(2\right)}}{1+R_2^{\left(1\right)}+R_2^{\left(2\right)}})---\left(12\right)$

Another is

${\mathrm{<figure-callout\; id="2"\; label="C"\; filenames="HDA0000467381200000032.png"\; state="\{\{state\}\}">C</figure-callout>}}_2=\frac{1}{2}{\log }_2(1+\frac{R_1^{\left(1\right)}{R}_1^{\left(2\right)}}{1+R_1^{\left(1\right)}+R_1^{\left(2\right)}})(1+\frac{R_2^{\left(1\right)}{R}_1^{\left(2\right)}}{1+R_2^{\left(1\right)}+R_1^{\left(2\right)}})---\left(13\right)$

Can obtain as drawn a conclusion

${\mathrm{<figure-callout\; id="1"\; label="C"\; filenames="HDA0000467381200000021.png,HDA0000467381200000032.png"\; state="\{\{state\}\}">C</figure-callout>}}_1>{\mathrm{<figure-callout\; id="2"\; label="C"\; filenames="HDA0000467381200000032.png"\; state="\{\{state\}\}">C</figure-callout>}}_2\&DoubleLeftRightArrow;R_1^{\left(1\right)}>R_1^{\left(2\right)},R_2^{\left(1\right)}>R_2^{\left(2\right)}---\left(14\right)$

Or

${\mathrm{<figure-callout\; id="1"\; label="C"\; filenames="HDA0000467381200000021.png,HDA0000467381200000032.png"\; state="\{\{state\}\}">C</figure-callout>}}_1>{\mathrm{<figure-callout\; id="2"\; label="C"\; filenames="HDA0000467381200000032.png"\; state="\{\{state\}\}">C</figure-callout>}}_2\&DoubleLeftRightArrow;R_1^{\left(1\right)}<R_1^{\left(2\right)},R_2^{\left(1\right)}<R_2^{\left(2\right)}---\left(15\right)$

In the time having multiple subcarrier, still can adopt induction to draw above-mentioned conclusion.Formula (14) and (15) expression, the subcarrier of BS end and MBS end accesses larger power system capacity according to Signal to Interference plus Noise Ratio and signal interference ratio pairing energy., the subcarrier pairing of the subcarrier of BS end Signal to Interference plus Noise Ratio maximum and MBS end signal interference ratio maximum, maximizing power system capacity.And signal interference ratio is not only relevant with channel coefficients with Signal to Interference plus Noise Ratio, also have sub-carrier power relevant.Therefore in carrying out subcarrier pairing, not only will consider channel coefficients, also should consider sub-carrier power, subcarrier pairing and power division influence each other.Because subcarrier is to distribute to user in the mode of Resource Block, so be first to carry out Resource Block pairing, and then carry out the subcarrier pairing of Resource Block inside carrying out carrier wave pairing.In the present invention, first according to Resource Block channel gain, the Resource Block in double bounce is matched, the SINR of definition BS end Resource Block is

The SNR of MBS end Resource Block is

BS end Resource Block and MBS end Resource Block match by following principle

Suppose

with pairing,

with

pairing

If

Step 210: the subcarrier internal to Resource Block matches.

If resource block pair, BS terminal carrier wave and MBS terminal carrier wave match by following principle:

Suppose

with

pairing,

with

pairing

If

Step 211: optimize BS end and MBS end power.

Under a certain user resources distributive condition, subcarrier pairing is determined, can adopt KKT condition to be optimized distribution to BS end and MBS end power, and the optimizing power that obtains BS end is $p_{m,n,k}^{\left(2\right)}\left(l_p\right)(m=\mathrm{1,2},...,N),$ The optimizing power of MSB end is $p_{m,n,k}^{\left(2\right)}\left(l_p\right)(n=\mathrm{1,2},...,N)$

Step 212: judge whether optimizing power restrains.

I.e. judgement

$\begin{array}{c}{p}_{m,n,k}^{\left(1\right)}\left(l_p\right)-p_{m,n,k}^{\left(1\right)}(l_p-1)\&le;{\mathrm{\&epsiv;}}_B\\ p_{m,n,k}^{\left(2\right)}\left(l_p\right)-p_{m,n,k}^{\left(2\right)}(l_p-1)\&le;{\mathrm{\&epsiv;}}_R\end{array}---\left(18\right)$

Whether set up, in the present embodiment, ε _b=ε _r=10 ^-6.If not, return to step 208, if so, enter step 213.

Step 213: if step 212 judges BS end and MBS end power convergence, output

$\begin{array}{c}{p}_{m,n,k}^{\left(1\right)}\left(l_k\right)=p_{m,n,k}^{\left(1\right)}\left(l_p\right)\\ p_{m,n,k}^{\left(2\right)}\left(l_k\right)-p_{m,n,k}^{\left(2\right)}\left(l_p\right)\end{array}---\left(19\right)$

And calculating power system capacity with this understanding

$C\left(l_k\right)=\underset{m=1}{\overset{N}{\mathrm{\&Sigma;}}}{C}_{m,n}^k---\left(19\right)$

And return to step 204.

Step 214: the relatively power system capacity under different user resource allocation condition, power system capacity maximum be the resource optimization condition obtaining, C=max{C (l _k).

Below in conjunction with Fig. 3, Fig. 4 analyzes the present invention program's beneficial effect.In Computer Simulation of the present invention, consider an OFDMA system, contain 16 subcarriers, BS-MBS link and MBS-UE obey respectively independently Rayleigh fading, and the power constraint of BS end is 40dBm, and the power constraint of MBS end is 5dBm.The average signal-to-noise ratio of definition BS-MBS link is

the average signal-to-noise ratio of MBS-UE link is

in addition, definition normalization Doppler describes the time-varying characteristics of BS-MBS link, and expression formula is f _n=NT _sf _d.

Accompanying drawing 3 illustrates, under different normalization Doppler conditions, the present invention's power system capacity in an embodiment changes and changes, and wherein transverse axis represents normalization Doppler, and the longitudinal axis represents power system capacity."○" represents BS end and MBS terminal carrier power mean allocation, and the power system capacity change curve (EP) while optimization without any subcarrier pairing, " ☆ " represents BS end and MBS terminal carrier power mean allocation, and the power system capacity change curve (EPCP) while carrying out Resource Block pairing, " " represents BS end and MBS terminal carrier power mean allocation, but the power system capacity change curve (EPCPSP) while carrying out the subcarrier pairing in Resource Block pairing and Resource Block, power system capacity change curve (I-OPCPSP) when the power division changing that " △ " expression BS end and MBS end employing the present invention propose and subcarrier match.From figure, can draw, first, along with the increase of normalization Doppler, power system capacity reduces, and illustrates that the Doppler being produced by speed can reduce systematic function.In addition, the subcarrier paring strategy (EPCPSP) in the Resource Block that employing the present invention proposes is compared with the scheme (EPCP) of only carrying out Resource Block pairing, further optimization system performance.Finally, the Resource Allocation Formula (I-OPCPSP) of employing iteration can maximized system capacity.

Accompanying drawing 4 illustrates, under different resource block length, power system capacity changes comparison diagram with the average SNR of BS-MBS link.In this emulation, normalization Doppler is made as 0.1, the signal to noise ratio of MBS-UE link is 20dB, and 16 subcarriers (N=16) are distributed to two users (K=2), considers that 16 subcarriers are divided into the situation of 2 Resource Block (2c), 4 Resource Block (4c) and 8 Resource Block (8c).By Tu Ke get, corresponding time domain channel coefficient correlation is respectively 0.976,0.993 and 0.998.Only exist under the condition of Resource Block pairing, the capacity maximum that channel correlation coefficient is high, the capacity minimum that channel correlation coefficient is low, be 8C EPCP>4C EPCP>2CEPCP, if the pairing of the subcarrier in employing Resource Block (2C EPCPSP, 4C EPCPSP, 8C EPCPSP), channel capacity under three kinds of conditions is more or less the same, and this explanation subcarrier pairing can make up the lower performance loss bringing of channel relevancy.Meanwhile, adopt the resource allocation algorithm (2C I-OPCPSP, 4C I-OPCPSP, 8CI-OPCPSP) of iteration, the capacity of submission system greatly, and capacity under three kinds of conditions is more or less the same.

Obviously; the above embodiment of the present invention is only for example of the present invention is clearly described; and be not the restriction to embodiments of the present invention; for those of ordinary skill in the field; can also make other changes in different forms on the basis of the above description; here cannot give all execution modes exhaustively, everyly belong to apparent variation or the still row in protection scope of the present invention of variation that technical scheme of the present invention extends out.

Claims (2)

1. the resource allocation methods under high ferro cell on wheels communication construction, is characterized in that, said method comprising the steps of:

Step 201: set up the system model under high ferro cell on wheels communication construction, calculate the target function about power system capacity, about the target function of power system capacity be

$C=\underset{p_{m,n,k}^{\left(1\right)}{p}_{m,n,k}^{\left(2\right)},\mathrm{SP}(m,n,k)}{\max }\left(\underset{k=1}{\overset{k}{\mathrm{\&Sigma;}}}{\mathrm{\&omega;}}_{k,\mathrm{mn}}\underset{m=1}{\overset{N}{\mathrm{\&Sigma;}}}\underset{n=1}{\overset{N}{\mathrm{\&Sigma;}}}{\mathrm{\&phi;}}_{m,n}{C}_{m,n}^k\right)---\left(a\right)$ Obey

$C1:\underset{m=1}{\overset{N}{\mathrm{\&Sigma;}}}{\mathrm{\&phi;}}_{m,n}=1,\&ForAll;n,\underset{n=1}{\overset{N}{\mathrm{\&Sigma;}}}{\mathrm{\&phi;}}_{m,n}=1,\&ForAll;m,{\mathrm{\&phi;}}_{m,n}\&Element;[o,1]$

$C2:\underset{k=1}{\overset{K}{\mathrm{\&Sigma;}}}{\mathrm{\&omega;}}_{k,\mathrm{mn}}=1,\&ForAll;m,n,{\mathrm{\&omega;}}_k\&Element;\left[\mathrm{0,1}\right]$

$C3:\underset{k=1}{\overset{K}{\mathrm{\&Sigma;}}}\underset{m=1}{\overset{N}{\mathrm{\&Sigma;}}}\underset{N=1}{\overset{N}{\mathrm{\&Sigma;}}}{p}_{m,n,k}^{\left(1\right)}\&le;P_B,p_{m,n,k}^{\left(1\right)}\&GreaterEqual;0$

$C4:\underset{k=1}{\overset{K}{\mathrm{\&Sigma;}}}\underset{m=1}{\overset{N}{\mathrm{\&Sigma;}}}\underset{N=1}{\overset{N}{\mathrm{\&Sigma;}}}{p}_{m,n,k}^{\left(2\right)}\&le;P_R,p_{m,n,k}^{\left(2\right)}\&GreaterEqual;0$

Wherein, P _band P _rrepresent that respectively BS end and MBS hold total transmitting power; constraints C1 and C2 guarantee that m subcarrier in the first jumping can only have n subcarrier pairing in the second jumping; constraints C3 guarantees that each subcarrier pair can only distribute to a user, and C4 and C5 are respectively that BS end and MBS hold total power constraint;

In expression formula (a)

${\mathrm{\&sigma;}}_{\mathrm{ICI}}^{2}E\left[{\left|\underset{j\&NotEqual;m}{\overset{N}{\underset{j=1}{\mathrm{\&Sigma;}}}}{\mathrm{\&alpha;}}_j\sqrt{p_j^{\left(1\right)}}\right|}^2\right]=\frac{{\left({\mathrm{NT}}_s{f}_d\right)}^2}{2}\underset{j\&NotEqual;m}{\overset{N}{\underset{j=1}{\mathrm{\&Sigma;}}}}\frac{p_j^{\left(1\right)}}{{(j-m)}^2}---\left(b\right)$

Expression formula (b), f _dfor maximum Doppler, T _sfor the systematic sampling cycle,

In expression formula (a),

$\begin{array}{c}{C}_{m,n}^k=\frac{1}{2}{\log }_2(1+\frac{r_{m,n,k}^{\left(1\right)}{p}_{m,n,k}^{\left(1\right)}{r}_{m,n,k}^{\left(2\right)}{p}_{m,n,k}^{\left(2\right)}}{1+r_{m,n,k}^{\left(1\right)}{p}_{m,n,k}^{\left(1\right)}+r_{m,n,k}^{\left(2\right)}{p}_{m,n,k}^{\left(2\right)}})\\ =\frac{1}{2}{\log }_2(1+\frac{p_{m,n,k}^{\left(1\right)}{p}_{m,n,k}^{\left(2\right)}}{(a_{m,n,k}+{\mathrm{\&rho;}}_{m,n,k}\underset{j\&NotEqual;m}{\overset{N}{\underset{j=1}{\mathrm{\&Sigma;}}}}\frac{p_{j,i,u}^{\left(1\right)}}{{(j-m)}^2})\&CenterDot;b_{m,n,k}+b_{m,n,k}{p}_{m,n,k}^{\left(1\right)}+(a_{m,n,k}+{\mathrm{\&rho;}}_{m,n,k}\underset{j\&NotEqual;m}{\overset{N}{\underset{j=1}{\mathrm{\&Sigma;}}}}\frac{p_{j,i,u}^{\left(1\right)}}{{(j-m)}^2})p_{m,n,k}^{\left(2\right)}})\end{array}---\left(c\right)$ In expression formula (c), $a_{m,n,k}=\frac{{\mathrm{\&sigma;}}_B^2}{{\left|h_{m,n,k}^1\right|}^2},b_{m,n,k}=\frac{{\mathrm{\&sigma;}}_R^2}{{\left|h_{m,n,k}^{\left(2\right)}\right|}^2},{\mathrm{\&rho;}}_{m,n,k}=\frac{{\left({\mathrm{NT}}_s{f}_d\right)}^2}{2{\left|h_{m,n,k}^1\right|}^2};$

Step 202: the subcarrier in OFDMA is carried out to resource blocking, and the length of Resource Block is relevant with the relativity of time domain of channel, meets Ε [h (l, w1) h ^h(l, w2)]>=ε, wherein h (l, w1) and h (l, w2) represent respectively BS channel tap l is at the coefficient of moment w1 and w2 in MBS link, ε is the degree of correlation parameter in 0 to 1 scope, ε=0.993;

Step 203: computational resource block length, channel is obeyed Jake ' s model, the length N of Resource Block _smeet

wherein

represent first-order bessel function;

Step 204: the scope l that the poor search of user resources is set _k=1:T _c;

The resource block length of subcarrier composition is N continuously _s, an OFDM symbol can be divided into

individual Resource Block, each CU

individual Resource Block, the possibility of the mode of distributing take Resource Block as unit has

$T_c=W_{N_c}^{N_{\mathrm{cperu}}}{W}_{N_c-N_{\mathrm{cperu}}}^{N_{\mathrm{cperu}}}...W_{N_{\mathrm{cperu}}}^{N_{\mathrm{cperu}}}---\left(e\right)$ Wherein,

represent to get the combination possibility of r number from R number;

Step 205: the search that user resources piece is distributed, determine MBS the channel response matrix of UE link;

K user's channel response matrix is that wherein h (k) is k user's channel coefficient matrix, is expressed as h (k)=[h ₁(k) ..., h _n(k) ..., h _n(k)], wherein hn (k) represents that MBS is in the channel of user k, the channel coefficients of n subcarrier;

If subcarrier is distributed take Resource Block as unit, K user's channel response matrix is $H_K^c=\left[\begin{array}{c}{h}^c\left(1\right)\\ \&CenterDot;\\ \&CenterDot;\\ \&CenterDot;\\ h^c\left(k\right)\\ \&CenterDot;\\ \&CenterDot;\\ \&CenterDot;\\ h^c\left(K\right)\end{array}\right],$ Wherein $h^c\left(k\right)=\left[h_1^c\left(k\right),...,h_{n_c}^c\left(k\right),...,h_{N_c}^c\left(k\right)\right],$ And $h_{n_c}^c\left(k\right)=[h_{(n_c-1)N_s+1}\left(k\right),h_{(n_c-1)N_s+2}\left(k\right),...,h_{n_c{N}_s}\left(k\right)],$

If N _{c_ID}=[1,2 ..., N _c], matrix $I_c=\left[\begin{array}{c}{I}_1\\ \&CenterDot;\\ \&CenterDot;\\ \&CenterDot;\\ I_{l_k}\\ \&CenterDot;\\ \&CenterDot;\\ \&CenterDot;\\ I_{T_c}\end{array}\right]$ A T _c× N _cmatrix, wherein (l _k=1,2 ..., T _c) be N _{c_ID}about a kind of possibility in (e) permutation and combination, the channel response matrix after the distribution of user resources piece can be expressed as $H_{C\_\max \mathrm{trix}}=\left[\begin{array}{c}{h}^{\left(2\right)}\left(1\right)\\ \&CenterDot;\\ \&CenterDot;\\ \&CenterDot;\\ h^{\left(2\right)}\left(l_k\right)\\ \&CenterDot;\\ \&CenterDot;\\ \&CenterDot;\\ h^{\left(2\right)}\left(T_c\right)\end{array}\right]$ , wherein $h^{\left(2\right)}\left(l_k\right)=[h_{B_1^{l_k}}^c\left(1\right),...,h_{B_{N_{c\_\mathrm{peru}}}^{l_k}}^c\left(1\right),h_{B_{N_{c\_\mathrm{peru}}+1}^{l_k}}^c\left(2\right),...,h_{B_{2N_{c\_\mathrm{peru}}}^{l_k}}^c\left(2\right),...],$ ${\left|h_{m,n,k}^{\left(2\right)}\right|}^2=h^{\left(2\right)}\left(l_k\right);$

Step 206: the power of initialization BS and MBS end

Step 207: the initial value lp=1 that BS end and MBS end power cycle are optimized is set;

Step 208: the SINR that calculates BS end is with the SNR of MBS end be $R_n^{\left(2\right)}=\frac{p_{m,n,k}^{\left(2\right)}{\left|h_{m,n,k}^{\left(2\right)}\right|}^2}{{\mathrm{\&sigma;}}_R^2};$

Step 209: Resource Block pairing, first according to Resource Block channel gain, the Resource Block in double bounce is matched, the average SINR of definition BS end Resource Block is

the average SNR of MBS end Resource Block is

bS end Resource Block and MBS end Resource Block match by following principle:

Suppose with

pairing,

with

pairing, if

Step 210: the subcarrier internal to Resource Block matches;

Resource block pair, BS terminal carrier wave and MBS terminal carrier wave match by following principle

Suppose

with pairing,

with pairing, if

Step 211: optimize BS end and MBS end power, adopt KKT condition to carry out iteration optimization distribution to BS end and MBS end power, the optimizing power that obtains BS end is

the optimizing power of MSB end is $p_{m,n,k}^{\left(2\right)}\left(l_p\right)(n=\mathrm{1,2},...,N);$

Step 212: judge whether optimizing power restrains;

I.e. judgement

$\begin{array}{c}{p}_{m,n,k}^{\left(1\right)}\left(l_p\right)-p_{m,n,k}^{\left(1\right)}(l_p-1)\&le;{\mathrm{\&epsiv;}}_B\\ p_{m,n,k}^{\left(2\right)}\left(l_p\right)-p_{m,n,k}^{\left(2\right)}(l_p-1)\&le;{\mathrm{\&epsiv;}}_R\end{array}---\left(h\right)$

Whether set up, preferred, ε _b=ε _r=10 ^-6, if not, return to step 208, if so, enter step 213;

Step 213: if step 212 judges BS end and MBS end power convergence, output

$\begin{array}{c}{p}_{m,n,k}^{\left(1\right)}\left(l_k\right)=p_{m,n,k}^{\left(1\right)}\left(l_p\right)\\ p_{m,n,k}^{\left(2\right)}\left(l_k\right)-p_{m,n,k}^{\left(2\right)}\left(l_p\right)\end{array}---\left(i\right)$

Calculate power system capacity with this understanding

and return to step 204;

Step 214: the relatively power system capacity under different user resource allocation condition, power system capacity maximum be the resource optimization condition obtaining, i.e. C=max{C (l _k).

2. the resource allocation methods under a kind of high ferro cell on wheels communication construction according to claim 1, it is characterized in that, in described step 201, system model is: the OFDMA system under a high ferro cell on wheels communication construction, contain N subcarrier, K user, ground BS is by the MBS of car body top and multiple telex networks of interior, suppose in continuous double bounce, the instantaneous gain of channel remains unchanged, in the first jumping, BS sends signal to MBS, in this process, consider allocation of carriers and power division, in the second jumping, MBS redistributes carrier wave and power, and signal is transmitted to K user, with SP (k, m, n) represent the subcarrier m (m=1 in the first jumping, 2, ..., N) the subcarrier n (n=1 and in the second jumping, 2, ..., N) pairing, subcarrier pair (m simultaneously, n) distribute to user, subcarrier pair SP (k, m, n) power comprises with

represent that respectively the BS that distributes to k user holds the power of m subcarrier and MBS to hold the power of n subcarrier, further defines

with be the channel coefficients of subcarrier pair SP (k, m, n) at BS-MBS link and MBS-UE link, the ratio of corresponding channel coefficients and interference noise is respectively

wherein

with

respectively the variance of additive white Gaussian noise in two-hop link, the interference power that ICI produces, definition subcarrier pairing parameter phi _m,n{ 0,1} works as φ to ∈ _m,nbe 1 to be to represent that the subcarrier m of BS end and the subcarrier n of MBS end match, otherwise φ _m,nbe 0, further define ω _{k, mn}for user assignment parameter, wherein ω _{k, mn}be that 1 expression subcarrier pair (m, n) is distributed to user k, otherwise be 0.3. the resource allocation methods under a kind of high ferro cell on wheels communication construction according to claim 1, is characterized in that, the initialization power of described BS end is

the initialization power of MBS end is $p_{m,n,k}^{\left(2\right)}\left(0\right)=\frac{P_R}{N},(n=\mathrm{1,2},...,N).$

Images