CN102244933A - Multi-service OFDM (Orthogonal Frequency Division Multiplexing) cross-layer dynamic resource distribution method based on evidence theory - Google Patents

Abstract

The invention discloses a multi-service OFDM (Orthogonal Frequency Division Multiplexing) cross-layer dynamic resource distribution method based on an evidence theory in the technical field of wireless communication. The method comprises the following steps of: establishing target sets for dividing data packets according to the scheduling level; setting evaluation indexes; calculating to obtain elementary probability distribution function values through which the data packets under each evaluation index are divided to the target sets according to membership functions through which the data packets under each evaluation index are divided to the target sets; according to the D-S (Dempster-Shafer) evidence theory, carrying out fusion on the elementary probability distribution function value through which the data packets under each evaluation index are divided to the target sets, thus obtaining a fused elementary probability distribution function value; and according to the fused elementary probability distribution function value, determining the scheduling level of the data packets; and scheduling the data packets according to the scheduling level and the fused elementary probability distribution function value. The multi-service OFDM cross-layer dynamic resource distribution method can be used for overcoming the defects of lower precision, reliability and accuracy when the resource scheduling priority is discriminated according to a single evaluation index.

Description

Multi-service OFDM based on evidence theory strides a layer dynamic resource allocation method

Technical field

The invention belongs to wireless communication technology field, relate in particular to a kind of multi-service OFDM and stride a layer dynamic resource allocation method based on evidence theory.

Background technology

Along with the development of radio multimedium business, communication more reliable, higher rate that wireless channel need be supported, and can guarantee multi-user's different business demand.OFDM (Orthogonal Frequency Division Multiplexing, OFDM) owing to anti-intersymbol interference ability by force, frequency domain channel control ability and provide high performance physical layer link ability to receive much concern flexibly.Physical layer is combined and then optimized allocation of resources becomes the key factor that improves systematic function with the upper-layer service demand, therefore studying in the ofdm system multi-service, to stride layer Dynamic Resource Allocation for Multimedia very urgent.In multi-user's resource allocation of differentiated service, traditional distribution method is based on the resource regulating method of delay requirement (EDF method) and based on the resource regulating method (PF method) of speed fairness.Scheduling of resource throughput based on time delay is low excessively, and it is excessive based on the scheduling of resource time delay of speed fairness, two kinds of methods all can not reach desirable distribution effects for the service-user of each feature request equalization of QoS (Quality of Service, service quality).The existing resources distribution method is to adjudicate priority of data packets according to a certain evaluation index mostly, defective of various degrees all at aspects such as accuracy, reliability and practicality.

Summary of the invention

The objective of the invention is to, provide a kind of multi-service OFDM to stride a layer dynamic resource allocation method based on evidence theory, by merging the evaluation index in the QoS component, come the scheduling grade of decision data bag, overcome precision, reliability and the lower defective of accuracy when scheduling of resource priority being differentiated according to the single evaluation index.

To achieve these goals, the technical scheme of the present invention's proposition is that a kind of multi-service OFDM based on evidence theory strides a layer dynamic resource allocation method, it is characterized in that described method comprises:

Step 1:, set up the goal set that is used for the dividing data bag according to the scheduling grade;

Step 2: set evaluation index;

Step 3: be divided into the membership function of goal set according to packet under each evaluation index, calculate the elementary probability partition function value that packet under each evaluation index is divided into goal set;

Step 4: according to the D-S evidence theory, the elementary probability partition function value that packet under each evaluation index is divided into goal set merges, and the packet after obtaining merging is divided into the elementary probability partition function value of goal set;

Step 5: be divided into the elementary probability partition function value of goal set, the scheduling grade of specified data bag according to the packet after merging;

Step 6: packet is dispatched according to the elementary probability partition function value that the packet after dispatching grade and merging is divided into goal set.

Described step 1 specifically is according to scheduling grade order from high to low, with packet be divided into anxious assemble close, fuzzy set and 3 goal sets of wait set.

Described setting evaluation index specifically is to set time delay, speed fairness and packet loss as evaluation index.

Described step 3 specifically is, according to packet under time delay, speed fairness and the packet loss evaluation index be divided into anxious assemble close, the membership function of fuzzy set and wait set, calculate packet under time delay, speed fairness and the packet loss evaluation index be divided into anxious assemble close, the elementary probability partition function value of fuzzy set and wait set;

Wherein, under the described time delay evaluation index packet be divided into anxious assemble close, the membership function of fuzzy set and wait set is:

$T\_{\mathrm{\&mu;}}_1\left(i\right)=\exp (-{\mathrm{<figure-callout\; id="1"\; label="C"\; filenames="HDA0000075533870000011.png,HDA0000075533870000021.png"\; state="\{\{state\}\}">C</figure-callout>}}_1*\mathrm{\&pi;}*{\left(\frac{T\_\mathrm{delay}\left(i\right)-(T\_\max \left(s\left(i\right)\right)/2)}{2}\right)}^2)$

$T\_{\mathrm{\&mu;}}_2\left(i\right)=\exp (-{\mathrm{<figure-callout\; id="2"\; label="C"\; filenames="HDA0000075533870000011.png"\; state="\{\{state\}\}">C</figure-callout>}}_2*\mathrm{\&pi;}*{\left(\frac{T\_\mathrm{delay}\left(i\right)-T\_\max \left(s\left(i\right)\right)}{2}\right)}^2)$

T_μ ₃(i)＝exp(-C ₃*π*(T_delay(i)) ²)

In the formula:

T_ μ ₁(i) be divided into the anxious membership function that closes of assembling for packet i under the time delay evaluation index;

T_ μ ₂(i) be the membership function that packet i is divided into fuzzy set under the time delay evaluation index;

T_ μ ₃(i) be divided into the membership function of waiting for set for packet i under the time delay evaluation index;

C ₁, C ₂, C ₃For regulating the parameter of membership function;

π is a circumference ratio;

T_delay (i) is the propagation delay time of packet i;

T_delay_max (s (i)) is the propagation delay time upper limit of professional s under the packet i;

Described according to packet under the time delay evaluation index be divided into anxious assemble close, the membership function of fuzzy set and wait set, calculate packet under the time delay evaluation index be divided into anxious assemble close, the computing formula of the elementary probability partition function value of fuzzy set and wait set is:

$\mathrm{BPAF}\_T_1\left(i\right)=\frac{T\_{\mathrm{\&mu;}}_1\left(i\right)}{\underset{j=\mathrm{1,2,3}}{\mathrm{\&Sigma;}}T\_{\mathrm{\&mu;}}_j\left(i\right)}$

$\mathrm{BPAF}\_T_2\left(i\right)=\frac{T\_{\mathrm{\&mu;}}_2\left(i\right)}{\underset{j=\mathrm{1,2,3}}{\mathrm{\&Sigma;}}T\_{\mathrm{\&mu;}}_j\left(i\right)}$

$\mathrm{BPAF}\_T_3\left(i\right)=\frac{T\_{\mathrm{\&mu;}}_3\left(i\right)}{\underset{j=\mathrm{1,2,3}}{\mathrm{\&Sigma;}}T\_{\mathrm{\&mu;}}_j\left(i\right)}$

In the formula:

BPAF_T ₁(i) be divided into the anxious elementary probability partition function value of closing of assembling for packet i under the time delay evaluation index;

BPAF_T ₂(i) be the elementary probability partition function value that packet i is divided into fuzzy set under the time delay evaluation index;

BPAF_T ₃(i) be divided into the elementary probability partition function value of waiting for set for packet i under the time delay evaluation index;

Under the described speed fairness evaluation index packet be divided into anxious assemble close, the membership function of fuzzy set and wait set is:

$R\_{\mathrm{\&mu;}}_1\left(i\right)=\frac{R\_\mathrm{total}\left(s\left(i\right)\right)-R\_\mathrm{aver}\left(s\left(i\right)\right)}{R\_\mathrm{total}\left(s\left(i\right)\right)}$

$R\_{\mathrm{\&mu;}}_2\left(i\right)=\frac{R\_\mathrm{aver}\left(s\left(i\right)\right)}{R\_\mathrm{total}\left(s\left(i\right)\right)}$

$R\_{\mathrm{\&mu;}}_3\left(i\right)=\left\{\begin{array}{cc}\frac{2R\_\mathrm{aver}\left(s\left(i\right)\right)}{R\_\mathrm{total}\left(s\left(i\right)\right)}& (0\&le;R\_\mathrm{aver}\left(s\left(i\right)\right)\&le;\frac{1}{2}R\_\mathrm{total}\left(s\left(i\right)\right))\\ \frac{2*R\_\mathrm{total}\left(s\left(i\right)\right)-2*R\_\mathrm{aver}\left(s\left(i\right)\right)}{R\_\mathrm{total}\left(s\left(i\right)\right)}& (\frac{1}{2}R\_\mathrm{total}\left(s\left(i\right)\right)\&le;R\_\mathrm{aver}\left(s\left(i\right)\right)\&le;1)\end{array}\right.$

In the formula:

R_ μ ₁(i) be divided into the anxious membership function that closes of assembling for packet i under the speed fairness evaluation index;

R_ μ ₂(i) be the membership function that packet i is divided into fuzzy set under the speed fairness evaluation index;

R_ μ ₃(i) be divided into the membership function of waiting for set for packet i under the speed fairness evaluation index;

R_aver (s (i)) is the average transmission rate of professional s under i the packet;

R_total (s (i)) is the speed sum of the affiliated professional s of i packet in whole subchannel transmission;

Described according to packet under the speed fairness evaluation index be divided into order anxious assemble close, the membership function of fuzzy set and wait set, calculate packet under the speed fairness evaluation index be divided into anxious assemble close, the computing formula of the elementary probability partition function value of fuzzy set and wait set is:

$\mathrm{BPAF}\_R_1\left(i\right)=\frac{R\_{\mathrm{\&mu;}}_1\left(i\right)}{\underset{j=\mathrm{1,2,3}}{\mathrm{\&Sigma;}}R\_{\mathrm{\&mu;}}_j\left(i\right)}$

$\mathrm{BPAF}\_R_2\left(i\right)=\frac{R\_{\mathrm{\&mu;}}_2\left(i\right)}{\underset{j=\mathrm{1,2,3}}{\mathrm{\&Sigma;}}R\_{\mathrm{\&mu;}}_j\left(i\right)}$

$\mathrm{BPAF}\_R_3\left(i\right)=\frac{R\_{\mathrm{\&mu;}}_3\left(i\right)}{\underset{j=\mathrm{1,2,3}}{\mathrm{\&Sigma;}}R\_{\mathrm{\&mu;}}_3\left(i\right)}$

In the formula:

BPAF_R ₁For packet i under the speed fairness evaluation index is divided into the anxious elementary probability partition function value of closing of assembling;

BPAF_R ₂Be divided into the elementary probability partition function value of fuzzy set for packet i under the speed fairness evaluation index;

BPAF_R ₃Be divided into the elementary probability partition function value of waiting for set for packet i under the speed fairness evaluation index;

Under the described packet loss evaluation index packet be divided into anxious assemble close, the membership function of fuzzy set and wait set is:

$L\_{\mathrm{\&mu;}}_1\left(i\right)=\exp (-{\mathrm{<figure-callout\; id="4"\; label="C"\; filenames="HDA0000075533870000011.png"\; state="\{\{state\}\}">C</figure-callout>}}_4*\mathrm{\&pi;}*{\left(\frac{(L\_\mathrm{loss}\left(s\left(i\right)\right)-(100*L\_\mathrm{loss}\_\max \left(s\left(i\right)\right))/2)}{2}\right)}^2)$

$L\_{\mathrm{\&mu;}}_2\left(i\right)=\exp (-{\mathrm{<figure-callout\; id="5"\; label="C"\; filenames="HDA0000075533870000011.png,HDA0000075533870000021.png"\; state="\{\{state\}\}">C</figure-callout>}}_5*\mathrm{\&pi;}*{\frac{(L\_\mathrm{loss}\left(s\left(i\right)\right)-(100*L\_\mathrm{loss}\_\max \left(s\left(i\right)\right))}{2})}^2)$

L_μ ₃(i)＝exp(-C ₆*π*(L_loss(s(i))) ²)

In the formula:

L_ μ ₁(i) be divided into the anxious membership function that closes of assembling for packet i under the packet loss evaluation index;

L_ μ ₂(i) be the membership function that packet i is divided into fuzzy set under the packet loss evaluation index;

L_ μ ₃(i) be divided into the membership function of waiting for set for packet i under the packet loss evaluation index;

C ₄, C ₅, C ₆For regulating the parameter of membership function;

L_loss (s (i)) is the packet loss of professional s before packet i sends under the packet i;

L_loss_max (s (i)) is the packet loss upper limit of professional s under the packet i;

Described according to packet under the packet loss evaluation index be divided into anxious assemble close, the membership function of fuzzy set and wait set, calculate packet under the packet loss evaluation index be divided into anxious assemble close, the computing formula of the elementary probability partition function value of fuzzy set and wait set is:

$\mathrm{BPAF}\_L_1\left(i\right)=\frac{L\_{\mathrm{\&mu;}}_1\left(i\right)}{\underset{j=\mathrm{1,2,3}}{\mathrm{\&Sigma;}}L\_{\mathrm{\&mu;}}_j\left(i\right)}$

$\mathrm{BPAF}\_L_2\left(i\right)=\frac{L\_{\mathrm{\&mu;}}_2\left(i\right)}{\underset{j=\mathrm{1,2,3}}{\mathrm{\&Sigma;}}L\_{\mathrm{\&mu;}}_j\left(i\right)}$

$\mathrm{BPAF}\_L_3\left(i\right)=\frac{L\_{\mathrm{\&mu;}}_3\left(i\right)}{\underset{j=\mathrm{1,2,3}}{\mathrm{\&Sigma;}}L\_{\mathrm{\&mu;}}_j\left(i\right)}$

In the formula:

BPAF_L ₁(i) be divided into the anxious elementary probability partition function value of closing of assembling for packet i under the packet loss evaluation index;

BPAF_L ₂(i) be the elementary probability partition function value that packet i is divided into fuzzy set under the packet loss evaluation index;

BPAF_L ₃(i) be divided into the elementary probability partition function value of waiting for set for packet i under the packet loss evaluation index.

In the described step 4, the elementary probability partition function value that packet under each evaluation index is divided into goal set merges, the elementary probability partition function value that packet after obtaining merging is divided into goal set specifically is, with packet under time delay, speed fairness and the packet loss evaluation index be divided into anxious assemble close, the elementary probability partition function value of fuzzy set and wait set merges, packet after obtaining merging be divided into anxious assemble close, the elementary probability partition function value of fuzzy set and wait set, computing formula is:

$\mathrm{BPAF}\_{\mathrm{TRL}}_1\left(i\right)=\frac{\mathrm{BPAF}\_{\mathrm{TR}}_1\left(i\right)*\mathrm{BPAF}\_L_2\left(i\right)+\mathrm{BPAF}\_{\mathrm{TR}}_2\left(i\right)*\mathrm{BPAF}\_L_1\left(i\right)+\mathrm{BPAF}\_{\mathrm{TR}}_1\left(i\right)*\mathrm{BPAF}\_L_2\left(i\right)}{1-K\left(i\right)}$

$\mathrm{BPAF}\_{\mathrm{TRL}}_2\left(i\right)=\frac{\mathrm{BPAF}\_{\mathrm{TR}}_2\left(i\right)*\mathrm{BPAF}\_L_2\left(i\right)}{1-K\left(i\right)}$

$\mathrm{BPAF}\_{\mathrm{TRL}}_3\left(i\right)=\frac{\mathrm{BPAF}\_{\mathrm{TR}}_2\left(i\right)*\mathrm{BPAF}\_L_3\left(i\right)+\mathrm{BPAF}\_{\mathrm{TR}}_3\left(i\right)*\mathrm{BPAF}\_L_2\left(i\right)*\mathrm{BPAF}\_{\mathrm{TR}}_3\left(i\right)*\mathrm{BPAF}\_L_3\left(i\right)}{1-K\left(i\right)}$

In the formula:

BPAF_TRL ₁(i) be divided into the anxious elementary probability partition function value of closing of assembling for merging back packet i;

BPAF_TRL ₂(i) for merging the elementary probability partition function value that back packet i is divided into fuzzy set;

BPAF_TRL ₃(i) be divided into the elementary probability partition function value of waiting for set for merging back packet i;

BPAF_TR ₁(i) for packet i under time delay and two evaluation indexes of speed fairness being divided into anxious elementary probability partition function value of assembling after the elementary probability partition function value of closing merges;

BPAF_TR ₂(i) be the elementary probability partition function value of the elementary probability partition function value that packet i under time delay and two evaluation indexes of speed fairness is divided into fuzzy set after merging;

BPAF_TR ₃(i) for packet i under time delay and two evaluation indexes of speed fairness being divided into the elementary probability partition function value after the elementary probability partition function value of waiting for set merges;

K (i) is the inconsistent factor of packet i under time delay, speed fairness and three evaluation indexes of packet loss, and its computing formula is:

K(i)＝BPAF_TR ₁(i)*BPAF_L ₃(i)+BPAF_L ₃(i)*BPAF_TR ₁(i)。

Describedly packet i under time delay and two evaluation indexes of speed fairness be divided into anxious computing formula of assembling the elementary probability partition function value after the elementary probability partition function value of closing merges be:

$\mathrm{BPAF}\_{\mathrm{TR}}_1\left(i\right)=\frac{\mathrm{BPAF}\_T_1\left(i\right)*\mathrm{BPAF}\_R_1\left(i\right)+\mathrm{BPAF}\_T_1\left(i\right)*\mathrm{BPAF}\_R_2\left(i\right)+\mathrm{BPAF}\_R_1\left(i\right)*\mathrm{BPAF}\_T_2\left(i\right)}{1-K_1\left(i\right)};$

The computing formula of the elementary probability partition function value after the described elementary probability partition function value that packet i under time delay and two evaluation indexes of speed fairness is divided into fuzzy set merges is:

$\mathrm{BPAF}\_{\mathrm{TR}}_2\left(i\right)=\frac{\mathrm{BPAF}\_T_2\left(i\right)*\mathrm{BPAF}\_R_2\left(i\right)}{1-K_1\left(i\right)};$

The described computing formula that packet i under time delay and two evaluation indexes of speed fairness is divided into the elementary probability partition function value after the elementary probability partition function value of waiting for set merges is:

$\mathrm{BPAF}\_{\mathrm{TR}}_3\left(i\right)=\frac{\mathrm{BPAF}\_T_3\left(i\right)*\mathrm{BPAF}\_R_3\left(i\right)+\mathrm{BPAF}\_T_2\left(i\right)*\mathrm{BPAF}\_R_3\left(i\right)+\mathrm{BPAF}\_R_2\left(i\right)*\mathrm{BPAF}\_T_3\left(i\right)}{1-K_1\left(i\right)};$

K ₁(i) be the inconsistent factor of packet i under time delay, two evaluation indexes of speed fairness, its computing formula is: K ₁(i)=BPAF_T ₁(i) * BPAF_R ₃(i)+BPAF_T ₃(i) * BPAF_R ₁(i).

Described step 6 is specifically: for the packet of different brackets, dispatch according to scheduling grade order from high to low; For same grade data package, when the scheduling grade of packet is anxious timing, is divided into according to the packet after merging and suddenly assembles the descending order of elementary probability partition function value of closing and dispatch; When the scheduling grade of packet when fuzzy, the ascending order of elementary probability partition function value that is divided into fuzzy set according to the packet after merging is dispatched; Bide one's time for waiting when the scheduling grade of packet, be divided into the ascending order of elementary probability partition function value of waiting for set according to the packet after merging and dispatch.

The scheduling grade that the present invention comes the decision data bag by the evaluation index that merges in the QoS component overcomes precision, reliability and the lower defective of accuracy when according to the single evaluation index scheduling of resource priority being differentiated.

Description of drawings

The multi-service OFDM that Fig. 1 is based on evidence theory strides a layer dynamic resource allocation method flow chart;

When Fig. 2 is T_max under the time delay evaluation index (s (i))=40ms packet i be divided into anxious assemble close, the membership function curve chart of fuzzy set and wait set;

When Fig. 3 is R_total under the speed fairness evaluation index (s (i))=300bps packet i be divided into anxious assemble close, the membership function curve chart of fuzzy set and wait set;

Fig. 4 be L_max under the packet loss evaluation index (s (i))=0.3 o'clock packet i be divided into anxious assemble close, the membership function curve chart of fuzzy set and wait set;

Fig. 5 is method provided by the invention and PF, EDF method mean transit delay correlation curve figure;

Fig. 6 is method provided by the invention and PF, EDF method average transmission rate correlation curve figure;

Fig. 7 is method provided by the invention and PF, EDF method packet loss correlation curve figure.

Embodiment

Below in conjunction with accompanying drawing, preferred embodiment is elaborated.Should be emphasized that following explanation only is exemplary, rather than in order to limit the scope of the invention and to use.

Optional packet i, class of service is s under this packet, T_delay (i)=20ms, R_aver (s)=200bps, L_loss (s)=0.2, T_max (s)=40ms, R_total (s)=300bps, L_max (s)=0.3.

The multi-service OFDM that Fig. 1 is based on evidence theory strides a layer dynamic resource allocation method flow chart.According to shown in Figure 1, method provided by the invention comprises following implementation step:

Step 1:, set up the goal set that is used for the dividing data bag according to the scheduling grade.

In the present embodiment, according to scheduling grade order from high to low, with packet be divided into anxious assemble close, fuzzy set and 3 goal sets of wait set.

Step 2: set evaluation index.

Consider the evaluation index in the QoS component, choose time delay, speed fairness and packet loss as the evaluation index in the present embodiment.

Step 3: be divided into the membership function of goal set according to packet under each evaluation index, calculate the elementary probability partition function value that packet under each evaluation index is divided into goal set.

According to step 1 and 2 goal set and the evaluation indexes of determining, according to packet under time delay, speed fairness and the packet loss evaluation index be divided into anxious assemble close, the membership function of fuzzy set and wait set, calculate packet under time delay, speed fairness and the packet loss evaluation index be divided into anxious assemble close, the elementary probability partition function value of fuzzy set and wait set.

Wherein, under the time delay evaluation index packet be divided into anxious assemble close, the membership function of fuzzy set and wait set is respectively:

$T\_{\mathrm{\&mu;}}_1\left(i\right)=\exp (-{\mathrm{<figure-callout\; id="1"\; label="C"\; filenames="HDA0000075533870000011.png,HDA0000075533870000021.png"\; state="\{\{state\}\}">C</figure-callout>}}_1*\mathrm{\&pi;}*{\left(\frac{T\_\mathrm{delay}\left(i\right)-(T\_\max \left(s\left(i\right)\right)/2)}{2}\right)}^2)---\left(1\right)$

$T\_{\mathrm{\&mu;}}_2\left(i\right)=\exp (-{\mathrm{<figure-callout\; id="2"\; label="C"\; filenames="HDA0000075533870000011.png"\; state="\{\{state\}\}">C</figure-callout>}}_2*\mathrm{\&pi;}*{\left(\frac{T\_\mathrm{delay}\left(i\right)-T\_\max \left(s\left(i\right)\right)}{2}\right)}^2)---\left(2\right)$

T_μ ₃(i)＝exp(-C ₃*π*(T_delay(i)) ²) (3)

In the formula:

T_ μ ₁(i) be divided into the anxious membership function that closes of assembling for packet i under the time delay evaluation index;

T_ μ ₂(i) be the membership function that packet i is divided into fuzzy set under the time delay evaluation index;

T_ μ ₃(i) be divided into the membership function of waiting for set for packet i under the time delay evaluation index;

C ₁, C ₂, C ₃For regulating the parameter of membership function;

π is a circumference ratio;

T_delay (i) is the propagation delay time of packet i;

T_delay_max (s (i)) is the propagation delay time upper limit of professional s under the packet i.

According to packet under the time delay evaluation index be divided into anxious assemble close, the membership function of fuzzy set and wait set, calculate packet under the time delay evaluation index be divided into anxious assemble close, the computing formula of the elementary probability partition function value of fuzzy set and wait set is:

$\mathrm{BPAF}\_T_1\left(i\right)=\frac{T\_{\mathrm{\&mu;}}_1\left(i\right)}{\underset{j=\mathrm{1,2,3}}{\mathrm{\&Sigma;}}T\_{\mathrm{\&mu;}}_j\left(i\right)}---\left(4\right)$

$\mathrm{BPAF}\_T_2\left(i\right)=\frac{T\_{\mathrm{\&mu;}}_2\left(i\right)}{\underset{j=\mathrm{1,2,3}}{\mathrm{\&Sigma;}}T\_{\mathrm{\&mu;}}_j\left(i\right)}---\left(5\right)$

$\mathrm{BPAF}\_T_3\left(i\right)=\frac{T\_{\mathrm{\&mu;}}_3\left(i\right)}{\underset{j=\mathrm{1,2,3}}{\mathrm{\&Sigma;}}T\_{\mathrm{\&mu;}}_j\left(i\right)}---\left(6\right)$

In the formula:

BPAF_T ₁(i) be divided into the anxious elementary probability partition function value of closing of assembling for packet i under the time delay evaluation index;

BPAF_T ₂(i) be the elementary probability partition function value that packet i is divided into fuzzy set under the time delay evaluation index;

BPAF_T ₃(i) be divided into the elementary probability partition function value of waiting for set for packet i under the time delay evaluation index.

Under the speed fairness evaluation index packet be divided into anxious assemble close, the membership function of fuzzy set and wait set is respectively:

$R\_{\mathrm{\&mu;}}_1\left(i\right)=\frac{R\_\mathrm{total}\left(s\left(i\right)\right)-R\_\mathrm{aver}\left(s\left(i\right)\right)}{R\_\mathrm{total}\left(s\left(i\right)\right)}---\left(7\right)$

$R\_{\mathrm{\&mu;}}_2\left(i\right)=\frac{R\_\mathrm{aver}\left(s\left(i\right)\right)}{R\_\mathrm{total}\left(s\left(i\right)\right)}---\left(8\right)$

$R\_{\mathrm{\&mu;}}_3\left(i\right)=\left\{\begin{array}{cc}\frac{2R\_\mathrm{aver}\left(s\left(i\right)\right)}{R\_\mathrm{total}\left(s\left(i\right)\right)}& (0\&le;R\_\mathrm{aver}\left(s\left(i\right)\right)\&le;\frac{1}{2}R\_\mathrm{total}\left(s\left(i\right)\right))\\ \frac{2*R\_\mathrm{total}\left(s\left(i\right)\right)-2*R\_\mathrm{aver}\left(s\left(i\right)\right)}{R\_\mathrm{total}\left(s\left(i\right)\right)}& (\frac{1}{2}R\_\mathrm{total}\left(s\left(i\right)\right)\&le;R\_\mathrm{aver}\left(s\left(i\right)\right)\&le;1)\end{array}\right.---\left(9\right)$

In the formula:

R_ μ ₁(i) be divided into the anxious membership function that closes of assembling for packet i under the speed fairness evaluation index;

R_ μ ₂(i) be the membership function that packet i is divided into fuzzy set under the speed fairness evaluation index;

R_ μ ₃(i) be divided into the membership function of waiting for set for packet i under the speed fairness evaluation index;

R_aver (s (i)) is the average transmission rate of professional s under i the packet;

R_total (s (i)) is the speed sum of the affiliated professional s of i packet in whole subchannel transmission.

According to packet under the speed fairness evaluation index be divided into anxious assemble close, the membership function of fuzzy set and wait set, calculate packet under the speed fairness evaluation index be divided into anxious assemble close, the computing formula of the elementary probability partition function value of fuzzy set and wait set is:

$\mathrm{BPAF}\_R_1\left(i\right)=\frac{R\_{\mathrm{\&mu;}}_1\left(i\right)}{\underset{j=\mathrm{1,2,3}}{\mathrm{\&Sigma;}}R\_{\mathrm{\&mu;}}_j\left(i\right)}---\left(10\right)$

$\mathrm{BPAF}\_R_2\left(i\right)=\frac{R\_{\mathrm{\&mu;}}_2\left(i\right)}{\underset{j=\mathrm{1,2,3}}{\mathrm{\&Sigma;}}R\_{\mathrm{\&mu;}}_j\left(i\right)}---\left(11\right)$

$\mathrm{BPAF}\_R_3\left(i\right)=\frac{R\_{\mathrm{\&mu;}}_3\left(i\right)}{\underset{j=\mathrm{1,2,3}}{\mathrm{\&Sigma;}}R\_{\mathrm{\&mu;}}_3\left(i\right)}---\left(12\right)$

In the formula:

BPAF_R ₁For packet i under the speed fairness evaluation index is divided into the anxious elementary probability partition function value of closing of assembling;

BPAF_R ₂Be divided into the elementary probability partition function value of fuzzy set for packet i under the speed fairness evaluation index;

BPAF_R ₃Be divided into the elementary probability partition function value of waiting for set for packet i under the speed fairness evaluation index.

Under the packet loss evaluation index packet be divided into anxious assemble close, fuzzy set and wait set membership function be:

$L\_{\mathrm{\&mu;}}_1\left(i\right)=\exp (-{\mathrm{<figure-callout\; id="4"\; label="C"\; filenames="HDA0000075533870000011.png"\; state="\{\{state\}\}">C</figure-callout>}}_4*\mathrm{\&pi;}*{\left(\frac{(L\_\mathrm{loss}\left(s\left(i\right)\right)-(100*L\_\mathrm{loss}\_\max \left(s\left(i\right)\right))/2)}{2}\right)}^2)---\left(13\right)$

$L\_{\mathrm{\&mu;}}_2\left(i\right)=\exp (-{\mathrm{<figure-callout\; id="5"\; label="C"\; filenames="HDA0000075533870000011.png,HDA0000075533870000021.png"\; state="\{\{state\}\}">C</figure-callout>}}_5*\mathrm{\&pi;}*{\frac{(L\_\mathrm{loss}\left(s\left(i\right)\right)-(100*L\_\mathrm{loss}\_\max \left(s\left(i\right)\right))}{2})}^2)---\left(14\right)$

L_μ ₃(i)＝exp(-C ₆*π*(L_loss(s(i))) ²) (15)

In the formula:

L_ μ ₁(i) be divided into the anxious membership function that closes of assembling for packet i under the packet loss evaluation index;

L_ μ ₂(i) be the membership function that packet i is divided into fuzzy set under the packet loss evaluation index;

L_ μ ₃(i) be divided into the membership function of waiting for set for packet i under the packet loss evaluation index;

C ₄, C ₅, C ₆For regulating the parameter of membership function;

L_loss (s (i)) is the packet loss of professional s before packet i sends under the packet i;

L_loss_max (s (i)) is the packet loss upper limit of professional s under the packet i.

According to packet under the packet loss evaluation index be divided into anxious assemble close, the membership function of fuzzy set and wait set, calculate packet under the packet loss evaluation index be divided into anxious assemble close, the computing formula of the elementary probability partition function value of fuzzy set and wait set is:

$\mathrm{BPAF}\_L_1\left(i\right)=\frac{L\_{\mathrm{\&mu;}}_1\left(i\right)}{\underset{j=\mathrm{1,2,3}}{\mathrm{\&Sigma;}}L\_{\mathrm{\&mu;}}_j\left(i\right)}---\left(16\right)$

$\mathrm{BPAF}\_L_2\left(i\right)=\frac{L\_{\mathrm{\&mu;}}_2\left(i\right)}{\underset{j=\mathrm{1,2,3}}{\mathrm{\&Sigma;}}L\_{\mathrm{\&mu;}}_j\left(i\right)}---\left(17\right)$

$\mathrm{BPAF}\_L_3\left(i\right)=\frac{L\_{\mathrm{\&mu;}}_3\left(i\right)}{\underset{j=\mathrm{1,2,3}}{\mathrm{\&Sigma;}}L\_{\mathrm{\&mu;}}_j\left(i\right)}---\left(18\right)$

In the formula:

BPAF_L ₁(i) be divided into the anxious elementary probability partition function value of closing of assembling for packet i under the packet loss evaluation index;

BPAF_L ₂(i) be the elementary probability partition function value that packet i is divided into fuzzy set under the packet loss evaluation index;

BPAF_L ₃(i) be divided into the elementary probability partition function value of waiting for set for packet i under the packet loss evaluation index.

In the present embodiment, earlier respectively with T_max (s)=40ms substitution formula (1) (2) (3), obtain the time delay membership function, as shown in Figure 2.With T_delay (i)=20ms, T_max (s)=40ms substitution formula (4) (5) (6), obtain packet under the time delay evaluation index be divided into anxious assemble close, the elementary probability partition function value BPAF_T of fuzzy set and wait set ₁(i)=0.9071, BPAF_T ₂(i)=0.0537, BPAF_T ₃(i)=0.0392.Judge that under the time delay evaluation index its scheduling grade is anxious the accent.

Again R_total (s)=300bps substitution formula (7) (8) (9) is obtained speed fairness membership function, as shown in Figure 3.With R_aver (s)=200bps, R_total (s)=300bps substitution formula (10) (11) (12) so that try to achieve packet under the speed fairness evaluation index be divided into anxious assemble close, the elementary probability partition function value BPAF_R of fuzzy set and wait set ₁(i)=0.1429, BPAF_R ₂(i)=0.2857, BPAF_R ₃(i)=0.5714.Judge that under speed fairness evaluation index its scheduling grade is for waiting for.

At last L_max (s)=0.3 substitution formula (13) (14) (15) is obtained the packet loss membership function, as shown in Figure 4.With L_loss (s)=0.2, L_max (s)=0.3 substitution formula (16) (17) (18) so that try to achieve packet under the packet loss evaluation index be divided into anxious assemble close, the elementary probability partition function value BPAF_T of fuzzy set and wait set ₁(i)=0.0209, BPAF_L ₂(i)=4.0583e-009, BPAF_T ₃(i)=0.9791.Judge that under the packet loss evaluation index its scheduling grade is for waiting for.

Step 4: according to the D-S evidence theory, the elementary probability partition function value that packet under each evaluation index is divided into goal set merges, and the packet after obtaining merging is divided into goal set elementary probability partition function value.

In the present embodiment, the elementary probability partition function value that packet under each evaluation index is divided into goal set merges, packet after obtaining merging is divided into goal set elementary probability partition function value specifically: with packet under time delay, speed fairness and the packet loss evaluation index be divided into anxious assemble close, the elementary probability partition function value of fuzzy set and wait set merges, packet after obtaining merging be divided into anxious assemble close, fuzzy set and wait set elementary probability partition function value, computing formula is:

$\mathrm{BPAF}\_{\mathrm{TRL}}_1\left(i\right)=\frac{\mathrm{BPAF}\_{\mathrm{TR}}_1\left(i\right)*\mathrm{BPAF}\_L_2\left(i\right)+\mathrm{BPAF}\_{\mathrm{TR}}_2\left(i\right)*\mathrm{BPAF}\_L_1\left(i\right)+\mathrm{BPAF}\_{\mathrm{TR}}_1\left(i\right)*\mathrm{BPAF}\_L_2\left(i\right)}{1-K\left(i\right)}---\left(19\right)$

$\mathrm{BPAF}\_{\mathrm{TRL}}_2\left(i\right)=\frac{\mathrm{BPAF}\_{\mathrm{TR}}_2\left(i\right)*\mathrm{BPAF}\_L_2\left(i\right)}{1-K\left(i\right)}---\left(20\right)$

$\mathrm{BPAF}\_{\mathrm{TRL}}_3\left(i\right)=\frac{\mathrm{BPAF}\_{\mathrm{TR}}_2\left(i\right)*\mathrm{BPAF}\_L_3\left(i\right)+\mathrm{BPAF}\_{\mathrm{TR}}_3\left(i\right)*\mathrm{BPAF}\_L_2\left(i\right)*\mathrm{BPAF}\_{\mathrm{TR}}_3\left(i\right)*\mathrm{BPAF}\_L_3\left(i\right)}{1-K\left(i\right)}---\left(21\right)$

In the formula:

BPAF_TRL ₁(i) be divided into the anxious elementary probability partition function value of closing of assembling for merging back packet i;

BPAF_TRL ₂(i) for merging the elementary probability partition function value that back packet i is divided into fuzzy set;

BPAF_TRL ₃(i) be divided into the elementary probability partition function value of waiting for set for merging back packet i;

BPAF_TR ₁(i) for packet i under time delay and two evaluation indexes of speed fairness being divided into anxious elementary probability partition function value of assembling after the elementary probability partition function value of closing merges;

BPAF_TR ₂(i) be the elementary probability partition function value of the elementary probability partition function value that packet i under time delay and two evaluation indexes of speed fairness is divided into fuzzy set after merging;

BPAF_TR ₃(i) for packet i under time delay and two evaluation indexes of speed fairness being divided into the elementary probability partition function value after the elementary probability partition function value of waiting for set merges;

K (i) is the inconsistent factor of packet i under time delay, speed fairness and three evaluation indexes of packet loss, and its computing formula is:

K(i)＝BPAF_TR ₁(i)*BPAF_L ₃(i)+BPAF_L ₃(i)*BPAF_TR ₁(i)。

Packet i under time delay and two evaluation indexes of speed fairness is divided into anxious computing formula of assembling the elementary probability partition function value after the elementary probability partition function value of closing merges is:

$\mathrm{BPAF}\_{\mathrm{TR}}_1\left(i\right)=\frac{\mathrm{BPAF}\_T_1\left(i\right)*\mathrm{BPAF}\_R_1\left(i\right)+\mathrm{BPAF}\_T_1\left(i\right)*\mathrm{BPAF}\_R_2\left(i\right)+\mathrm{BPAF}\_R_1\left(i\right)*\mathrm{BPAF}\_T_2\left(i\right)}{1-K_1\left(i\right)};$

The computing formula of the elementary probability partition function value after the elementary probability partition function value that packet i under time delay and two evaluation indexes of speed fairness is divided into fuzzy set merges is:

$\mathrm{BPAF}\_{\mathrm{TR}}_2\left(i\right)=\frac{\mathrm{BPAF}\_T_2\left(i\right)*\mathrm{BPAF}\_R_2\left(i\right)}{1-K_1\left(i\right)};$

The computing formula that packet i under time delay and two evaluation indexes of speed fairness is divided into the elementary probability partition function value after the elementary probability partition function value of waiting for set merges is:

$\mathrm{BPAF}\_{\mathrm{TR}}_3\left(i\right)=\frac{\mathrm{BPAF}\_T_3\left(i\right)*\mathrm{BPAF}\_R_3\left(i\right)+\mathrm{BPAF}\_T_2\left(i\right)*\mathrm{BPAF}\_R_3\left(i\right)+\mathrm{BPAF}\_R_2\left(i\right)*\mathrm{BPAF}\_T_3\left(i\right)}{1-K_1\left(i\right)};$

K ₁(i) be the inconsistent factor of packet i under time delay, two evaluation indexes of speed fairness, its computing formula is: K ₁(i)=BPAF_T ₁(i) * BPAF_R ₃(i)+BPAF_T ₃(i) * BPAF_R ₁(i).

In the present embodiment, packet under the time delay evaluation index that calculates above is divided into anxious assembling to be closed, the elementary probability partition function value of fuzzy set and wait set, packet is divided into suddenly to assemble and closes under the speed fairness evaluation index, the elementary probability partition function value of fuzzy set and wait set, packet is divided into suddenly to assemble and closes under the packet loss evaluation index, the elementary probability partition function value substitution formula (19) (20) (21) of fuzzy set and wait set, the packet after obtaining merging are divided into anxious assembling and close, the elementary probability partition function value of fuzzy set and wait set: BPAF_TRL ₁(i)=0.8935, BPAF_TRL ₂(i)=0.0765, BPAF_TRL ₃(i)=0.0380.

Step 5: be divided into the elementary probability partition function value of goal set, the scheduling grade of specified data bag according to the packet after merging.

With the packet after merging be divided into anxious assemble close, the scheduling grade of set that the elementary probability partition function value of fuzzy set and wait set is big is as the scheduling grade of packet.

In the present embodiment, as following table:

	The anxious accent	Fuzzy	Wait for	Court verdict
					Time delay	0.9071	0.0537	0.0392	The anxious accent
The speed fairness	0.1429	0.2857	0.5714	Wait for
					Packet loss	0.0209	4.0583e-009	0.9791	Wait for
After the fusion	0.8935	0.0765	0.0380	The anxious accent

Assembling the elementary probability partition function value maximum of closing because fusion back packet is divided into suddenly, is 0.8935, therefore will " suddenly transfer " the scheduling grade as packet.

Step 6: packet is dispatched according to the elementary probability partition function value that the packet after dispatching grade and merging is divided into goal set.

Specifically:, dispatch according to scheduling grade order from high to low for the packet of different brackets; For same grade data package, when the scheduling grade of packet is anxious timing, is divided into according to the packet after merging and suddenly assembles the descending order of elementary probability partition function value of closing and dispatch; When the scheduling grade of packet when fuzzy, the ascending order of elementary probability partition function value that is divided into fuzzy set according to the packet after merging is dispatched; Bide one's time for waiting when the scheduling grade of packet, be divided into the ascending order of elementary probability partition function value of waiting for set according to the packet after merging and dispatch.

As shown in the table: suppose to have 6 packets, according to they grade and merge after be divided into anxious assemble close, the elementary probability partition function value of fuzzy set and wait set, determine that its dispatching sequence is 2,1,6,5,3,4.

Sequence number	The anxious accent	Fuzzy	Wait for	Court verdict
					1	0.528	0.419	0.053	The anxious accent
2	0.894	0.077	0.038	The anxious accent
					3	0.190	0.127	0.683	Wait for
4	0.054	0.014	0.932	Wait for
					5	0.120	0.806	0.074	Fuzzy
6	0.134	0.720	0.146	Fuzzy

Fig. 5, Fig. 6 and Fig. 7 are respectively method provided by the invention and PF, EDF method mean transit delay, average transmission rate and packet loss correlation curve figure, be not difficult to find out that by Fig. 5, Fig. 6 and Fig. 7 method provided by the invention has overcome precision, reliability and the lower defective of accuracy when according to the single evaluation index scheduling of resource priority being differentiated.

The above; only for the preferable embodiment of the present invention, but protection scope of the present invention is not limited thereto, and anyly is familiar with those skilled in the art in the technical scope that the present invention discloses; the variation that can expect easily or replacement all should be encompassed within protection scope of the present invention.Therefore, protection scope of the present invention should be as the criterion with the protection range of claim.

Claims (7)

1. the multi-service OFDM based on evidence theory strides a layer dynamic resource allocation method, it is characterized in that described method comprises:

Step 1:, set up the goal set that is used for the dividing data bag according to the scheduling grade;

Step 2: set evaluation index;

Step 3: be divided into the membership function of goal set according to packet under each evaluation index, calculate the elementary probability partition function value that packet under each evaluation index is divided into goal set;

Step 4: according to the D-S evidence theory, the elementary probability partition function value that packet under each evaluation index is divided into goal set merges, and the packet after obtaining merging is divided into the elementary probability partition function value of goal set;

Step 5: be divided into the elementary probability partition function value of goal set, the scheduling grade of specified data bag according to the packet after merging;

Step 6: packet is dispatched according to the elementary probability partition function value that the packet after dispatching grade and merging is divided into goal set.

2. a kind of multi-service OFDM based on evidence theory according to claim 1 strides a layer dynamic resource allocation method, it is characterized in that described step 1 specifically is according to scheduling grade order from high to low, with packet be divided into anxious assemble close, fuzzy set and 3 goal sets of wait set.

3. a kind of multi-service OFDM based on evidence theory according to claim 2 strides a layer dynamic resource allocation method, it is characterized in that described setting evaluation index specifically is to set time delay, speed fairness and packet loss as evaluation index.

4. a kind of multi-service OFDM based on evidence theory according to claim 3 strides a layer dynamic resource allocation method, it is characterized in that described step 3 specifically is, according to packet under time delay, speed fairness and the packet loss evaluation index be divided into anxious assemble close, the membership function of fuzzy set and wait set, calculate packet under time delay, speed fairness and the packet loss evaluation index be divided into anxious assemble close, the elementary probability partition function value of fuzzy set and wait set;

Wherein, under the described time delay evaluation index packet be divided into anxious assemble close, the membership function of fuzzy set and wait set is:

$T\_{\mathrm{\&mu;}}_1\left(i\right)=\exp (-C_1*\mathrm{\&pi;}*{\left(\frac{T\_\mathrm{delay}\left(i\right)-(T\_\max \left(s\left(i\right)\right)/2)}{2}\right)}^2)$

$T\_{\mathrm{\&mu;}}_2\left(i\right)=\exp (-C_2*\mathrm{\&pi;}*{\left(\frac{T\_\mathrm{delay}\left(i\right)-T\_\max \left(s\left(i\right)\right)}{2}\right)}^2)$

T_μ ₃(i)＝exp(-C ₃*π*(T_delay(i)) ²)

In the formula:

T_ μ ₁(i) be divided into the anxious membership function that closes of assembling for packet i under the time delay evaluation index;

T_ μ ₂(i) be the membership function that packet i is divided into fuzzy set under the time delay evaluation index;

T_ μ ₃(i) be divided into the membership function of waiting for set for packet i under the time delay evaluation index;

C ₁, C ₂, C ₃For regulating the parameter of membership function;

π is a circumference ratio;

T_delay (i) is the propagation delay time of packet i;

T_delay_max (s (i)) is the propagation delay time upper limit of professional s under the packet i;

Described according to packet under the time delay evaluation index be divided into anxious assemble close, the membership function of fuzzy set and wait set, calculate packet under the time delay evaluation index be divided into anxious assemble close, the computing formula of the elementary probability partition function value of fuzzy set and wait set is:

$\mathrm{BPAF}\_T_1\left(i\right)=\frac{T\_{\mathrm{\&mu;}}_1\left(i\right)}{\underset{j=\mathrm{1,2,3}}{\mathrm{\&Sigma;}}T\_{\mathrm{\&mu;}}_j\left(i\right)}$

$\mathrm{BPAF}\_T_2\left(i\right)=\frac{T\_{\mathrm{\&mu;}}_2\left(i\right)}{\underset{j=\mathrm{1,2,3}}{\mathrm{\&Sigma;}}T\_{\mathrm{\&mu;}}_j\left(i\right)}$

$\mathrm{BPAF}\_T_3\left(i\right)=\frac{T\_{\mathrm{\&mu;}}_3\left(i\right)}{\underset{j=\mathrm{1,2,3}}{\mathrm{\&Sigma;}}T\_{\mathrm{\&mu;}}_j\left(i\right)}$

In the formula:

BPAF_T ₁(i) be divided into the anxious elementary probability partition function value of closing of assembling for packet i under the time delay evaluation index;

BPAF_T ₂(i) be the elementary probability partition function value that packet i is divided into fuzzy set under the time delay evaluation index;

BPAF_T ₃(i) be divided into the elementary probability partition function value of waiting for set for packet i under the time delay evaluation index;

Under the described speed fairness evaluation index packet be divided into anxious assemble close, the membership function of fuzzy set and wait set is:

$R\_{\mathrm{\&mu;}}_1\left(i\right)=\frac{R\_\mathrm{total}\left(s\left(i\right)\right)-R\_\mathrm{aver}\left(s\left(i\right)\right)}{R\_\mathrm{total}\left(s\left(i\right)\right)}$

$R\_{\mathrm{\&mu;}}_2\left(i\right)=\frac{R\_\mathrm{aver}\left(s\left(i\right)\right)}{R\_\mathrm{total}\left(s\left(i\right)\right)}$

$R\_{\mathrm{\&mu;}}_3\left(i\right)=\left\{\begin{array}{cc}\frac{2R\_\mathrm{aver}\left(s\left(i\right)\right)}{R\_\mathrm{total}\left(s\left(i\right)\right)}& (0\&le;R\_\mathrm{aver}\left(s\left(i\right)\right)\&le;\frac{1}{2}R\_\mathrm{total}\left(s\left(i\right)\right))\\ \frac{2*R\_\mathrm{total}\left(s\left(i\right)\right)-2*R\_\mathrm{aver}\left(s\left(i\right)\right)}{R\_\mathrm{total}\left(s\left(i\right)\right)}& (\frac{1}{2}R\_\mathrm{total}\left(s\left(i\right)\right)\&le;R\_\mathrm{aver}\left(s\left(i\right)\right)\&le;1)\end{array}\right.$

In the formula:

R_ μ ₁(i) be divided into the anxious membership function that closes of assembling for packet i under the speed fairness evaluation index;

R_ μ ₂(i) be the membership function that packet i is divided into fuzzy set under the speed fairness evaluation index;

R_ μ ₃(i) be divided into the membership function of waiting for set for packet i under the speed fairness evaluation index;

R_aver (s (i)) is the average transmission rate of professional s under i the packet;

R_total (s (i)) is the speed sum of the affiliated professional s of i packet in whole subchannel transmission;

Described according to packet under the speed fairness evaluation index be divided into anxious assemble close, the membership function of fuzzy set and wait set, calculate packet under the speed fairness evaluation index be divided into anxious assemble close, the computing formula of the elementary probability partition function value of fuzzy set and wait set is:

$\mathrm{BPAF}\_R_1\left(i\right)=\frac{R\_{\mathrm{\&mu;}}_1\left(i\right)}{\underset{j=\mathrm{1,2,3}}{\mathrm{\&Sigma;}}R\_{\mathrm{\&mu;}}_j\left(i\right)}$

$\mathrm{BPAF}\_R_2\left(i\right)=\frac{R\_{\mathrm{\&mu;}}_2\left(i\right)}{\underset{j=\mathrm{1,2,3}}{\mathrm{\&Sigma;}}R\_{\mathrm{\&mu;}}_j\left(i\right)}$

$\mathrm{BPAF}\_R_3\left(i\right)=\frac{R\_{\mathrm{\&mu;}}_3\left(i\right)}{\underset{j=\mathrm{1,2,3}}{\mathrm{\&Sigma;}}R\_{\mathrm{\&mu;}}_3\left(i\right)}$

In the formula:

BPAF_R ₁For packet i under the speed fairness evaluation index is divided into the anxious elementary probability partition function value of closing of assembling;

BPAF_R ₂Be divided into the elementary probability partition function value of fuzzy set for packet i under the speed fairness evaluation index;

BPAF_R ₃Be divided into the elementary probability partition function value of waiting for set for packet i under the speed fairness evaluation index;

Under the described packet loss evaluation index packet be divided into anxious assemble close, the membership function of fuzzy set and wait set is:

$L\_{\mathrm{\&mu;}}_1\left(i\right)=\exp (-C_4*\mathrm{\&pi;}*{\left(\frac{(L\_\mathrm{loss}\left(s\left(i\right)\right)-(100*L\_\mathrm{loss}\_\max \left(s\left(i\right)\right))/2)}{2}\right)}^2)$

$L\_{\mathrm{\&mu;}}_2\left(i\right)=\exp (-C_5*\mathrm{\&pi;}*{\frac{(L\_\mathrm{loss}\left(s\left(i\right)\right)-(100*L\_\mathrm{loss}\_\max \left(s\left(i\right)\right))}{2})}^2)$

L_μ ₃(i)＝exp(-C ₆*π*(L_loss(s(i))) ²)

In the formula:

L_ μ ₁(i) be divided into the anxious membership function that closes of assembling for packet i under the packet loss evaluation index;

L_ μ ₂(i) be the membership function that packet i is divided into fuzzy set under the packet loss evaluation index;

L_ μ ₃(i) be divided into the membership function of waiting for set for packet i under the packet loss evaluation index;

C ₄, C ₅, C ₆For regulating the parameter of membership function;

L_loss (s (i)) is the packet loss of professional s before packet i sends under the packet i;

L_loss_max (s (i)) is the packet loss upper limit of professional s under the packet i;

Described according to packet under the packet loss evaluation index be divided into anxious assemble close, the membership function of fuzzy set and wait set, calculate packet under the packet loss evaluation index be divided into anxious assemble close, the computing formula of the elementary probability partition function value of fuzzy set and wait set is:

$\mathrm{BPAF}\_L_1\left(i\right)=\frac{L\_{\mathrm{\&mu;}}_1\left(i\right)}{\underset{j=\mathrm{1,2,3}}{\mathrm{\&Sigma;}}L\_{\mathrm{\&mu;}}_j\left(i\right)}$

$\mathrm{BPAF}\_L_2\left(i\right)=\frac{L\_{\mathrm{\&mu;}}_2\left(i\right)}{\underset{j=\mathrm{1,2,3}}{\mathrm{\&Sigma;}}L\_{\mathrm{\&mu;}}_j\left(i\right)}$

$\mathrm{BPAF}\_L_3\left(i\right)=\frac{L\_{\mathrm{\&mu;}}_3\left(i\right)}{\underset{j=\mathrm{1,2,3}}{\mathrm{\&Sigma;}}L\_{\mathrm{\&mu;}}_j\left(i\right)}$

In the formula:

BPAF_L ₁(i) be divided into the anxious elementary probability partition function value of closing of assembling for packet i under the packet loss evaluation index;

BPAF_L ₂(i) be the elementary probability partition function value that packet i is divided into fuzzy set under the packet loss evaluation index;

BPAF_L ₃(i) be divided into the elementary probability partition function value of waiting for set for packet i under the packet loss evaluation index.

5. a kind of multi-service OFDM based on evidence theory according to claim 4 strides a layer dynamic resource allocation method, it is characterized in that in the described step 4, the elementary probability partition function value that packet under each evaluation index is divided into goal set merges, the elementary probability partition function value that packet after obtaining merging is divided into goal set specifically is, with time delay, packet is divided into suddenly to assemble and closes under speed fairness and the packet loss evaluation index, the elementary probability partition function value of fuzzy set and wait set merges, and the packet after obtaining merging is divided into anxious assembling and closes, the elementary probability partition function value of fuzzy set and wait set; Computing formula is:

$\mathrm{BPAF}\_{\mathrm{TRL}}_1\left(i\right)=\frac{\mathrm{BPAF}\_{\mathrm{TR}}_1\left(i\right)*\mathrm{BPAF}\_L_2\left(i\right)+\mathrm{BPAF}\_{\mathrm{TR}}_2\left(i\right)*\mathrm{BPAF}\_L_1\left(i\right)+\mathrm{BPAF}\_{\mathrm{TR}}_1\left(i\right)*\mathrm{BPAF}\_L_2\left(i\right)}{1-K\left(i\right)}$

$\mathrm{BPAF}\_{\mathrm{TRL}}_2\left(i\right)=\frac{\mathrm{BPAF}\_{\mathrm{TR}}_2\left(i\right)*\mathrm{BPAF}\_L_2\left(i\right)}{1-K\left(i\right)}$

$\mathrm{BPAF}\_{\mathrm{TRL}}_3\left(i\right)=\frac{\mathrm{BPAF}\_{\mathrm{TR}}_2\left(i\right)*\mathrm{BPAF}\_L_3\left(i\right)+\mathrm{BPAF}\_{\mathrm{TR}}_3\left(i\right)*\mathrm{BPAF}\_L_2\left(i\right)*\mathrm{BPAF}\_{\mathrm{TR}}_3\left(i\right)*\mathrm{BPAF}\_L_3\left(i\right)}{1-K\left(i\right)}$

In the formula:

BPAF_TRL ₁(i) be divided into the anxious elementary probability partition function value of closing of assembling for merging back packet i;

BPAF_TRL ₂(i) for merging the elementary probability partition function value that back packet i is divided into fuzzy set;

BPAF_TRL ₃(i) be divided into the elementary probability partition function value of waiting for set for merging back packet i;

BPAF_TR ₁(i) for packet i under time delay and two evaluation indexes of speed fairness being divided into anxious elementary probability partition function value of assembling after the elementary probability partition function value of closing merges;

BPAF_TR ₂(i) be the elementary probability partition function value of the elementary probability partition function value that packet i under time delay and two evaluation indexes of speed fairness is divided into fuzzy set after merging;

BPAF_TR ₃(i) for packet i under time delay and two evaluation indexes of speed fairness being divided into the elementary probability partition function value after the elementary probability partition function value of waiting for set merges;

K (i) is the inconsistent factor of packet i under time delay, speed fairness and three evaluation indexes of packet loss, and its computing formula is:

K(i)＝BPAF_TR ₁(i)*BPAF_L ₃(i)+BPAF_L ₃(i)*BPAF_TR ₁(i)。

6. a kind of multi-service OFDM based on evidence theory according to claim 5 strides a layer dynamic resource allocation method, it is characterized in that describedly packet i under time delay and two evaluation indexes of speed fairness is divided into anxious computing formula of assembling the elementary probability partition function value after the elementary probability partition function value of closing merges is:

$\mathrm{BPAF}\_{\mathrm{TR}}_1\left(i\right)=\frac{\mathrm{BPAF}\_T_1\left(i\right)*\mathrm{BPAF}\_R_1\left(i\right)+\mathrm{BPAF}\_T_1\left(i\right)*\mathrm{BPAF}\_R_2\left(i\right)+\mathrm{BPAF}\_R_1\left(i\right)*\mathrm{BPAF}\_T_2\left(i\right)}{1-K_1\left(i\right)};$

The computing formula of the elementary probability partition function value after the described elementary probability partition function value that packet i under time delay and two evaluation indexes of speed fairness is divided into fuzzy set merges is:

$\mathrm{BPAF}\_{\mathrm{TR}}_2\left(i\right)=\frac{\mathrm{BPAF}\_T_2\left(i\right)*\mathrm{BPAF}\_R_2\left(i\right)}{1-K_1\left(i\right)};$

The described computing formula that packet i under time delay and two evaluation indexes of speed fairness is divided into the elementary probability partition function value after the elementary probability partition function value of waiting for set merges is:

$\mathrm{BPAF}\_{\mathrm{TR}}_3\left(i\right)=\frac{\mathrm{BPAF}\_T_3\left(i\right)*\mathrm{BPAF}\_R_3\left(i\right)+\mathrm{BPAF}\_T_2\left(i\right)*\mathrm{BPAF}\_R_3\left(i\right)+\mathrm{BPAF}\_R_2\left(i\right)*\mathrm{BPAF}\_T_3\left(i\right)}{1-K_1\left(i\right)};$

K ₁(i) be the inconsistent factor of packet i under time delay, two evaluation indexes of speed fairness, its computing formula is: K ₁(i)=BPAF_T ₁(i) * BPAF_R ₃(i)+BPAF_T ₃(i) * BPAF_R ₁(i).

7. a kind of multi-service OFDM based on evidence theory according to claim 6 strides a layer dynamic resource allocation method, it is characterized in that described step 6 specifically: for the packet of different brackets, dispatch according to scheduling grade order from high to low; For same grade data package, when the scheduling grade of packet is anxious timing, is divided into according to the packet after merging and suddenly assembles the descending order of elementary probability partition function value of closing and dispatch; When the scheduling grade of packet when fuzzy, the ascending order of elementary probability partition function value that is divided into fuzzy set according to the packet after merging is dispatched; Bide one's time for waiting when the scheduling grade of packet, be divided into the ascending order of elementary probability partition function value of waiting for set according to the packet after merging and dispatch.

Images