CN103079273A - Resource management method suitable for multilayer satellite system - Google Patents

Abstract

The invention relates to a resource management method suitable for a multilayer satellite system. The resource management method adopts the following design principle that high altitude satellites are used as reserved resources and are specially used for distributing resources for users with high priorities and high QOS (Quality of Service) requirements and distributing the resources for the user as required in a real-time manner; and meanwhile, low earth orbit satellites are combined and are used as ground network supplements, the characteristics of high-speed motion and the like of the low earth orbit satellites are considered and a method for dynamically distributing the resources for users with low priorities by using efficiency, cost and user publicity as evaluation indexes is designed. According to the design, the characteristics of the high altitude satellites and the low earth orbit satellites are combined; the priorities of the users are considered; the resources are distributed for the users with high priorities as required; a multi-target evaluation function is designed for the users with low priorities; and the multi-target evaluation function is nondimensionalized, so that in the resource management process, the evaluation indexes of other targets are not ignored due to an excessive function value of a certain index, and thus, a multi-target optimization method which has the advantages of short time delay, high bandwidth utilization rate and balanced resource distribution is integrally implemented.

Description

A kind of method for managing resource that is applicable to the multilayer satellite system

Technical field

The present invention relates to the method for managing resource of the multilayer satellite network of a kind of high rail satellite that is applied to have inter satellite link-low rail, a kind of QoS requirement that ensures different priority users specifically, to high-priority users distribution according to need resource, and design the method for managing resource of multiple-objection optimization for low priority user based on genetic Algorithm Design, belong to technical field of satellite communication.

Background technology

Satellite network has the ability of flexibly bandwidth on demand, and extensively various data communication service can be provided, and for the business demand of unpredictable bandwidth, satellite can be used as alternate channel.Compare with ground network, low-track satellite network has that topological structure dynamic change, communication link are frequently switched, propagation delay time is long, the satellite node number is relatively less, Constellation Network has systematicness and the characteristics such as harmonious, because satellite orbit is fixed, network topology change has again periodically and foreseeability simultaneously.High rail satellite mainly is responsible for exchange and the forwarding of packets of information, the position is fixed substantially, the cover time is also longer, when a large amount of user terminals is realized broadcasting service, have irreplaceable advantage, can provide access service for the user who possesses the relatively high power emitter simultaneously.These researchs for management method on the star had both brought certain difficulty.Following wireless communication system is all had higher requirement to traffic rate, service quality and type service, and the finiteness of Internet resources, i.e. the restriction such as system power, bandwidth, becoming affects the bottleneck that wireless communication system further develops.

The consumption degree of user's number and Internet resources is dynamic changes in the network, less when number of users, and the offered load amount is very high if the price of Internet resources is still kept when lighter, the resource quantity that end subscriber can be applied for is limited very low, thereby caused the idle of ample resources; When the number of users of system is larger, and network consumption is when comparatively serious, and the user can not get enough constraints to the application of resource, excessive resource so that the chance of network congestion greatly increase.Because Internet resources are always limited, it is the maximum service ability that total business volume that integrated service or Differentiated Services all must guarantee to enter network is no more than network.Therefore must be according to the QoS of customer requirement in the base station, judge type that this is professional in case allocation of channel resources may to different users.

The method of Resource Allocation in Networks can be divided into static allocation and dynamic assignment: static allocation will keep for session connection the resource of some, until conversation end; And dynamic assignment will be according to network condition and the application demand adjustresources quantity of distributing dynamically.Traditional Resource Management Algorithm has: Round Robin algorithm (polling dispatching algorithm), Max C/I(heap(ed) capacity/utilance) algorithm and M-LWDF(real time business) algorithm etc.Round Robin algorithm ensures the fairness between the user, but do not consider channel the time become characteristics, do not consider to exist different service quality to need situation between the user yet, can not satisfy the requirement of part communication system in a lot of situations.The design object of Max C/I algorithm is to obtain maximum power system capacity and the highest resource utilization, has abandoned the fairness between the user, and the user in the best state in each time slot selective channel serves, thereby reaches high as far as possible transmission rate.M-LWDF algorithm target is the quality of service requirement that as far as possible ensures the user on the basis of system resource taking full advantage of, and can take into account fairness and transmission rate, but variation that but can not the dynamically adapting Radio Link, and the wireless channel utilance is not high.

According to existing resource management method described above being not suitable in the inserting of satellite network application, along with the growing diversification to resource requirement of network, the user requires the finiteness of (QoS) and Internet resources itself with high quality-of-service to becoming increasingly abundant of network resource requirement, will become affects the bottleneck that wireless communication system further develops.According to high low orbit satellite own characteristic, and different user is different to the qos requirement of resource requirement, when the design resource management algorithm, realize the maximization of network resource utilization, here we can not ignore the demand of high user's high priority and high QoS, can not consider domestic consumer, therefore want the two to take into account in the design of resource management scheme.

Summary of the invention

Technical problem: in view of the user mainly contains efficient, cost and fairness to the QoS parameter of network resource requirement, general evaluation index take throughput, average packet time-delay, packet loss rate, resource utilization etc. as network efficiency.These optimization problems are interrelated conflicting again, the invention provides in a kind of multilayer satellite system (high rail satellite and low orbit satellite) access network the multiple-objection optimization method for managing resource based on genetic algorithm, in the As soon as possible Promising Policy different priority users in network under the prerequisite to the resource different QoS requirements, the maximization network resource utilization, and solved the Resources Management of multi-user under the high dynamic topological structure.

Technical scheme: according to above characteristic, we consider the advantage of network dynamic topology structure, User Priority and resource regulating method, propose a kind of method for managing resource of multiple-objection optimization of the consideration User Priority based on genetic algorithm.Its design principle is: take high rail satellite as reserving resource, be responsible for high preferential high QoS demand user's resource specially and distribute, be user's real-time Resources allocation as required; Replenish as ground network in conjunction with low orbit satellite simultaneously, consider the characteristics such as himself high-speed motion, design a kind of method of the low priority user dynamic assignment resource take efficient, cost and user's public character as evaluation index.This design combines high rail satellite and low orbit satellite characteristics, and considered User Priority, high-priority users is distributed according to need, to low priority user design Multi Goal Opinion Function, and it is carried out nondimensionalization, so that when resource management not because of the excessive evaluation index of ignoring all the other targets of the functional value of certain index, thereby realize that on the whole time-delay is less, bandwidth availability ratio is high and the Multipurpose Optimal Method of resource partition equilibrium.This algorithm had both ensured the QoS of high-priority users, ensured again the fairness between the low priority user, under the prerequisite that satisfies the different service quality needs situation of all users, improved as far as possible network resource utilization, optimized the whole network resource.

The present invention is applied in the method for managing resource of multilayer satellite system, and described multilayer satellite system comprises a high rail satellite at least, some low orbit satellite groups with inter satellite link, and the customer group of some different access waies; Be mainly used in solving different priority users problem of Resources allocation how in the limited situation of Internet resources, its concrete steps are as follows:

Step 1. initial time is waited for user access network;

Step 2. judges whether the user is military user, is to turn step 3, otherwise, turn step 6;

Step 3. accesses high rail satellite link with military user;

Step 4. judges that current high rail satellite has or not available resources, allowable resource is arranged then by being required to be user resource allocation, until waiting list is when be empty, high rail satellite resource distributes end;

Step 5. does not have resource can divide timing, and the user enters high rail satellite resource waiting list, waits for the resource distribution, turns step 4;

Step 6. accesses the low orbit satellite link with domestic consumer;

Step 7. judges whether current time is new snapshot, is to turn step 8, the no step 10 that turns;

Step 8. is got new snapshot, upgrades network topology structure figure;

Step 9. ground control centre is according to the recurrence period of satellite characteristics of motion, current topological result and access customer number, allocation of network resources,

Step 10. is judged User Priority, and the high-priority users for access then allows to take bandwidth reserved; If without bandwidth reserved, then allow to seize current low priority user bandwidth;

Step 11. then only allows to distribute common remaining bandwidth for the low priority user of access; If without remaining bandwidth, enter and wait in line formation execution poll;

Step 12. is current to be not new snapshot, does not then change current Resource Allocation Formula, judges to have or not available surplus resources, then distributes surplus resources if having, if without surplus resources, enters and waits in line formation execution poll;

Step 13. is user resource allocation according to current scheme;

Step 14. is waited for next constantly arrival of signal.

The allocation of network resources detailed process is in the step 9:

1). some access users, sent service request to low orbit satellite respectively, the analysis user demand, matrix b represents that the user is to the bandwidth demand amount of low orbit satellite resource, wherein b ⁱ _jExpression user i needs the bandwidth of low orbit satellite j,

2). nondimensionalization time-delay evaluation function f _t

The resource of user's request is more, the time of exclusive resource is longer so, consider service quality problem, estimate the priority that the user applies for resource with time-delay, i user assigned in required time of whole resources and the whole network the required time T of user compare nondimensionalization time-delay evaluation function f _t, t wherein ⁱ _KThe time from K low orbit satellite Gains resources of expression user i, q ⁱ _KRepresent that i user needs from K low orbit satellite Gains resources b ⁱ _KThe bandwidth from K low orbit satellite Gains resources of expression user i, T represents the required time of user in the whole network;

$f_t=\frac{t_k^i}{t_k^i+T}=\frac{q_k^i/b_k^i}{q_k^i/b_k^i+T}---\left(2\right)$

3). nondimensionalization bandwidth evaluation function f _b

The resource of user's request is more, and the total resources occupation proportion is higher so, considers service quality problem, and we can estimate the priority that the user applies for resource with the required time-delay of user.I user assigned to the required resource bandwidth B of user compares in the required bandwidth of whole resources and the whole network, nondimensionalization time-delay evaluation function f _b, q wherein ⁱ _KRepresent that i user needs from K low orbit satellite Gains resources, B represents user's resource requirement bandwidth in the whole network;

$f_b=\frac{b_k^i}{b_k^i+B}=\frac{b_k^i}{b_k^i+\underset{k=1}{\overset{k}{\mathrm{\Σ}}}{q}_k^i}---\left(3\right)$

4). nondimensionalization Cost Evaluation function f _c

The resource of user's request is directly proportional with its spent cost, there is multiple link form in the access network, each is inconsistent for its cost, originally estimated the priority that the user applies for resource with the one-tenth that user's resource requirement expends, i user assigned to the required cost C of user compares in the required cost of whole resources and the whole network, nondimensionalization time-delay evaluation function f _c, c wherein ⁱ _KRepresent that i user is from the cost of K low orbit satellite Gains resources, b ⁱ _KThe bandwidth from K low orbit satellite Gains resources of expression user i, C represents the required cost of user in the whole network;

$f_c=\frac{c_k^i{b}_k^i}{c_k^i{b}_k^i+C}=\frac{c_k^i{b}_k^i}{c_k^i{b}_k^i+\underset{i=1}{\overset{n}{\mathrm{\Σ}}}\underset{k=1}{\overset{k}{\mathrm{\Σ}}}{c}_k^i{q}_k^i}---\left(4\right)$

5). nondimensionalization User Priority evaluation function f _p

User and business are defined the level, access user according to priority is divided into emergency user, real-time, interactive user, domestic consumer from high to low successively.The User Priority part here is relevant with the implementation status of customer service, and the time of the resource to be used such as customer service is more of a specified duration, and then the priority of this customer service is just higher; A part is to use the situation of resource relevant with customer service, in emergency circumstances identical, oneself is through higher than the priority of the customer service of also not using resource in the customer service of using resource, and short ratio resource service time resource service time, long priority was high; Another part is own relevant with customer service, and namely in order to distinguish the execution sequence of different user business, artificial or system adds a value;

f _p＝a*δR _ole+b*δ _Res+c*δ _Wait(0<a，b，c<1;a+b+c=1) （5）

F wherein _pBe the User Priority evaluation function; δ _RoleExpression user's initial priority value is divided into Three Estate according to user role: emergency user, real-time, interactive user, domestic consumer; δ _ResExpression user requested service takies the situation of resource, δ _WaitThe pending times such as expression customer service; A, b, c are different user weight in User Priority evaluation function value.

6). multiple objective function

One of processing mode of multiple-objection optimization is exactly to say that target is nondimensional, so just has comparativity; The span of the evaluation function of each target is more close, thereby avoids a certain target function value excessive and cover the evaluation function value of all the other targets.Below four evaluation function f of our faces _t, f _c, f _b, f _pCorresponding weight is made as w _t, w _c, w _b, w _p, satisfy w _t+ w _c+ w _b+ w _p=1, target function is so

Min(F)=Min(w _t×f _t+w _t×f _t+w _t×f _t+w _t×f _t+w _t×f _t) （6）

s.t. $\left\{\begin{array}{c}\underset{k=1}{\overset{k}{\mathrm{\&Sigma;}}}{t}_k^i\&le;\max \left(t\right)\\ \underset{k=1}{\overset{k}{\mathrm{\&Sigma;}}}{b}_k^i\&le;\max \left(b\right)\\ \underset{k=1}{\overset{k}{\mathrm{\&Sigma;}}}{c}_k^i\&le;\max \left(c\right)\end{array}\right.$

Wherein, time-delay, bandwidth and cost all can not surpass practical application time-delay maximum max (t), maximum value bandwidth max (b), cost maximum max (c);

7). determine user's Request Priority, the user enters waiting list with certain priority, waits for bandwidth resources.

Beneficial effect: the present invention proposes a kind of method for managing resource that is applicable to the multilayer satellite system, be mainly used in solving multilayer satellite system net environment, when providing limited Internet resources for different priority users, the problem such as maximization network resource utilization how.

Usually the Resource Management Algorithm on the satellite system usually is difficult to take into account the high and domestic consumer's QoS requirement of high-priority users, often with sacrifice the two one of the raising of acquisition performance.The present invention has considered the demand of different priority users, under the prerequisite of meeting consumers' demand, and the maximization network resource utilization.Specifically: the reservation resource take high rail as high-priority users, distribution according to need resource; As in real time domestic consumer's resource management, and ensure user fairness and efficient based on the genetic Algorithm Design multi-objective optimization algorithm with low orbit satellite.

The method had both ensured the QoS of high-priority users, ensured again the fairness between the low priority user, under the prerequisite that satisfies the different service quality needs situation of all users, improved as far as possible network resource utilization, optimized the whole network resource.

Description of drawings

Fig. 1 is the block diagram according to multilayer satellite system formant of the present invention, mainly is comprised of high rail, the low orbit satellite with inter satellite link and dissimilar three parts such as terminal use.

Fig. 2 is the block diagram according to LEO Satellite formant of the present invention, mainly partly is comprised of base station controller, subbase station and terminal use etc.

Fig. 3 is that low orbit satellite is under a snapshot, based on the schematic diagram of the multiple-objection optimization dynamic resource management algorithm of genetic algorithm.

Fig. 4 is the resource management flow chart that is applied to whole multilayer satellite system network.

Embodiment

One, architecture

Based on multilayer satellite network architecture of the present invention as shown in Figure 1, specifically composed as follows: key territory, input field and three parts of user domain.Key routed domain is comprised of several the high rail satellites of backbone, mainly is responsible for exchange and the forwarding of packets of information, can provide access service for the user who possesses the relatively high power emitter simultaneously.Input field is comprised of many low rail constellations, is responsible for the intelligence access of different access objects and different business stream, comprises that specifically the user authenticates, calls out determining etc. of access control, channel allocation, the grade of service and priority.User domain is comprised of dissimilar user, comprising: fixed-line subscriber, mobile subscriber, airborne user and vehicle-mounted user etc.

● high rail satellite system

Main exchange and the forwarding of being responsible for packets of information can provide access service for the user who possesses the relatively high power emitter simultaneously.The main task here is to be responsible for the resource management of the high QoS demand of high-priority users, by being required to be user resource allocation.

● LEO Satellite

Low rail system partly is comprised of space segment, ground segment and user segment usually, and space segment refers to satellite constellation, and ground segment comprises satellite control center, the network control center, gateway station (or being called juncture station) and satellite portable terminal.Being responsible for the real-time resource bid of low priority user distributes.As shown in Figure 2.

● customer group

User domain is comprised of dissimilar user, comprising: fixed-line subscriber, mobile subscriber, airborne user and vehicle-mounted user etc.Different user priority is different, and the service request of different stage is arranged.

Two, method flow

1, current snapshot bends down rail inserting of satellite method for managing resource flow process, as shown in Figure 3.

Receive a collection of user's service request when the low orbit satellite access network, understand for convenient, suppose here K low orbit satellite, between have some inter satellite links, N end subscriber.q ⁱ _KRepresent that i user needs from K low orbit satellite Gains resources; c ⁱ _KRepresent that i user is from the cost of K low orbit satellite Gains resources; b ⁱ _KThe bandwidth from K low orbit satellite Gains resources of expression user i; t ⁱ _KThe time from K low orbit satellite Gains resources of expression user i.When large-capacity user proposes resource request to the base station, system will determine Subscriber Queue priority by multi-objective Optimization Genetic Algorithm, and its concrete steps are as follows:

The some access users of step 1. have sent service request to low orbit satellite respectively, the analysis user demand.Matrix b represents that the user is to the bandwidth demand amount of low orbit satellite resource, wherein b ⁱ _jExpression user i needs the bandwidth of low orbit satellite j.

Step 2. nondimensionalization time-delay evaluation function f _t

The resource of user's request is more, so the time of exclusive resource longer, consider service quality problem, we can estimate the priority that the user applies for resource with time-delay.I user assigned in required time of whole resources and the whole network the required time T of user compare nondimensionalization time-delay evaluation function f _t, t wherein ⁱ _KThe time from K low orbit satellite Gains resources of expression user i, q ⁱ _KRepresent that i user needs from K low orbit satellite Gains resources b ⁱ _KThe bandwidth from K low orbit satellite Gains resources of expression user i, T represents the required time of user in the whole network.

$f_t=\frac{t_k^i}{t_k^i+T}=\frac{q_k^i/b_k^i}{q_k^i/b_k^i+T}---\left(2\right)$

Step 3. nondimensionalization bandwidth evaluation function f _b

The resource of user's request is more, and the total resources occupation proportion is higher so, considers service quality problem, and we can estimate the priority that the user applies for resource with the required time-delay of user.I user assigned to the required resource bandwidth B of user compares in the required bandwidth of whole resources and the whole network, nondimensionalization time-delay evaluation function f _b, q wherein ⁱ _KRepresent that i user needs from K low orbit satellite Gains resources, B represents user's resource requirement bandwidth in the whole network.

$f_b=\frac{b_k^i}{b_k^i+B}=\frac{b_k^i}{b_k^i+\underset{k=1}{\overset{k}{\mathrm{\&Sigma;}}}{q}_k^i}---\left(3\right)$

Step 4. nondimensionalization Cost Evaluation function f _c

The resource of user's request is directly proportional with its spent cost, has multiple link form in the access network, and each is inconsistent for its cost, and we can estimate the priority that the user applies for resource originally with the one-tenth that user's resource requirement expends.I user assigned to the required cost C of user compares in the required cost of whole resources and the whole network, nondimensionalization time-delay evaluation function f _c, c wherein ⁱ _KRepresent that i user is from the cost of K low orbit satellite Gains resources, b ⁱ _KThe bandwidth from K low orbit satellite Gains resources of expression user i, C represents the required cost of user in the whole network.

$f_c=\frac{c_k^i{b}_k^i}{c_k^i{b}_k^i+C}=\frac{c_k^i{b}_k^i}{c_k^i{b}_k^i+\underset{i=1}{\overset{n}{\mathrm{\&Sigma;}}}\underset{k=1}{\overset{k}{\mathrm{\&Sigma;}}}{c}_k^i{q}_k^i}---\left(4\right)$

Step 5. nondimensionalization User Priority evaluation function f _p

User and business are defined the level, access user according to priority is divided into emergency user, real-time, interactive user, domestic consumer from high to low successively.The User Priority part here is relevant with the implementation status of customer service, and the time of the resource to be used such as customer service is more of a specified duration, and then the priority of this customer service is just higher; A part is to use the situation of resource relevant with customer service, in emergency circumstances identical, oneself is through higher than the priority of the customer service of also not using resource in the customer service of using resource, and short ratio resource service time resource service time, long priority was high; Another part is own relevant with customer service, and namely in order to distinguish the execution sequence of different user business, artificial or system adds a value.

f _p＝a*δR _ole+b*δ _Res+c*δ _Wait(0<a，b，c<1;a+b+c=1)（5）

F wherein _pBe the User Priority evaluation function; δ _RoleExpression user's initial priority value is divided into Three Estate according to user role: emergency user, real-time, interactive user, domestic consumer; δ _ResExpression user requested service takies the situation of resource, δ _WaitThe pending times such as expression customer service; A, b, c are different user weight in User Priority evaluation function value.

Step 6. multiple objective function

One of processing mode of multiple-objection optimization is exactly to say that target is nondimensional, so just has comparativity; The span of the evaluation function of each target is more close, thereby avoids a certain target function value excessive and cover the evaluation function value of all the other targets.Below four evaluation function f of our faces _t, f _c, f _b, f _pCorresponding weight is made as w _t, w _c, w _b, w _p, satisfy w _t+ w _c+ w _b+ w _p=1, target function is so

Min(F)=Min(w _t×f _t+w _t×f _t+w _t×f _t+w _t×f _t+w _t×f _t)（6）

s.t. $\left\{\begin{array}{c}\underset{k=1}{\overset{k}{\mathrm{\&Sigma;}}}{t}_k^i\&le;\max \left(t\right)\\ \underset{k=1}{\overset{k}{\mathrm{\&Sigma;}}}{b}_k^i\&le;\max \left(b\right)\\ \underset{k=1}{\overset{k}{\mathrm{\&Sigma;}}}{c}_k^i\&le;\max \left(c\right)\end{array}\right.$

Wherein, time-delay, bandwidth and cost all can not surpass practical application time-delay maximum max (t), maximum value bandwidth max (b), cost maximum max (c).

Step 7. is determined user's Request Priority, and the user enters waiting list with certain priority, waits for base station controller distribution bandwidth resource.

When step 8. enters formation when new user's request, repeating step.

Example: be applied to the method for managing resource of whole multilayer satellite system, schematic diagram as shown in Figure 4, its concrete steps are as follows:

Step 1. initial time is waited for user access network.

Step 2. judges whether the user is military user, is to turn step 3, otherwise, turn step 6.

Step 3. accesses high rail satellite link with military user.

Step 4. judges that current high rail satellite has or not available resources, allowable resource is arranged then by being required to be user resource allocation

Step 5. does not have resource to distribute, and the user enters high rail satellite resource waiting list, waits for the resource distribution

Step 6. accesses the low orbit satellite link with domestic consumer

Step 7. judges whether current time is new snapshot, is to turn step 8, the no step 10 that turns

Step 8. is got new snapshot, upgrades network topology structure figure

Step 9. ground control centre is according to the recurrence period of satellite characteristics of motion, current topological result and access customer number, allocation of network resources, and its detailed process is:

(1). some access users, sent service request to low orbit satellite respectively, the analysis user demand.Matrix b represents that the user is to the bandwidth demand amount of low orbit satellite resource, wherein b ⁱ _jExpression user i needs the bandwidth of low orbit satellite j.

(2). nondimensionalization time-delay evaluation function f _t

The resource of user's request is more, so the time of exclusive resource longer, consider service quality problem, we can estimate the priority that the user applies for resource with time-delay.I user assigned in required time of whole resources and the whole network the required time T of user compare nondimensionalization time-delay evaluation function f _t, t wherein ⁱ _KThe time from K low orbit satellite Gains resources of expression user i, q ⁱ _KRepresent that i user needs from K low orbit satellite Gains resources b ⁱ _KThe bandwidth from K low orbit satellite Gains resources of expression user i, T represents the required time of user in the whole network.

$f_t=\frac{t_k^i}{t_k^i+T}=\frac{q_k^i/b_k^i}{q_k^i/b_k^i+T}---\left(2\right)$

(3). nondimensionalization bandwidth evaluation function f _b

The resource of user's request is more, and the total resources occupation proportion is higher so, considers service quality problem, and we can estimate the priority that the user applies for resource with the required time-delay of user.I user assigned to the required resource bandwidth B of user compares in the required bandwidth of whole resources and the whole network, nondimensionalization time-delay evaluation function f _b, q wherein ⁱ _KRepresent that i user needs from K low orbit satellite Gains resources, B represents user's resource requirement bandwidth in the whole network.

$f_b=\frac{b_k^i}{b_k^i+B}=\frac{b_k^i}{b_k^i+\underset{k=1}{\overset{k}{\mathrm{\&Sigma;}}}{q}_k^i}---\left(3\right)$

(4). nondimensionalization Cost Evaluation function f _c

The resource of user's request is directly proportional with its spent cost, has multiple link form in the access network, and each is inconsistent for its cost, and we can estimate the priority that the user applies for resource originally with the one-tenth that user's resource requirement expends.I user assigned to the required cost C of user compares in the required cost of whole resources and the whole network, nondimensionalization time-delay evaluation function f _c, c wherein ⁱ _KRepresent that i user is from the cost of K low orbit satellite Gains resources, b ⁱ _KThe bandwidth from K low orbit satellite Gains resources of expression user i, C represents the required cost of user in the whole network.

$f_c=\frac{c_k^i{b}_k^i}{c_k^i{b}_k^i+C}=\frac{c_k^i{b}_k^i}{c_k^i{b}_k^i+\underset{i=1}{\overset{n}{\mathrm{\&Sigma;}}}\underset{k=1}{\overset{k}{\mathrm{\&Sigma;}}}{c}_k^i{q}_k^i}---\left(4\right)$

(5). nondimensionalization User Priority evaluation function f _p

User and business are defined the level, access user according to priority is divided into emergency user, real-time, interactive user, domestic consumer from high to low successively.The User Priority part here is relevant with the implementation status of customer service, and the time of the resource to be used such as customer service is more of a specified duration, and then the priority of this customer service is just higher; A part is to use the situation of resource relevant with customer service, in emergency circumstances identical, oneself is through higher than the priority of the customer service of also not using resource in the customer service of using resource, and short ratio resource service time resource service time, long priority was high; Another part is own relevant with customer service, and namely in order to distinguish the execution sequence of different user business, artificial or system adds a value.

f _p＝a*δ _Role+b*δ _Res+c*δ _Wait(0<a，b，c<1;a+b+c=1)（5）

F wherein _pBe the User Priority evaluation function; δ _RoleExpression user's initial priority value is divided into Three Estate according to user role: emergency user, real-time, interactive user, domestic consumer; δ _ResExpression user requested service takies the situation of resource, δ _WaitThe pending times such as expression customer service; A, b, c are different user weight in User Priority evaluation function value.

(6). multiple objective function

One of processing mode of multiple-objection optimization is exactly to say that target is nondimensional, so just has comparativity; The span of the evaluation function of each target is more close, thereby avoids a certain target function value excessive and cover the evaluation function value of all the other targets.Below four evaluation function f of our faces _t, f _c, f _b, f _pCorresponding weight is made as w _t, w _c, w _b, w _p, satisfy w _t+ w _c+ w _b+ w _p=1, target function is so

Min(F)=Min(w _t×f _t+w _t×f _t+w _t×f _t+w _t×f _t+w _t×f _t)（6）

s.t. $\left\{\begin{array}{c}\underset{k=1}{\overset{k}{\mathrm{\&Sigma;}}}{t}_k^i\&le;\max \left(t\right)\\ \underset{k=1}{\overset{k}{\mathrm{\&Sigma;}}}{b}_k^i\&le;\max \left(b\right)\\ \underset{k=1}{\overset{k}{\mathrm{\&Sigma;}}}{c}_k^i\&le;\max \left(c\right)\end{array}\right.$

Wherein, time-delay, bandwidth and cost all can not surpass practical application time-delay maximum max (t), maximum value bandwidth max (b), cost maximum max (c).

(7). determine user's Request Priority, the user enters waiting list with certain priority, waits for bandwidth resources.

Step 10. is judged User Priority, and the high-priority users for access then allows to take bandwidth reserved; If without bandwidth reserved, then allow to seize current low priority user bandwidth.

Step 11. then only allows to distribute common remaining bandwidth for the low priority user of access; If without remaining bandwidth, enter and wait in line formation execution poll.

Step 12. is current to be not new snapshot, does not then change current Resource Allocation Formula, judges to have or not available surplus resources, then distributes surplus resources if having, if without surplus resources, enters and waits in line formation execution poll.

Step 13. is user resource allocation according to current scheme.

Step 14. is waited for next constantly arrival of signal.

Claims (2)

1. a method for managing resource that is applied to the multilayer satellite system is characterized in that described multilayer satellite system comprises a high rail satellite at least, some low orbit satellite groups with inter satellite link, and the customer group of some different access waies; Be mainly used in solving different priority users problem of Resources allocation how in the limited situation of Internet resources, its concrete steps are as follows:

Step 1. initial time is waited for user access network;

Step 2. judges whether the user is military user, is to turn step 3, otherwise, turn step 6;

Step 3. accesses high rail satellite link with military user;

Step 4. judges that current high rail satellite has or not available resources, allowable resource is arranged then by being required to be user resource allocation, until waiting list is when be empty, high rail satellite resource distributes end;

Step 5. does not have resource can divide timing, and the user enters high rail satellite resource waiting list, waits for the resource distribution, turns step 4;

Step 6. accesses the low orbit satellite link with domestic consumer;

Step 7. judges whether current time is new snapshot, is to turn step 8, the no step 10 that turns;

Step 8. is got new snapshot, upgrades network topology structure figure;

Step 9. ground control centre is according to the recurrence period of satellite characteristics of motion, current topological result and access customer number, allocation of network resources,

Step 10. is judged User Priority, and the high-priority users for access then allows to take bandwidth reserved; If without bandwidth reserved, then allow to seize current low priority user bandwidth;

Step 11. then only allows to distribute common remaining bandwidth for the low priority user of access; If without remaining bandwidth, enter and wait in line formation execution poll;

Step 12. is current to be not new snapshot, does not then change current Resource Allocation Formula, judges to have or not available surplus resources, then distributes surplus resources if having, if without surplus resources, enters and waits in line formation execution poll;

Step 13. is user resource allocation according to current scheme;

Step 14. is waited for next constantly arrival of signal.

2. the method for managing resource that is applied to the multilayer satellite system according to claim 1 is characterized in that, the allocation of network resources detailed process is in the step 9:

1). some access users, sent service request to low orbit satellite respectively, the analysis user demand, matrix b represents that the user is to the bandwidth demand amount of low orbit satellite resource, wherein b ⁱ _jExpression user i needs the bandwidth of low orbit satellite j,

2). nondimensionalization time-delay evaluation function f _t

The resource of user's request is more, the time of exclusive resource is longer so, consider service quality problem, estimate the priority that the user applies for resource with time-delay, i user assigned in required time of whole resources and the whole network the required time T of user compare nondimensionalization time-delay evaluation function f _t, t wherein ⁱ _KThe time from K low orbit satellite Gains resources of expression user i, q ⁱ _KRepresent that i user needs from K low orbit satellite Gains resources b ⁱ _KThe bandwidth from K low orbit satellite Gains resources of expression user i, T represents the required time of user in the whole network;

$f_t=\frac{t_k^i}{t_k^i+T}=\frac{q_k^i/b_k^i}{q_k^i/b_k^i+T}---\left(2\right)$

3). nondimensionalization bandwidth evaluation function f _b

The resource of user's request is more, and the total resources occupation proportion is higher so, considers service quality problem, and we can estimate the priority that the user applies for resource with the required time-delay of user.I user assigned to the required resource bandwidth B of user compares in the required bandwidth of whole resources and the whole network, nondimensionalization time-delay evaluation function f _b, q wherein ⁱ _KRepresent that i user needs from K low orbit satellite Gains resources, B represents user's resource requirement bandwidth in the whole network;

$f_b=\frac{b_k^i}{b_k^i+B}=\frac{b_k^i}{b_k^i+\underset{k=1}{\overset{k}{\mathrm{\&Sigma;}}}{q}_k^i}---\left(3\right)$

4). nondimensionalization Cost Evaluation function f _c

The resource of user's request is directly proportional with its spent cost, there is multiple link form in the access network, each is inconsistent for its cost, originally estimated the priority that the user applies for resource with the one-tenth that user's resource requirement expends, i user assigned to the required cost C of user compares in the required cost of whole resources and the whole network, nondimensionalization time-delay evaluation function f _c, c wherein ⁱ _KRepresent that i user is from the cost of K low orbit satellite Gains resources, b ⁱ _KThe bandwidth from K low orbit satellite Gains resources of expression user i, C represents the required cost of user in the whole network;

$f_c=\frac{c_k^i{b}_k^i}{c_k^i{b}_k^i+C}=\frac{c_k^i{b}_k^i}{c_k^i{b}_k^i+\underset{i=1}{\overset{n}{\mathrm{\&Sigma;}}}\underset{k=1}{\overset{k}{\mathrm{\&Sigma;}}}{c}_k^i{q}_k^i}---\left(4\right)$

5). nondimensionalization User Priority evaluation function f _p

User and business are defined the level, access user according to priority is divided into emergency user, real-time, interactive user, domestic consumer from high to low successively.The User Priority part here is relevant with the implementation status of customer service, and the time of the resource to be used such as customer service is more of a specified duration, and then the priority of this customer service is just higher; A part is to use the situation of resource relevant with customer service, in emergency circumstances identical, oneself is through higher than the priority of the customer service of also not using resource in the customer service of using resource, and short ratio resource service time resource service time, long priority was high; Another part is own relevant with customer service, and namely in order to distinguish the execution sequence of different user business, artificial or system adds a value;

f _p＝a*δ _Role+b*δ _Res+c*δ _Wait(0<a，b，c<1;a+b+c=1)（5）

F wherein _pBe the User Priority evaluation function; δ _RoleExpression user's initial priority value is divided into Three Estate according to user role: emergency user, real-time, interactive user, domestic consumer; δ _ResExpression user requested service takies the situation of resource, δ _WaitThe pending times such as expression customer service; A, b, c are different user weight in User Priority evaluation function value.

6). multiple objective function

One of processing mode of multiple-objection optimization is exactly to say that target is nondimensional, so just has comparativity; The span of the evaluation function of each target is more close, thereby avoids a certain target function value excessive and cover the evaluation function value of all the other targets.Below four evaluation function f of our faces _t, f _c, f _b, f _pCorresponding weight is made as w _t, w _c, w _b, w _p, satisfy w _t+ w _c+ w _b+ w _p=1, target function is so

Min(F)=Min(w _t×f _t+w _t×f _t+w _t×f _t+w _t×f _t+w _t×f _t)（6）

s.t. $\left\{\begin{array}{c}\underset{k=1}{\overset{k}{\mathrm{\&Sigma;}}}{t}_k^i\&le;\max \left(t\right)\\ \underset{k=1}{\overset{k}{\mathrm{\&Sigma;}}}{b}_k^i\&le;\max \left(b\right)\\ \underset{k=1}{\overset{k}{\mathrm{\&Sigma;}}}{c}_k^i\&le;\max \left(c\right)\end{array}\right.$

Wherein, time-delay, bandwidth and cost all can not surpass practical application time-delay maximum max (t), maximum value bandwidth max (b), cost maximum max (c);

7). determine user's Request Priority, the user enters waiting list with certain priority, waits for bandwidth resources.

Images