CN101516093B - Method and device for judging comprehensive priority of mobile communication system - Google Patents

Method and device for judging comprehensive priority of mobile communication system Download PDF

Info

Publication number
CN101516093B
CN101516093B CN2009100768636A CN200910076863A CN101516093B CN 101516093 B CN101516093 B CN 101516093B CN 2009100768636 A CN2009100768636 A CN 2009100768636A CN 200910076863 A CN200910076863 A CN 200910076863A CN 101516093 B CN101516093 B CN 101516093B
Authority
CN
China
Prior art keywords
user
priority
value
priority value
professional
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Expired - Fee Related
Application number
CN2009100768636A
Other languages
Chinese (zh)
Other versions
CN101516093A (en
Inventor
王卫东
张英海
周瑶
赵新蕾
王宗文
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Beijing University of Posts and Telecommunications
Original Assignee
Beijing University of Posts and Telecommunications
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Beijing University of Posts and Telecommunications filed Critical Beijing University of Posts and Telecommunications
Priority to CN2009100768636A priority Critical patent/CN101516093B/en
Publication of CN101516093A publication Critical patent/CN101516093A/en
Application granted granted Critical
Publication of CN101516093B publication Critical patent/CN101516093B/en
Expired - Fee Related legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Landscapes

  • Data Exchanges In Wide-Area Networks (AREA)

Abstract

The invention discloses a method and a device for judging comprehensive priority of a mobile communication system. The method includes the following steps: service QoS parameters of each service and user call type parameters of the service are searched; static comprehensive priority of each service is calculated according to service attribute and the user call type; and dynamic comprehensive priority value of each service is obtained by calculation by combining with multi-aspect factors such as network, service and user. According to the network condition, the algorithm of static or dynamic comprehensive priority is properly selected as the judgment basis of the service access service. The method and device for judging comprehensive priority of the mobile communication system provided in the invention comprehensively considers multi-aspect factors, provides and establishes static and dynamic comprehensive priority systems, and can carry out flexible and accurate tracking judgment on user service priority, thereby guaranteeing the comprehensive efficiency of the mobile communication system to be optimum and meeting the requirements of each party in the mobile communication system.

Description

The judging comprehensive priority method and apparatus of mobile communication system
Technical field
The present invention relates to mobile communication system priority decision technology, be meant a kind of judging comprehensive priority method and apparatus of mobile communication system especially.
Background technology
Along with combining of IP and mobile communication standard, future mobile communication system will be an all-IP transport communication system that supports the high-speed multimedia business.It will support the higher rate of information throughput and reduce system transmissions significantly to postpone, for the user provides more flexile individual business.Simultaneously, differentiation requires also strictness more to future mobile communication system to QoS.And priority for satisfying the different business demand, guarantees quality of services for users as a ring important among the QoS, and system is that miscellaneous service or different user have distributed service order.Particularly when system loading is higher, different priority is set to reducing the system transmissions cost, guaranteeing that QoS of survice seems and be even more important.
Toward research, priority can be classified from following two angles with according to this.According to the classification difference, priority can be divided into user call type priority and service attribute priority.User call type priority is the type of call according to the user, divides as accessed user, switching user, new calling party.In research in the past, accessed user's priority be higher than the back both, the priority of switching the user is higher than the new calling party of Target cell; Service attribute priority is the basic qos parameter according to business, and as decisions such as time delay, delay variation, loss rate, throughputs, wherein delay character is paid close attention to maximum factors.Such as, the priority of the real time business of delay sensitive being higher than non-real-time service, can guarantee real-time is required to provide high rank than higher business, user based on the priority location of time delay, obtaining the effect of real-time, interactive, thereby improve its transmission quality.
In addition, present studies show that, in order to obtain better service quality and to pay littler cost, the priority professional and user can change in transmission course.For example, the priority of non-real-time service can increase gradually along with the growth of waiting for service time.Whether change according to the transmission course medium priority, priority can be divided into static again and dynamic two kinds.Static priority refers to exist in the process in business, and is constant relatively from the system perspective User Priority, for example connect User Priority and has been higher than the switching User Priority, is higher than new User Priority, and active user priority is higher than the non real-time User Priority; And dynamic priority refers in service process, and User Priority changes according to the change of professional stand-by period in the transmission course, and for example the priority of switching the user along with the increase of stand-by period promotes gradually, is higher than similar user's priority.
There are some following limitations in existing priority localization method: 1, user call type priority and service attribute priority are broken up the consideration of coming, ignored the situation associated with each other of the two.Download the switching user's of (to time delay less demanding) priority and might not be higher than the voice or the video user of newly exhaling such as carrying out data file; Though 2 have set forth the thought of priority dynamic change in some document, considered the influence of stand-by period to user priority, but the variation of priority only limits to the non real-time user, do not consider the change of priority relativeness between non-real-time service and the real time business, and the stand-by period herein only be that the attend to the basic or the fundamental characteristics of body of working are set out, do not consider the characteristics of operation demand.3, be higher than when the not high non real-time of delay requirement connect user's priority when the priority of switching user or new user, the former can temporarily seize the resource that non real-time has met the user, access service, it is then out of service to meet the non real-time user, enter buffering area and wait for that this kind phenomenon is " the non real-time user dodges phenomenon ".Because the non-real-time data business will account for very big proportion in the future communications system, and real time business is too high with respect to the priority of non-real-time service at present, and this can cause that too much non-real-time service dodges phenomenon, and non real-time user's interests can not effectively be guaranteed.4, the priority influencing factor of existing priority thought consideration is limited, can not satisfy the requirement that the dynamic many indexs of future communications system QoS are weighed, and also need consider the influence of how new factor to priority.The measurement of QoS is dynamic many index systems in the future communications system: different business has the feature set of being made up of the various features parameter; Portable terminal all is among the continuous variation with the access cell status, and the parameter of multiple dynamic changes such as user's Move Mode, cell load state all exists direct influence to satisfying of QoS.Therefore, except the stand-by period, other factors can change the size of priority equally, and service priority should be integrated network, business, user's multiple factor dynamic change.
Under the situation that priority between the same line of business (as speech business) compares or other influencing factors variations are little, traditional priority locating rule still is suitable for, and does not need too much to consider the influence of other factors.But along with the introducing of the development of mobile communication technology, novel business and novel I P transmission technology, the class of business that future mobile communication system provided is more and more.The appearance of new business such as game on line, video telephone, Streaming Media, mobile TV, VoIP makes that tradition is only no longer suitable to the priority decision mechanism of time sensitivity based on customer call type or service attribute.Be in particular in:
(1) judgement of traditional priority mainly is to carry out at the difference of service attribute or type of call separately, and following class of business is various, the judgement of priority need combine user call type and service attribute two factors, just can be overthrown by the above-mentioned priority orders that both determine like this.As shown in Figure 1, when switch the user carry out non-real-time service, to delay requirement when not high, can be higher than the priority of switching the user to the real time new User Priority of delay sensitive.
(2) the future communications system can be the user the more business of horn of plenty is provided, and is also more and more higher to the requirement of QoS, and therefore, it is more careful that the division of priority needs, and should be able to reflect the demand of QoS of survice and the variation of status transmission.
(3) non-real-time service occupies very big proportion in future mobile communication system.For ensureing non real-time user's interests, should avoid too much " the non real-time user dodges phenomenon " occurring, break the routine of " priority of non-real-time service is lower than real time business ".
(4) design of priority mainly comes from the time sensitivity of user call type and service attribute, and is relatively simple.And future mobile communication system is complicated more, and the factor that influences priority can be more.
To sum up, the priority of future communications system determines that system should be a kind of dynamic, decision mechanism of taking all factors into consideration many-sided factor, need take all factors into consideration many-sided influences such as network environment, business, user.
Summary of the invention
In view of this, the present invention is directed to that mobile communication system all-IP, QoS are variable, the current situation of multi-service coexistence, propose a kind of dynamic, comprehensive mobile communication system priority decision method and device.
Judging comprehensive priority method based on above-mentioned purpose a kind of mobile communication system provided by the invention comprises:
A. collect each professional QoS of survice parameter and user call type parameter;
B. according to each professional service attribute priority value of each QoS of survice calculation of parameter, obtain each professional user call type priority value according to the user call type calculation of parameter;
C. the result according to step B calculates each professional comprehensive priority value.
Optionally, the described professional OoS parameter of this method comprises: peak transfer rate, user's average throughput;
Step B is described further to be comprised according to each professional service attribute priority value of each QoS of survice calculation of parameter:
The service attribute priority P of user i I-scCalculate by following formula
Figure G2009100768636D00041
Wherein, R iBe user's peak transfer rate, λ iBe user's average throughput, α iFor the priority weights of differentiated service, when user i is the real time business user, α iBy packet loss rate δ iThe maximum delay T that can tolerate with user i iDecision; z iFor preestablishing;
Step B is described to be obtained each professional user call type priority value according to the user call type calculation of parameter and further comprises:
User i type of call priority value P I, sctCalculate by following formula
Figure G2009100768636D00042
β wherein 1, β 2, β 3Be to have met the user, switch the user and newly exhaled the user for distinguishing, and the different weights of giving.
Optionally, the described comprehensive priority value of this method step C is calculated by following formula:
P i=w 1*P i,sc+w 2*P i,sct
Wherein, P iBe the static comprehensive priority value of user i, w 1And w 2Weighted value for service attribute priority and user call type priority.
Optionally, the described w of this method 1And w 2Calculate by following formula:
w 1 = 1 1 + e - ( P i , sc - P i , sct )
w 2 = 1 - 1 1 + e - ( P i , sc - P i , sct ) ,
Wherein, P I, scBe the service attribute priority value of user i, P I, sctBe user i type of call priority value.
Optionally, this method is if current network conditions is bad, and then step B further comprises:
Add up institute's patient maximum latency and actual stand-by period, user's Move Mode, the cell load state of each business;
Patient maximum latency of institute and actual stand-by period by each business, calculate each professional period of reservation of number priority value;
Judge user's Move Mode that each is professional, calculate each professional user's Move Mode priority value;
Can provide stock number by each professional resource requirement amount and access subzone network, calculate each professional user and intend inserting cell load priority of status value;
The service attribute priority value and the user call type priority value that further obtain among the step C by step B, and period of reservation of number priority value, user's Move Mode priority value and user intend inserting cell load priority of status value, calculates each professional comprehensive priority value.
Optionally, the described QoS of survice parameter of this method comprises: peak transfer rate, user's average throughput, the maximum delay that packet loss rate and user can tolerate;
Described period of reservation of number priority value is calculated by following formula:
P i , queue = T wait * K , T wait < T max n [ T wait + ( T wait - T max 2 ) m ] * K , T wait &GreaterEqual; T max n
M wherein, n=1,2........
P I, queueBe user i stand-by period priority value, T WaitBe actual stand-by period of user, T MaxBe the patient maximum delay of user, K is the weighted value of different business;
Described each professional user's Move Mode priority value is calculated by following formula:
Figure G2009100768636D00061
Wherein, P I, speedBe user's Move Mode priority value, ε 1, ε 2, ε 3User's priority valve when distinguishing corresponding high speed, low speed and inactive state;
Described user intends inserting cell load priority of status value and is calculated by following formula:
Figure G2009100768636D00062
η wherein 1>η 2
P I, celloadBe user i corresponding priorities value when the different districts load condition, η 1, η 2Give the user different priority for provide resource whether to satisfy professional required bandwidth according to the sub-district.
Optionally, described each the professional comprehensive priority value of this method is finally determined jointly by above-mentioned five factors, is obtained by following formula:
G i=k 1*P i,sc+k 2*P i,sct+k 3*P i,queue+k 4*P i,speed+k 5*P i,celload (i>=0)
Wherein, G iThe dynamic comprehensive priority value of representative of consumer i, P I, sctBe service attribute priority value P I, scWith user call type priority value, P I, queueBe period of reservation of number priority value, P I, speedBe user's Move Mode priority value, P I, celloadFor the user intends inserting cell load priority of status value, obtain each professional dynamic comprehensive priority value; k 1, k 2, k 3, k 4And k 5Be respectively the weighted value of service attribute, user call type, period of reservation of number, user's Move Mode and five factors of cell load state.
Optionally, the weighted value of the described service attribute of this method, user call type, period of reservation of number, user's Move Mode and five factors of cell load state adopts Fuzzy AHP to analyze definite.
Optionally, further comprise behind the described step C of this method:
D. insert Target cell according to the comprehensive priority value selective sequential optimal user that obtains.
Optionally, the described step D of this method further comprises:
According to described service attribute, user call type, period of reservation of number, user's Move Mode and five factor (r of cell load state 1, r 2, r 3, r 4, r 5), make relatively judging in twos between factor, establish γ Iji/ γ j, expression γ i, γ jTo the ratio of service selection influence degree, obtain parameter (r 1, r 2, r 3, r 4, r 5) fuzzy matrix A
A = 0.5 0.5 0.6 0.7 0.65 0.5 0.5 0.6 0.7 0.65 0.4 0.4 0.5 0.6 0.55 0.3 0.3 0.4 0.5 0.45 0.35 0.35 0.45 0.55 0.5
Wherein, element satisfies among the A: γ Ij>0, γ Ij+ γ Ji=1 (i ≠ j), γ Ii=0.5, and i, j=1,2....n.
To the fuzzy Complementary Judgement Matrix that obtains, calculate
V i &prime; = &Sigma; j = 1 n r ij + n 2 - 1 n ( n - 1 )
To V i' make normalized after, promptly get evaluating weight V={0.2225,0.2225,0.1975,0.1725,0.185};
Check the method for its consistency principle by the compatibility of fuzzy judgment matrix.
With V=(V 1, V 2, V 3, V 4, V 5) TWeight vectors for fuzzy judgment matrix A makes V Ij=V i/ (V i+ V j), then claim n rank matrix V *=(V Ij) N*nBe the eigenmatrix of A, calculate compatibility index I
I ( A , V * ) = 1 n 2 &Sigma; i = 1 n &Sigma; j = 1 n | A ij + V ij - 1 |
Provide each evaluation of indexes value of different schemes, { ω α(β) | α=1,2 ... .., n; β=1,2 ... .m}, wherein β represents the number of different schemes, sets up the fuzzy evaluation matrix of relative degree of membership by evaluating, and each parameter value is carried out standardization
s &alpha; ( &beta; ) = w &alpha; ( &beta; ) w max ( &alpha; ) + w min ( &alpha; )
Wherein, ω Max(α), ω Min(α) be respectively maximum, the minimum value of α evaluating; ω α(β) be evaluating value before the standardization, s α(β) for the relative degree of membership of standardization post-evaluation parameter;
Calculate total degree of membership
S ( &beta; ) = &Sigma; &alpha; v &alpha; s &alpha; ( &beta; )
α={ r wherein 1, r 2, r 3, r 4, r 5,
β=1,...m
V wherein αThe weight of expression evaluating, s α(β) the relative degree of membership of expression evaluating;
G=S(β) (15)
According to the priority of each customer service request, selecting the user of the professional correspondence of total degree of membership maximum in the formation is optimal user.
Optionally, further comprise behind the described step D of this method: upgrade all parameter sets, after priority is judged blanking time, proceed the circulation of next round.
Based on above-mentioned purpose, the present invention also provides a kind of judging comprehensive priority device of mobile communication system, comprising:
The priority critical parameter is collected module, collects each QoS of survice parameter and user call type parameter;
Service attribute priority computing module is used for according to each QoS of survice calculation of parameter service attribute priority value;
User call type priority computing module is used for according to user call type calculation of parameter user call type priority value;
The comprehensive priority computing module according to the result of calculation of service attribute priority computing module, user call type priority computing module, output, calculates each service integration priority value.
Optionally, this installs described professional OoS parameter and comprises: peak transfer rate, user's average throughput;
Described customer service attribute priority value calculates by following formula
Figure G2009100768636D00082
Wherein, P I, scBe the service attribute priority of user i, R iBe user's peak transfer rate, λ iBe user's average throughput, α iFor the priority weights of differentiated service, when user i is the real time business user, α iBy packet loss rate δ iThe maximum delay T that can tolerate with user i iDecision; z iFor preestablishing;
Described user call type priority value calculates by following formula
Figure G2009100768636D00091
Wherein, P I, sctBe user i type of call priority value, β 1, β 2, β 3Be to have met the user, switch the user and newly exhaled the user for distinguishing, and the different weights of giving.
Optionally, this installs described comprehensive priority value and calculates by following formula:
P i=w 1*P i,sc+w 2*P i,sct
Wherein, P iBe the static comprehensive priority value of user i, w 1And w 2Weighted value for service attribute priority and user call type priority.
Optionally, this device also comprises:
Subzone network condition judgement module is judged the subzone network state, for comprehensive priority computing module, priority judge that time interval computing module provides the subzone network state information;
Period of reservation of number priority computing module by the patient maximum latency of institute and the actual stand-by period of each business, calculates period of reservation of number priority;
User's Move Mode priority computing module is judged user's Move Mode that each is professional, calculates user's Move Mode priority;
Cell load priority of status computing module can provide stock number, calculation plot load condition priority by each professional resource requirement amount and access subzone network;
Comprise the static comprehensive priority calculating sub module in the described comprehensive priority computing module, by service attribute priority value that obtains and user call type priority value calculations static comprehensive priority value; Dynamic comprehensive priority calculating sub module, by the service attribute priority value that obtains, user call type priority value, period of reservation of number priority value, user's Move Mode priority value and user intend inserting cell load priority of status value, calculate each professional dynamic comprehensive priority value; And this comprehensive priority computing module is chosen corresponding submodule according to the subzone network state information that subzone network condition judgement module provides.
Optionally, this installs described QoS of survice parameter and comprises: peak transfer rate, user's average throughput, the maximum delay that packet loss rate and user can tolerate;
Described period of reservation of number priority value is calculated by following formula:
P i , queue = T wait * K , T wait < T max n [ T wait + ( T wait - T max 2 ) m ] * K , T wait &GreaterEqual; T max n
M wherein, n=1,2......
P I, queueBe user i stand-by period priority value, T WaitBe actual stand-by period of user, T MaxBe the patient maximum delay of user, K is the weighted value of different business;
Described each professional user's Move Mode priority value is calculated by following formula:
Figure G2009100768636D00102
Wherein, P I, speedBe user's Move Mode priority value, ε 1, ε 2, ε 3User's priority valve when distinguishing corresponding high speed, low speed and inactive state;
Described user intends inserting cell load priority of status value and is calculated by following formula:
Figure G2009100768636D00103
η wherein 1>η 2
P I, celloadBe user i corresponding priorities value when the different districts load condition, η 1, η 2Give the user different priority for provide resource whether to satisfy professional required bandwidth according to the sub-district.
Optionally, described each the professional comprehensive priority value of this device is finally determined jointly by above-mentioned five factors, is obtained by following formula:
G i=k 1*P i,sc+k 2*P i,sct+k 3*P i,queue+k 4*P i,speed+k 5*P i,celload (i>=0)
Wherein, G iThe dynamic comprehensive priority value of representative of consumer i, P I, sctBe service attribute priority value P I, scWith user call type priority value, P I, queueBe period of reservation of number priority value, P I, speedBe user's Move Mode priority value, P I, celloadFor the user intends inserting cell load priority of status value, obtain each professional dynamic comprehensive priority value; k 1, k 2, k 3, k 4And k 5Be respectively the weighted value of service attribute, user call type, period of reservation of number, user's Move Mode and five factors of cell load state.
Optionally, this weighted value of installing described service attribute, user call type, period of reservation of number, user's Move Mode and five factors of cell load state adopts Fuzzy AHP to analyze definite.
Optionally, this device also comprises: priority is judged time interval computing module, by the subzone network state information that subzone network condition judgement module provides, sets priority and judges the time interval.
From above as can be seen, the judging comprehensive priority method and apparatus of mobile communication system provided by the invention, taken all factors into consideration many-sided factor, propose and set up static, dynamic comprehensive priority system, can follow the tracks of judgement flexibly accurately to customer service priority, thereby guarantee mobile communication system comprehensive effectiveness optimum, satisfy the demand of each side in the mobile communication system, as user's subjective demand of accepting business, objectively demand and the operation demand considered from aspects such as cost control, market strategies of operator etc. in the professional transmission course.And avoided too much non real-time user to dodge phenomenon, guaranteed non real-time user's interests.
Description of drawings
Fig. 1 unites the schematic diagram of definite priority for prior art type of call and service attribute;
Fig. 2 is the schematic diagram that concerns of static comprehensive priority of the present invention and dynamic comprehensive priority;
Fig. 3 is an embodiment of the invention priority structure system schematic diagram;
Fig. 4 is an embodiment of the invention priority decision maker schematic flow sheet;
Fig. 5 judges the structural representation of dress system for embodiment of the invention priority.
Embodiment
With reference to the accompanying drawings the present invention is described more fully, exemplary embodiment of the present invention wherein is described.
Priority decision method of the present invention and device have carried out more comprehensively considering to period of reservation of number, have not only considered traffic performance but also consider the operation demand.The technical scheme that is provided mainly comprises: collect each professional QoS of survice parameter and this professional user call type parameter; Calculate each professional static comprehensive priority according to service attribute and user call type; Calculate each professional dynamic comprehensive priority value in conjunction with network, business, the many-sided factor of user.Suitably choose static state or dynamic comprehensive priority algorithm according to network condition, as the basis for estimation of this service access service.
At first, the stand-by period is to the influence of priority.For guaranteeing the service quality of non-real-time service in the practical communication process, the increase of stand-by period can significantly improve the priority of non-real-time service, makes it surpass the priority of real time business.But the effect of stand-by period not only is limited in the similar non-real-time service, and the priority of real time business also can increase along with the increase of stand-by period.Priority is dynamic change between real-time and non-real-time service.
Secondly, user's Move Mode is to the influence of service priority.In the data transmission procedure, the user may become high-speed moving state from inactive state, and under different mobile statuss, user's transmission characteristic influences difference to what Serving cell caused.With respect to low speed or stationary user, for keeping identical FER, high speed mobile terminal must improve transmit signal power, can cause its interference to other users of sub-district to strengthen like this.In addition, user's Move Mode also can produce certain influence to switching.At the cell boarder place, the user Yi Fasheng " ping-pong " of high-speed mobile causes switching times to increase.In sum, for the performance of elevator system should be considered the influence of user's mobile status to service priority, reduce high-speed mobile user's priority.
Once more, cell load is to the influence of service priority.The required bandwidth difference of different business.High high speed business inserts if the available resource in sub-district can not satisfy priority, and high speed business can be in wait state always so, discharges up to more resources, causes all business of formation back all can't insert.From the entire system interests, the lower low speed business of bandwidth requirement that comes the formation back under the equal conditions should preferentially insert.That is to say that customer service priority should be given different value according to the stock number of sub-district.
Because above-mentioned all kinds of factors are different to the influence degree of user's comprehensive priority.These two kinds of factors of service attribute and user's type of call, it is professional in case selected its exists in the process in business can not change, be absolute factor, the priority definition that the present invention determines absolute factor is static comprehensive priority (Static integrated priority); What factors such as stand-by period, user's Move Mode and cell load reflected is that subjective demand of user and objective environment change the influence to priority, they change in communication process constantly, it is relative factor, variation along with relative factor, it can produce bigger influence to the final priority of user, comprehensive absolute factor and relative factor are determined the priority that the user is final, i.e. dynamic comprehensive priority (Dynamicintegrated priority).As shown in phantom in Figure 2, business 1 static comprehensive priority of being determined by absolute factor is greater than professional 2, but because the influence of factor relatively, along with the priority of these two business of carrying out of business constantly changes, shown in solid line.And the dynamic comprehensive priority in some moment (zone shown in figure A, B) professional 2 has surpassed professional 1.Professional dynamic comprehensive priority not only by absolute factor restriction, also is subjected to the influence of relative factor, its each side such as network environment, business, user influence that has been concentrated expression.When network condition better, can determine priority relationship between each business by static comprehensive priority, and when network condition is relatively poor, need follow the tracks of judgement to each service priority with dynamic comprehensive priority, in time to reflect the priority demand between each business.The structural system of priority of the present invention as shown in Figure 3.
As shown in Figure 4, this novel judging comprehensive priority procedure is as follows:
Step 401 is collected the parameters that is used for the priority judgement, as: each QoS of survice parameter (comprising peak transfer rate, user's average throughput, the maximum delay that packet loss rate and user can tolerate), user call type parameter etc.
Step 402, computing service attribute priority value and user call type priority value.
By each QoS of survice parameter (peak transfer rate, user's average throughput, the maximum delay that packet loss rate and user can tolerate), obtain each professional service attribute priority value.
Described service attribute priority is obtained by following (1) formula:
Definition P I, scBe the service attribute priority value of user i, then
Wherein, R iBe user's peak transfer rate, λ iBe user's average throughput, α iFor the priority weights of differentiated service, when user i is the real time business user, α iBy packet loss rate δ iThe maximum delay T that can tolerate with user i iDecision.Concerning non-real-time service, it only is required to be it and sets proper weight z time delay and insensitive i, guarantee that it can obtain enough bandwidth.
Simultaneously, according to its type of call of user call type parameter decision (newly exhale, switch, connect), and calculate each professional user call type priority value.
Described user call type priority value is obtained by following (2) formula:
Definition P I, sctBe user i type of call priority value
Figure G2009100768636D00132
β 1, β 2, β 3Be to have met the user, switch the user and newly exhaled the user for distinguishing, and the different weights of giving.
Step 403 is judged network condition, if network condition is good, enters step 404, otherwise, skip to step 405.
Described network condition refers to that well network condition is relatively stable, and information change such as offered load, system response time, reliability are little.Under the good situation of network condition, can judge user's comprehensive priority, not need too much to consider the influence of other factors by user call type and service attribute.In concrete the enforcement, the setup parameter threshold value is judged the network condition quality according to actual needs, and for example, when offered load was 80%, in the certain hour scope, its domain of walker only was 5%, promptly 75%~85%, and it is good then to be judged as network condition.
Step 404 by service attribute priority value and the user call type priority value that step 402 obtains, calculates each professional static comprehensive priority value, directly enters step 408.
Described static comprehensive priority, the initial priority standard as the user exists in the process in business, and User Priority is by absolute factor (service attribute and user call type) decision, and is constant relatively.Being different from traditional static priority, is the result by service attribute and user's type of call acting in conjunction decision.
Definition P iBe the static comprehensive priority value of user i, the described static comprehensive priority value of customer service determines jointly by service attribute priority value and user call type priority value, as shown in Equation (3).
P i=w 1*P i,sc+w 2*P i,sct (3)
W wherein 1And w 2Be the weighted value of service attribute priority and user call type priority, they can be constants, also can be that variable (is w 1, w 2Respectively by independent variable P I, scAnd P I, sctDetermine).
Work as w 1And w 2During for constant, this situation can occur: the priority to the very high new user of delay requirement is lower than the switching User Priority not too high to delay requirement, and this is irrational in actual applications.For this reason, according to the fuzzy control principle of prior art and BP neural net method, utilize neuronal activation function s ((4) as the following formula), in the algorithm that the embodiment of the invention adopts with w with feedback characteristics 1, w 2Be configured to P I, scAnd P I, sctNonlinear function, make weight factor w 1, w 2For about P I, scAnd P I, sctBinary function, specifically see following formula (5).
s ( x ) = 1 1 + e - x , x &Element; ( - &infin; , + &infin; ) - - - ( 4 )
w 1 = 1 1 + e - ( P i , sc - P i , sct )
w 2 = 1 - 1 1 + e - ( P i , sc - P i , sct ) - - - ( 5 )
Like this, because neural net has the characteristic of function approximation, by the neural network learning process, reduce the difference of actual output and theoretical output, error is reduced in the tolerable scope, so the selected weight of present embodiment can more rational definite static comprehensive priority, embodied the characteristic that weight factor can be adjusted voluntarily according to the size of service attribute priority and user call type priority.
Static comprehensive priority is applicable to the unconspicuous situation of relative factors vary under the different business, it has reflected user's comprehensive priority, avoided the situation of appearance " carrying out data file download the switching user's of (to time delay less demanding) priority be higher than the voice user who newly exhales all the time ", take all factors into consideration service attribute and user call type, help to reduce new user's blocking rate, reduce those under certain conditions the not high non real-time of delay requirement is switched user's priority, preferentially satisfy real time new and exhale professional demand.
But when changing, static comprehensive priority just can not correctly reflect user's comprehensive priority when relative factor (stand-by period, user's Move Mode and cell load etc.), and promptly there is limitation in the tolerance index of static priority.Along with the growth of stand-by period, the priority of non-real-time service needs to improve relatively; Along with the variation of user moving speed, user's priority also is to need to change; When the heavy duty of sub-district, the service priority of this sub-district to be accessed needs to reduce relatively, alleviates the situation of heavy duty, improves the service priority of underloading to be accessed sub-district, thereby reaches balanced loaded effect.The present invention takes all factors into consideration the influence of absolute factor and relative factor on the basis of static comprehensive priority algorithm, proposed dynamic priority algorithm (DIPC, Dynamic IPC), and it can more fully estimate User Priority.
Dynamic comprehensive priority: the priority criteria of using during the final access service of user, by absolute factor (service attribute and user call type) and relative factor (stand-by period, user's Move Mode and cell load etc.) acting in conjunction decision, constantly change.
DIPC algorithm basic thought: consider in the embodiment of the invention that professional priority in formation is dynamically to adjust, realize the priority adjustment of formation at base station receiving terminal place, and professional unified scheduling.Base station place not only considers the influence to priority of service attribute, two kinds of factors of user call type, also take all factors into consideration the influence of period of reservation of number, user's Move Mode and cell load state three aspect factor, periodically service priority is calculated, determine under the varying environment with this, under the different business condition, the relation of different type of call users' service priority.The calculating of service attribute and user call type priority can be prolonged with the formula in the above-mentioned simple algorithm.
405-407 specifically introduces below by step, the calculating of dynamic comprehensive priority value.
Step 405 is added up institute's patient maximum latency and actual stand-by period, user's Move Mode, the cell load state of each business.
Step 406 is calculated period of reservation of number priority value, user's Move Mode priority value and user and is intended inserting cell load priority of status value.
Patient maximum latency of institute and actual stand-by period by each business, calculate each professional period of reservation of number priority value.
Described period of reservation of number priority value is obtained by following (6) formula:
Definition P I, queueBe user i stand-by period priority value, then
P i , queue = T wait * K , T wait < T max n [ T wait + ( T wait - T max 2 ) m ] * K , T wait &GreaterEqual; T max n
M wherein, n=1,2........ (6)
Wherein, T WaitBe actual stand-by period of user, T MaxBy the patient maximum latency of user, work as T MaxValue when surpassing a certain thresholding, T is pressed in its influence WaitThe power power increase, wherein the n in the thresholding expression formula is according to circumstances different and get different values.And K is the weighted value of different business, and it is a underrange.Real time business is K 1, and non-real-time service is K 2,, suppose K in order to guarantee the priority of real time business 1>K 2
Simultaneously, judge each professional user's Move Mode (comprising high speed, low speed and inactive state), calculate each professional user's Move Mode priority value.
Described user's Move Mode priority value is obtained by following (7) formula:
Definition P I, speedBe user's Move Mode priority value, then
Figure G2009100768636D00171
Wherein, ε 1, ε 2, ε 3User's priority valve when distinguishing corresponding high speed, low speed and inactive state.
And simultaneously, can provide stock number to calculate each professional user and intend inserting cell load priority of status value by each professional resource requirement amount and access subzone network.Wherein, professional resource requirement amount can be determined traffic performance and then learn professional resource requirement amount according to the type of service in the service parameter (browse as basic speech business, visual telephone, generic web page etc.), described service parameter can be collected when step 401 is collected the parameters that is used for the priority judgement, also can be at other times such as collecting in this step; Described access subzone network can provide stock number to obtain from the target cell base station end by further increasing and the mutual step of target cell base station end.
Described user intends inserting cell load priority of status value and is obtained by following (8) formula:
Definition P I, celloadBe user i corresponding priorities value when the different districts load condition, then
Figure G2009100768636D00172
η wherein 1>η 2
η wherein 1, η 2Give the user different priority for provide resource whether to satisfy professional required bandwidth according to the sub-district.
Step 407, the service attribute priority value P that obtains by step 402 I, scWith user call type priority value P I, sct, and the period of reservation of number priority value P that obtains by step 406 I, Queue, user's Move Mode priority value P I, speedIntend inserting cell load priority of status value P with the user I, celload, obtain each professional dynamic comprehensive priority value.At this moment, with the dynamic comprehensive priority value as each professional comprehensive priority value.
Described each professional dynamic comprehensive priority value is finally determined jointly by above-mentioned five factors, is obtained by following formula (9):
G i=k 1*P i,sc+k 2*P i,sct+k 3*P i,queue+k 4*P i,speed+k 5*P i,celload (i>=0) (9)
Wherein, G iThe dynamic comprehensive priority value of representative of consumer i, k 1, k 2, k 3, k 4And k 5Be respectively the weighted value of service attribute, user call type, period of reservation of number, user's Move Mode and five factors of cell load state, be defined as constant at this, it is definite to adopt Fuzzy AHP to analyze.
Step 408 after static comprehensive priority value or dynamic comprehensive priority value calculations are finished, can select optimal user to insert Target cell according to priority orders.
This step specifically comprises:
(1) the fuzzy Complementary Judgement Matrix of structure
According to above-mentioned service attribute, user call type, period of reservation of number, user's Move Mode and five factor (r of cell load state 1, r 2, r 3, r 4, r 5), make relatively judging in twos between factor.If γ Iji/ γ j, expression γ i, γ jTo the ratio of service selection influence degree, thus the parameter of obtaining (r 1, r 2, r 3, r 4, r 5) fuzzy matrix A.
A = 0.5 0.5 0.6 0.7 0.65 0.5 0.5 0.6 0.7 0.65 0.4 0.4 0.5 0.6 0.55 0.3 0.3 0.4 0.5 0.45 0.35 0.35 0.45 0.55 0.5 - - - ( 10 )
Element satisfies among the A: γ Ij>0, γ Ij+ γ Ji=1 (i ≠ j), γ Ii=0.5, and i, j=1,2....n.
γ IjObtaining value method ginseng be shown in Table 1:
Influence degree The strongest Very strong By force Stronger Equate
γ ij 0.9 0.8 0.7 0.6 0.5
Table 1
(2) calculate weight
To the fuzzy Complementary Judgement Matrix that obtains, calculate the weight combination V of evaluating.
V i &prime; = &Sigma; j = 1 n r ij + n 2 - 1 n ( n - 1 ) - - - ( 11 )
To V i' make normalized after, promptly get evaluating weight V={0.2225,0.2225,0.1975,0.1725,0.185}.
(3) consistency check
Check the method for its consistency principle by the compatibility of fuzzy judgment matrix.
V=(V 1, V 2, V 3, V 4, V 5) TWeight vectors for fuzzy judgment matrix A makes V Ij=V i/ (V i+ V j), then claim n rank matrix V *=(V Ij) N*nEigenmatrix for A.As fuzzy judgment matrix A and its eigenmatrix V *Complete when compatible, fuzzy judgment matrix A just is a crash consistency.Compatibility index I as shown in Equation (12).
I ( A , V * ) = 1 n 2 &Sigma; i = 1 n &Sigma; j = 1 n | A ij + V ij - 1 | - - - ( 12 )
Can get actual I (A, V by calculating *Therefore)=0.0604<0.1 thinks that judgment matrix is conforming for meeting.
(4) degree of membership is calculated relatively
At first provide each evaluation of indexes value of different schemes, { ω α(β) | α=1,2 ... .., n; β=1,2 ... .m}, wherein β represents the number of different schemes, is set up the fuzzy evaluation matrix of relative degree of membership then by evaluating.For making modeling have versatility, keep the change information of each parameter value simultaneously as far as possible, each parameter value is carried out standardization.
s &alpha; ( &beta; ) = w &alpha; ( &beta; ) w max ( &alpha; ) + w min ( &alpha; ) - - - ( 13 )
In the formula, ω Max(α), ω Min(α) be respectively maximum, the minimum value of α evaluating; ω α(β) be evaluating value before the standardization, s α(β) for the relative degree of membership of standardization post-evaluation parameter.
(5) optimal user is determined
Can calculate total degree of membership by formula (14).
S ( &beta; ) = &Sigma; &alpha; v &alpha; s &alpha; ( &beta; )
α={ r wherein 1, r 2, r 3, r 4, r 5,
β=1,...m (14)
V wherein αThe weight of expression evaluating, s α(β) the relative degree of membership of expression evaluating.
G=S(β) (15)
According to the priority of each customer service request, select the call request of priority maximum to insert, promptly select the service access network of total degree of membership maximum in the formation to get final product, this user is optimal user.
Step 409, treat that the user inserts after, upgrade all parameter sets.After priority is judged T blanking time, proceed the circulation of next round.
In addition, when network condition was good, priority judged that blanking time, T can increase, utilize the static comprehensive priority algorithm to judge User Priority, opposite but when the network condition difference, T just need reduce, and adopt the dynamic comprehensive priority algorithm to judge, improve the priority calculated rate.
Need to prove, because the DIPC algorithm always selects the highest user of priority to insert Target cell, when the target surplus resources can not satisfy the user's request with limit priority, by chance there is not the non real-time user who has inserted again in this sub-district this moment, there is not available resource to discharge, all users' priority will constantly be judged by system so, as long as this user's priority is high always, will cause the user of its back also will to connect to advance this sub-district, thereby cause the raising of cutting off rate or blocking rate.At this, the present invention is " bulkycongestion " problem with this problem definition.Have only priority as other users to be higher than this user's priority, and system's residual resource be can satisfy the user's that priority catches up from behind demand again the time, cutting off rate or blocking rate just can not improve, otherwise top problem will occur.For this reason can be in step 409, after treating that the user inserts, further add following steps: when Target cell can not satisfy priority during high user's demand, the number of times of this User Priority of restriction inquiry, when reaching inquiry number of times threshold value, just no longer this user is done to insert and handle, continue the next user of inquiry.Will reduce cutting off rate or blocking rate for the low slightly user of priority provides more access chance like this, make the combination property optimum of system.
For two said methods example in actual applications, specified below.
Example 1
Suppose to have had in the Zone switched queue queue user's 1 request call real-time voice repeatedly newly exhale wait to be accessed, the non real-time that is carrying out simultaneously the file download switches user 2 and moves to Zone switched, also enter in the switching and queuing formation Target cell to be accessed such as the two is common.And there is in the sub-district this moment resource to discharge but is not enough to provide both to serve, need carry out priority and calculate, judge actually which user to provide service for earlier according to the various parameters of two kinds of business.Priority was calculated once for phase weekly, had maximum person and was optimal user.
As shown in Figure 4, user 1, user 2 are carried out input set and collect step.The input parameter collection comprises average throughput, peak transfer rate, time delay, packet loss rate and user call type.Because the traffic performance parameter is not change in the process that business exists, so this Pi, the sc value can draw when business produces.Carry out judging comprehensive priority process 1, draw each user's service attribute priority, wherein P 1, sc=8, P 2, sc=10.ω then Max(sc)=10, ω Min(sc)=8, degree of membership is respectively relatively: S Sc(1)=8/ (8+10)=0.444, S Sc(2)=10/ (8+10)=0.556.
s &alpha; ( &beta; ) = w &alpha; ( &beta; ) w max ( &alpha; ) + w min ( &alpha; ) - - - ( 13 )
From user perspective, user call type can change, but user call type also is constant always before service is not processed.For example, type of call is always for newly exhaling the user before access service not, insert the back for just for meeting the user, do not switch to other sub-districts and become the switching user but need only this user, its type of call also is constant, and user call type only just changes after service processing.For this example, to cross in the stroke in the wait access service, user call type priority is changeless, according to the priority decision flowchart, carries out judging comprehensive priority process 2, determines each user's P 1, sct, P wherein 1, sct=1, P 2, sct=2.According to formula (13), draw relative degree of membership: S respectively equally Sct(1)=1/ (1+2)=0.333, S Sct(2)=2/ (1+2)=0.667.
Standby service attribute priority and user call type priority need to judge whether network state is good after calculating and finishing, and select quiet/dynamic comprehensive priority computational methods.The hypothetical network situation is bad herein, promptly need carry out dynamic comprehensive priority decision mechanism.According to DIPC, stand-by period, user's Move Mode and cell load state are change at random, need to continue 3 parameters of statistics, carry out judging comprehensive priority process 5,6 and 7, draw P I, queue, P I, speed, P I, celloadSo far each user has one group of 5 different classes of priority { service attribute priority, type of call priority, stand-by period priority, Move Mode priority, cell load priority of status }, carry out judging comprehensive priority process 8, draw the dynamic comprehensive priority of user 1 and 2 according to above 2 groups of five different priority.Wherein, P 1, queue=5, P 2, queue=3, relative degree of membership S Queue(1)=5/ (5+3)=0.625, S Queue(2)=3/ (5+3)=0.375; P 1, speed=1, P 2, speed=0.5, relative degree of membership S Speed(1)=1/ (1+0.5)=0.667, S Speed(2)=0.5/ (1+0.5)=0.333; P 1, celload=0.6, P 2, celload=0.4, S Celload(1)=0.6, S Celload(2)=0.4.
Determine formula (14) according to optimal user, the evaluating weight V={0.2225 that draws by Fuzzy AHP again, 0.2225,0.1975,0.1725,0.185} finally determines each user's comprehensive priority: S (1)=0.522, S (2)=0.478.Specifically see Table customer parameter table and the degree of membership shown in 2.User 1 priority preferentially selects request call real-time voice repeatedly newly to exhale user access network greater than user 2 priority for this reason.
S ( &beta; ) = &Sigma; &alpha; v &alpha; s &alpha; ( &beta; )
α={ r wherein 1, r 2, r 3, r 4, r 5,
β=1,...m (14)
V wherein αThe weight of expression evaluating, s α(β) the relative degree of membership of expression evaluating.
Figure G2009100768636D00222
Table 2
In table 2, for switching the user, no matter whether it enters queuing, for its at the initial value 3s of the accessed user of queuing time place.
After treating newly to exhale user 1 to insert Target cell, its user call type is updated at once and has met the user, and the traffic performance parameter is still constant, other parameters can specifically be upgraded according to actual conditions, promptly carry out the final step of flow chart---upgrade all parameter sets, after judging interval T, carry out the priority of new lower whorl and judge.
Example 2
To be switched non real-time file data downloaded user 1 such as suppose to have only in the Zone switched queue queue in the formation, in the wait process in the sub-district always No Assets discharge.Network state is still bad, according to DIPC, relatively this user with inserted carry out the priority that file is downloaded the user 2 of transport service.
Figure G2009100768636D00231
Figure G2009100768636D00241
Table 3
In table 3, the accessed user does not have the stand-by period, for assurance accessed user's service quality so at the initial value 5s of the accessed user of queuing time place.
Computational methods are the same, can be got by the result of customer parameter table and degree of membership in the table 3, and user 1 numerical priority value is 0.55, user 2 numerical priority value is 0.45, user's 1 access network, and user 2 is forced to stop transfer of data, waits for the comparison of priority of following one-period.And user 1 type of call becomes and meets the user this moment, and user 2 is forced into buffering area and waits for, same traffic performance parameter does not all change, and other parameters can be upgraded according to other concrete conditions.
When the priority of switching user or new user is higher than when the not high non real-time of delay requirement connect user's priority, the former can temporarily seize the resource that non real-time has met the user, access service, meet the non real-time user and then temporarily withdraw from service, enter buffering area, this kind phenomenon is called " the non real-time user dodges phenomenon ".The ratio that the non real-time that has inserted is dodged number of users and all accessed user's numbers is the non real-time user and dodges rate.Thus, will produce the non real-time user and dodge, reduce connecing the integrality that the user accepts to serve.But it is not completely without benefit that the non real-time user dodges the generation of rate, and it has reduced new user's blocking rate to a certain extent and has switched the conversation loss rate, has improved the performance of system's flexible configuration; But too much non real-time user dodges and can cause the service quality of non-real-time service to descend.For this reason, in the middle of real system, should reduce the non real-time user as far as possible and dodge rate preferentially satisfying new user's blocking rate and switching under the prerequisite of conversation loss rate.According to the present invention, when being forced in buffering area etc., the non real-time user who has inserted bides one's time, can increase the non real-time User Priority along with the increase of waiting for time delay improves, avoid occurring too much non real-time user and dodge phenomenon.
By above interpretation of result as can be known, the priority of customer service is with stand-by period, User Status, the dynamic change of cell load state.The size of priority is traffic performance, user call type, stand-by period, user's Move Mode and five coefficient results of factor of cell load state.Broken and connect user priority, switching user priority and the preferential thought of real time business in traditional priority, under the different business condition, the priority of switching the user is not necessarily greater than new calling party's priority, and accessed user's priority is not necessarily greater than switching, new calling party.Ensure that with this miscellaneous service, various user can both obtain satisfied service, the overall performance of elevator system.
Based on the inventive method, the present invention also provides a kind of judging comprehensive priority device of mobile communication system.Referring to shown in Figure 5, Fig. 5 is apparatus of the present invention modular structure figure.The functional description of each module is as follows:
The priority critical parameter is collected module 501, collection is used for the parameters that priority is judged, as each QoS of survice parameter (peak transfer rate, user's average throughput, the maximum delay that packet loss rate and user can tolerate), user call type etc. are for service attribute priority computing module 502, user call type priority computing module 503, period of reservation of number priority computing module 504, user's Move Mode priority computing module 505, cell load priority of status computing module 506 provide corresponding calculating parameter.
Service attribute priority computing module 502, computing service attribute priority value.
User call type priority computing module 503 calculates the user call type priority value.
Period of reservation of number priority computing module 504 calculates the period of reservation of number priority value.
User's Move Mode priority computing module 505 calculates user's Move Mode priority value.
Cell load priority of status computing module 506, calculation plot load condition priority value.
Comprehensive priority computing module 507 comprises being used to the dynamic comprehensive priority calculating sub module 5072 calculating the static comprehensive priority calculating sub module 5071 of static comprehensive priority value and be used to calculate the dynamic comprehensive priority value; The subzone network state information that this comprehensive priority computing module 507 provides according to subzone network condition judgement module 508, choose the next result of calculation of corresponding submodule, calculate each service integration priority according to service attribute priority computing module 502, user call type priority computing module 503, period of reservation of number priority computing module 504, user's Move Mode priority computing module 505,506 outputs of cell load priority of status computing module.
Subzone network condition judgement module 508 is judged the subzone network state, for comprehensive priority computing module 507, priority judge that time interval computing module 509 provides the subzone network state information.
Priority is judged time interval computing module 509, by the subzone network state information that subzone network condition judgement module 507 provides, sets priority and judges the time interval.
Description of the invention is in order to provide for the purpose of example and the explanation, and is not exhaustively or limit the invention to disclosed form.Many modifications and variations are obvious for the ordinary skill in the art.Selecting and describing embodiment is for better explanation principle of the present invention and practical application, thereby and makes those of ordinary skill in the art can understand the various embodiment that have various modifications that the present invention's design is suitable for special-purpose.

Claims (14)

1. the judging comprehensive priority method of a mobile communication system is characterized in that, comprising:
A. collect each professional QoS of survice parameter and user call type parameter;
B. according to each professional service attribute priority value of each QoS of survice calculation of parameter, obtain each professional user call type priority value according to the user call type calculation of parameter;
C. the result according to step B calculates each professional comprehensive priority value, wherein
By formula P i=w 1* P I, sc+ w 2* P I, sctCalculate the static comprehensive priority value of user i, wherein, P iBe the static comprehensive priority value of user i, P I, scBe the service attribute priority value of user i, P I, sctBe user i type of call priority value, w 1And w 2Weighted value for service attribute priority and user call type priority;
By formula G i=k 1* P I, sc+ k 2* P I, sct+ k 3* P I, queue+ k 4* P I, speed+ k 5* P I, celload(i>=0) calculates the dynamic comprehensive priority value of user i, wherein P I, queueBe period of reservation of number priority value, P I, speedBe user's Move Mode priority value, P I, celloadFor the user intends inserting cell load priority of status value, obtain each professional dynamic comprehensive priority value; k 1, k 2, k 3, k 4And k 5Be respectively the weighted value of service attribute, user call type, period of reservation of number, user's Move Mode and five factors of cell load state.
2. method according to claim 1 is characterized in that, described QoS of survice parameter comprises: peak transfer rate, user's average throughput;
Step B is described further to be comprised according to each professional service attribute priority value of each QoS of survice calculation of parameter:
The service attribute priority P of user i I, scCalculate by following formula
P i , sc = &alpha; i R i &lambda; i ,
Figure FSB00000409653600012
Wherein, R iBe user's peak transfer rate, λ iBe user's average throughput, α iFor the priority weights of differentiated service, when user i is the real time business user, α iBy packet loss rate δ iThe maximum delay T that can tolerate with user i iDecision; z iFor preestablishing;
Step B is described to be obtained each professional user call type priority value according to the user call type calculation of parameter and further comprises:
User i type of call priority value P I, sctCalculate by following formula
β wherein 1, β 2, β 3Be to have met the user, switch the user and newly exhaled the user for distinguishing, and the different weights of giving.
3. method according to claim 1 is characterized in that, described w 1And w 2Calculate by following formula:
w 1 = 1 1 + e - ( P i , sc - P i , sct )
w 2 = 1 - 1 1 + e - ( P i , sc - P i , sct ) ,
Wherein, P I, scBe the service attribute priority value of user i, P I, sctBe user i type of call priority value.
4. method according to claim 1 and 2 is characterized in that, if current network conditions is bad, then step B further comprises:
Add up institute's patient maximum latency and actual stand-by period, user's Move Mode, the cell load state of each business;
Patient maximum latency of institute and actual stand-by period by each business, calculate each professional period of reservation of number priority value;
Judge user's Move Mode that each is professional, calculate each professional user's Move Mode priority value;
Can provide stock number by each professional resource requirement amount and access subzone network, calculate each professional user and intend inserting cell load priority of status value;
The service attribute priority value and the user call type priority value that further obtain among the step C by step B, and period of reservation of number priority value, user's Move Mode priority value and user intend inserting cell load priority of status value, calculates each professional comprehensive priority value.
5. method according to claim 4 is characterized in that, described QoS of survice parameter comprises: peak transfer rate, user's average throughput, the maximum delay that packet loss rate and user can tolerate;
Described period of reservation of number priority value is calculated by following formula:
P i , queue = T wait * K , T wait < T max n [ T wait + ( T wait - T max 2 ) m ] * K , T wait &GreaterEqual; T max n
1m wherein, n=1,2........
P I, queueBe user i stand-by period priority value, T WaitBe actual stand-by period of user, T MaxBe the patient maximum delay of user, K is the weighted value of different business;
Described each professional user's Move Mode priority value is calculated by following formula:
Figure FSB00000409653600032
Wherein, P I, speedBe user's Move Mode priority value, ε 1, ε 2, ε 3User's priority valve when distinguishing corresponding high speed, low speed and inactive state;
Described user intends inserting cell load priority of status value and is calculated by following formula:
η wherein 1>η 2
P I, celloadBe user i corresponding priorities value when the different districts load condition, η 1, η 2Give the user different priority for provide resource whether to satisfy professional required bandwidth according to the sub-district.
6. method according to claim 1 is characterized in that, the weighted value of described service attribute, user call type, period of reservation of number, user's Move Mode and five factors of cell load state adopts Fuzzy AHP to analyze definite.
7. method according to claim 1 is characterized in that, further comprises behind the described step C:
D. insert Target cell according to the comprehensive priority value selective sequential optimal user that obtains.
8. method according to claim 7 is characterized in that, described step D further comprises:
According to described service attribute, user call type, period of reservation of number, user's Move Mode and five factor r of cell load state 1, r 2, r 3, r 4, r 5, make relatively judging in twos between factor, establish r Ij=r i/ r j, expression r i, r jTo the ratio of service selection influence degree, obtain parameter r 1, r 2, r 3, r 4, r 5Fuzzy complementary judgement matrix A
A = 0.5 0.5 0.6 0.7 0.65 0.5 0.5 0.6 0.7 0.65 0.4 0.4 0.5 0.6 0.55 0.3 0.3 0.4 0.5 0.45 0.35 0.35 0.45 0.55 0.5
Wherein, element satisfies among the A: r Ij>0, r Ij+ r Ji=1 (i ≠ j), r Ii=0.5, and i, j=1,2....n;
To the fuzzy Complementary Judgement Matrix that obtains, calculate
V i &prime; = &Sigma; j = 1 n r ij + n 2 - 1 n ( n - 1 )
To V i' make normalized after, promptly get evaluating weight V={0.2225,0.2225,0.1975,0.1725,0.185};
Check the method for its consistency principle by the compatibility of fuzzy Complementary Judgement Matrix;
With V=(V 1, V 2, V 3, V 4, V 5) TWeight vectors for fuzzy Complementary Judgement Matrix A makes V Ij=V i/ (V i+ V j), then claim n rank matrix V *=(V Ij) N*nBe the eigenmatrix of A, calculate compatibility index I
I ( A , V * ) = 1 n 2 &Sigma; i = 1 n &Sigma; j = 1 n | A ij + V ij - 1 |
Provide each evaluation of indexes value of different schemes, { ω α(β) | α=1,2 ... .., n; β=1,2 ... .m}, wherein β represents the number of different schemes, sets up the fuzzy evaluation matrix of relative degree of membership by evaluating, and each parameter value is carried out standardization
S &alpha; ( &beta; ) = &omega; &alpha; ( &beta; ) &omega; max ( &alpha; ) + &omega; min ( &alpha; )
Wherein, ω Max(α), ω Min(α) be respectively maximum, the minimum value of α evaluating; ω α(β) be evaluating value before the standardization, s α(β) for the relative degree of membership of standardization post-evaluation parameter;
Calculate total degree of membership
S ( &beta; ) = &Sigma; &alpha; v &alpha; s &alpha; ( &beta; )
α={ r wherein 1, r 2, r 3, r 4, r 5,
β=1,...m
V wherein αThe weight of expression evaluating, s α(β) the relative degree of membership of expression evaluating;
According to the priority of each customer service request, selecting the user of the professional correspondence of total degree of membership maximum in the formation is optimal user.
9. method according to claim 7 is characterized in that, further comprises behind the described step D: upgrade all parameter sets, after priority is judged blanking time, proceed the circulation of next round.
10. the judging comprehensive priority device of a mobile communication system is characterized in that, comprising:
The priority critical parameter is collected module, collects each QoS of survice parameter and user call type parameter;
Service attribute priority computing module is used for according to each QoS of survice calculation of parameter service attribute priority value;
User call type priority computing module is used for according to user call type calculation of parameter user call type priority value;
The comprehensive priority computing module according to the result of calculation of service attribute priority computing module, user call type priority computing module, output, calculates each service integration priority value, wherein
By formula P i=w 1* P I, sc+ w 2* P I, sctCalculate the static comprehensive priority value of user i, wherein, P iBe the static comprehensive priority value of user i, P I, scBe the service attribute priority value of user i, P I, sctBe user i type of call priority value, w 1And w 2Weighted value for service attribute priority and user call type priority;
By formula G i=k 1* P I, sc+ k 2* P I, sct+ k 3* P I, queue+ k 4* P I, speed+ k 3* P I, celload(i>=0) calculates the dynamic comprehensive priority value of user i, wherein P I, queueBe period of reservation of number priority value, P I, speedBe user's Move Mode priority value, P I, celloadFor the user intends inserting cell load priority of status value, obtain each professional dynamic comprehensive priority value; k 1, k 2, k 3, k 4And k 5Be respectively the weighted value of service attribute, user call type, period of reservation of number, user's Move Mode and five factors of cell load state.
11. device according to claim 10 is characterized in that, described QoS of survice parameter comprises: peak transfer rate, user's average throughput;
Described customer service attribute priority value calculates by following formula
P i , sc = &alpha; i R i &lambda; i ,
Figure FSB00000409653600062
Wherein, P I, scBe the service attribute priority of user i, R iBe user's peak transfer rate, λ iBe user's average throughput, α iFor the priority weights of differentiated service, when user i is the real time business user, α iBy packet loss rate δ iThe maximum delay T that can tolerate with user i iDecision; z iFor preestablishing;
Described user call type priority value calculates by following formula
Figure FSB00000409653600063
Wherein, P I, sctBe user i type of call priority value, β 1, β 2, β 3Be to have met the user, switch the user and newly exhaled the user for distinguishing, and the different weights of giving.
12. device according to claim 10 is characterized in that, this device also comprises:
Priority is judged time interval computing module, by the subzone network state information that subzone network condition judgement module provides, sets priority and judges the time interval;
Subzone network condition judgement module is judged the subzone network state, for comprehensive priority computing module, priority judge that time interval computing module provides the subzone network state information;
Period of reservation of number priority computing module by the patient maximum latency of institute and the actual stand-by period of each business, calculates period of reservation of number priority;
User's Move Mode priority computing module is judged user's Move Mode that each is professional, calculates user's Move Mode priority;
Cell load priority of status computing module can provide stock number, calculation plot load condition priority by each professional resource requirement amount and access subzone network;
Comprise the static comprehensive priority calculating sub module in the described comprehensive priority computing module, by service attribute priority value that obtains and user call type priority value calculations static comprehensive priority value; Dynamic comprehensive priority calculating sub module, by the service attribute priority value that obtains, user call type priority value, period of reservation of number priority value, user's Move Mode priority value and user intend inserting cell load priority of status value, calculate each professional dynamic comprehensive priority value; And this comprehensive priority computing module is chosen corresponding submodule according to the subzone network state information that subzone network condition judgement module provides.
13. device according to claim 12 is characterized in that, described QoS of survice parameter comprises: peak transfer rate, user's average throughput, the maximum delay that packet loss rate and user can tolerate;
Described period of reservation of number priority value is calculated by following formula:
P i , queue = T wait * K , T wait < T max n [ T wait + ( T wait - T max 2 ) m ] * K , T wait &GreaterEqual; T max n
M wherein, n=1,2........
P I, queueBe user i stand-by period priority value, T WaitBe actual stand-by period of user, T MaxBe the patient maximum delay of user, K is the weighted value of different business;
Described each professional user's Move Mode priority value is calculated by following formula:
Figure FSB00000409653600072
Wherein, P I, speedBe user's Move Mode priority value, ε 1, ε 2, ε 3User's priority valve when distinguishing corresponding high speed, low speed and inactive state;
Described user intends inserting cell load priority of status value and is calculated by following formula:
Figure FSB00000409653600073
η wherein 1>η 2
P I, celloadBe user i corresponding priorities value when the different districts load condition, η 1, η 2Give the user different priority for provide resource whether to satisfy professional required bandwidth according to the sub-district.
14. device according to claim 10 is characterized in that, the weighted value of described service attribute, user call type, period of reservation of number, user's Move Mode and five factors of cell load state adopts Fuzzy AHP to analyze definite.
CN2009100768636A 2009-01-23 2009-01-23 Method and device for judging comprehensive priority of mobile communication system Expired - Fee Related CN101516093B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN2009100768636A CN101516093B (en) 2009-01-23 2009-01-23 Method and device for judging comprehensive priority of mobile communication system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN2009100768636A CN101516093B (en) 2009-01-23 2009-01-23 Method and device for judging comprehensive priority of mobile communication system

Publications (2)

Publication Number Publication Date
CN101516093A CN101516093A (en) 2009-08-26
CN101516093B true CN101516093B (en) 2011-11-16

Family

ID=41040331

Family Applications (1)

Application Number Title Priority Date Filing Date
CN2009100768636A Expired - Fee Related CN101516093B (en) 2009-01-23 2009-01-23 Method and device for judging comprehensive priority of mobile communication system

Country Status (1)

Country Link
CN (1) CN101516093B (en)

Families Citing this family (20)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101707788B (en) * 2009-10-27 2014-04-02 北京邮电大学 Differential pricing strategy based dynamic programming method of multilayer network services
CN102395208B (en) * 2011-06-29 2019-02-15 中兴通讯股份有限公司 A kind of adjust automatically voice sends the methods, devices and systems of parameter
US8909836B2 (en) * 2012-10-08 2014-12-09 Andes Technology Corporation Interrupt controller, apparatus including interrupt controller, and corresponding methods for processing interrupt request event(s) in system including processor(s)
CN103152795B (en) * 2013-03-26 2015-04-29 哈尔滨工业大学 Heterogeneous network access control method based on multi-objective optimization
CN104185236A (en) * 2013-05-23 2014-12-03 普天信息技术研究院有限公司 Radio communication method and system based on multi-type broadband mobile communication technology
CN104378773A (en) * 2013-08-12 2015-02-25 中兴通讯股份有限公司 Working mode determining method and device, network element, EMS and NMS
CN104427623B (en) * 2013-08-30 2018-03-23 中国移动通信集团设计院有限公司 A kind of business scheduling method and device
CN105306366B (en) * 2014-05-29 2019-07-05 中国移动通信集团公司 Data routing method and device
CN104168660B (en) * 2014-08-27 2017-09-29 中国联合网络通信集团有限公司 Determine the method and device of priority
CN104598739B (en) * 2015-01-26 2018-05-01 西安电子科技大学 A kind of index system construction method of network-oriented overall efficiency
WO2016161582A1 (en) * 2015-04-08 2016-10-13 华为技术有限公司 Quality of service measurement method and apparatus, and base station
CN105306826A (en) * 2015-11-20 2016-02-03 小米科技有限责任公司 Camera function setting method and device
CN105578503B (en) * 2015-12-10 2019-03-01 中国联合网络通信集团有限公司 A kind of method and device of service access
CN105681717B (en) * 2016-03-04 2019-09-17 广州星唯信息科技有限公司 The distributed storage method and system of public transport vehicle-mounted monitoring video
CN105681219A (en) * 2016-03-09 2016-06-15 上海斐讯数据通信技术有限公司 WIFI network message processing method, system and WIFI network system
CN107786956B (en) * 2016-08-29 2021-06-08 中国移动通信集团安徽有限公司 Method and device for determining real-time priority of package
CN108668296B (en) * 2017-03-29 2021-07-27 中国移动通信集团湖北有限公司 Method, device and equipment for determining circuit switched fallback perception difference cell
CN111479176B (en) * 2019-11-27 2022-05-27 重庆前卫表业有限公司 Self-adaptive fuzzy data uploading method based on NB-IoT
CN112055382B (en) * 2020-07-28 2024-04-12 西安空间无线电技术研究所 Service access method based on refined distinction
CN112867166B (en) * 2021-01-18 2022-05-20 重庆邮电大学 Service data scheduling method for micro-power wireless communication network in smart power grid

Also Published As

Publication number Publication date
CN101516093A (en) 2009-08-26

Similar Documents

Publication Publication Date Title
CN101516093B (en) Method and device for judging comprehensive priority of mobile communication system
CN1327668C (en) System and method for user scheduling in a communication network
CN109600262A (en) Resource self-configuring and self-organization method and device in URLLC transmission network slice
CN100583901C (en) Service quality guarantee method for broadband radio access network media accessing control layer
CN101600231A (en) In packet scheduling, determine the method and system of the priority of subscriber equipment
CN106973413A (en) The Adaptive QoS control method of Wireless Sensor Networks
CN106100907A (en) A kind of MEC server selection algorithm based on fairness
CN104219167A (en) Network resource scheduling method and server
CN105873112A (en) Multi-mode terminal vertical switching method in heterogeneous network
CN112203345B (en) D2D communication energy efficiency maximization power distribution method based on deep neural network
CN103687025B (en) Resource allocation method used for LTE system
CN101662810A (en) Management method and management system for multiple services and multiple connections of mobile communication
CN102724763A (en) Time domain group scheduling method on basis of two-dimensional priority level
CN106095529A (en) A kind of carrier wave emigration method under C RAN framework
CN106413015A (en) Network access switching method and device
CN100440871C (en) Packet scheduling method and device for wireless communication system
CN107197530A (en) The downlink resource allocation algorithm that the fluctuation of LTE system lower channel is perceived
CN106878967A (en) Note transmission method and device
CN101841915A (en) Method for scheduling and distributing resources in communication system and scheduler
CN101137163A (en) Data service capacity planning method of honeycomb wireless communication network
CN1347253A (en) Algorithm for scheduling and distributing packet data service resources in mobile environment
CN101568154B (en) Method and device for determining scheduling priority
CN101902818A (en) Wireless resource scheduling method and system
CN102025539A (en) Agent-based available resource assessment method for cognitive network
CN105338575B (en) A kind of control method, device and server

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
C14 Grant of patent or utility model
GR01 Patent grant
CF01 Termination of patent right due to non-payment of annual fee

Granted publication date: 20111116

Termination date: 20160123

EXPY Termination of patent right or utility model