CN103607737A - Heterogeneous network service flow-dividing method and system - Google Patents

Abstract

The invention provides a heterogeneous network service flow-dividing method and system. The method comprises: receiving service request information of a multi-mode terminal; performing information detection on each link in a heterogeneous network, and periodically obtaining the queuing length information and data packet quantity of the cache region of each link; according to the periodically obtained queuing length information and data packet quantity of the cache region of each link, determining the service flow-dividing portion of each link; and according to the service flow-dividing portion of each link, periodically performing service flow dividing. By using the method and system provided by the invention, the service feedback of the cache region of each link can be periodically obtained and the detected information can be integrated into a theoretical model so that the dynamic combination of a flow-dividing decision-making method of theoretical and engineering practice can be realized, and the accuracy and initiative of a flow-dividing method can be improved.

Description

A kind of heterogeneous-network service shunt method and system

Technical field

The present invention relates to wireless technical field, relate in particular to a kind of heterogeneous-network service shunt method and system.

Background technology

Along with the blowout formula growth of multimedia service and the diversity of user's request class of business change, rely on the network of single standard to provide data, services can not guarantee that user is in all demands of the aspects such as throughput, time delay to user.Simultaneously due to the development of radio network technique, no matter the fusion between heterogeneous network and cooperative transmission still in engineering practice, have all obtained some achievements in theoretical research.Isomery merges and will become the main flow trend of next generation wireless communication network, and 2G, 3G and 4G network the situation of depositing begin to take shape, and this will cause same geographic area to be covered by multiple isomery subnet.

Along with the evolution development of terminal technology, the multimode terminal that can simultaneously access a plurality of heterogeneous networks reception and transmission data is progressively commercial.Therefore in future soon, in order to promote network efficiency and to guarantee user's business experience, Virtual network operator, when providing business service for user, may transmit data from a plurality of heterogeneous networks simultaneously.In this course, how to distribute the service distributing ratio between each heterogeneous network that the Diffluence Algorithm in minute streaming server is determined, this is the core of whole separate system, directly affects the efficiency of business transmission and user's service quality (QoS) experience.

In the heterogeneous network separate system having existed at present, there are some achievements in theory with in engineering.In theoretical research, the transmission of Business Stream in network is modeled as general queueing theory, and the network delay of usining minimizes as optimization aim sets up Optimized model, solves the optimum shunt ratio that draws heterogeneous networks.The method has been determined the shunt ratio in whole business transfer process from the beginning, by the abstract approximate modeling of wireless channel being tried to achieve to the optimal transmission solution of whole transmitting procedure; And in engineering, adopt configuration information collection and feedback node in network, situation about periodically transmitting to shunting server feedback business, comprise time delay, packet loss etc., for a minute state for current each heterogeneous network of streaming server judgement, and then formulate rational service distributing decision-making.The method does not have strict theoretical foundation, relies on the real-time network information of obtaining to analyze and adjudicate.

As can be seen here, heterogeneous network separate system of the prior art, be no matter based on theory or based on engineering practice, all come with some shortcomings: the separate system based on theoretical research lacks the dynamically adapting to real system, separate system based on engineering practice lacks certain theoretical foundation, prior art cannot be accomplished the effective combination of the two, can not meet the shunting accuracy of decision-making and the requirement of initiative simultaneously.

Summary of the invention

(1) technical problem that will solve

The invention provides a kind of heterogeneous-network service shunt method and system, to solve in prior art, cannot be by the technical problem based on accuracy and the effective combination of initiative in theory and the shunting decision-making technique based on engineering practice.

(2) technical scheme

For solving the problems of the technologies described above, the invention provides a kind of heterogeneous-network service shunt method, comprising:

Receive the service requesting information of multimode terminal;

Every link in heterogeneous network is carried out to information detection, periodically obtain every link buffer area queuing long message and data packet number;

According to the every link buffer area queuing long message periodically getting and data packet number, determine the service distributing ratio of every link;

According to the service distributing ratio of described every link, periodically carry out service distributing.

Further, described according to the every link buffer area queuing long message periodically getting and data packet number, determine that the service distributing ratio of every link comprises:

By every link queuing long message l _nwith the long threshold value L of link queuing comparison, determine the service speed μ of every link _n, n is the number of links in heterogeneous network;

According to the data packet number m of every the link periodically getting _n, extract the data packet number m of every link when queuing up long threshold value L _nL;

According to the service speed μ of every link _nwith the data packet number m when queuing up long threshold value L _nL, the packet that obtains every link postpones D _n;

According to the principle of minimization system delay, determine the service distributing ratio of every link.

Further, described by every link queuing long message l _nwith the long threshold value L of link queuing comparison, determine the service speed μ of every link _ncomprise:

Profit is determined the service speed μ of every link with the following method _n:

${\mathrm{\&mu;}}_n=\left\{\begin{array}{cc}{\mathrm{\&mu;}}_{\mathrm{nl}}& {\mathrm{ifl}}_n<L\\ {\mathrm{\&mu;}}_{\mathrm{nm}}& \mathrm{otherwise}\end{array}\right.;$

μ wherein _nlfor the service speed higher limit of this link, μ _nmservice speed lower limit for this link.

Further, described according to the service speed μ of every link _nwith the data packet number m when queuing up long threshold value L _l, the packet that obtains every link postpones D _ncomprise:

When this link is long-term evolving network link,

λ wherein _nfor the arrival rate of this link, this link traffic stream obedience parameter is λ _npoisson flow;

When this link is long-term evolving network during through WLAN conversion link,

$D_n=D_n^{\mathrm{WLAN}}+D_n^{\mathrm{LTE}},$

Wherein

for the packet delay of considering that WLAN disturbs flow to cause,

wherein λ ' is the parameter of WLAN interference flow Poisson flow, μ ' _narrival rate for this link after consideration WLAN interference flow;

the average data packet delay causing for long-term evolving network:

$\begin{array}{c}\begin{array}{c}{D}_n^{\mathrm{LTE}}={(\frac{1-{\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nl}}}\right)}^{m_{\mathrm{nL}}+1}}{1-\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nL}}}\right)}+\frac{{\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nl}}}\right)}^{m_{\mathrm{nL}}}\&CenterDot;\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{um}}}}{1-\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nm}}}\right)})}^{-1}\&CenterDot;\{\frac{\left(\frac{1}{{\mathrm{\&mu;}}_{\mathrm{nl}}}\right)(1+(m_{\mathrm{nL}}-1){\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nl}}}\right)}^{m_{\mathrm{nL}}}-m_{\mathrm{nL}}{\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nl}}}\right)}^{m_{\mathrm{nL}}-1})}{{(1-\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nl}}}\right))}^2}\\ +\frac{{\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nl}}}\right)}^{m_{\mathrm{nL}}-1}(m_{\mathrm{nL}}-(m_{\mathrm{nL}}-1)\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nm}}}\right))}{{(1-\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nm}}}\right))}^2}\}\end{array}\end{array}$

λ wherein _nfor the arrival rate of this link in long-term evolving network.

Further,

The described principle postponing according to minimization system, determine that the service distributing ratio of every link comprises:

The principle postponing according to minimization system, solves following optimization problem:

make

and meet 0≤λ in long-term evolving network link _n≤ μ _n, long-term evolving network is 0≤λ in WLAN conversion link _n<min (μ _n, μ _n')-λ ', wherein λ is the parameter of total service traffics Poisson flow, vector

Described according to the service distributing ratio of described every link, periodically carry out service distributing and comprise: according to the service distributing ratio of resulting every link

, periodically carry out service distributing.

On the other hand, the present invention also provides a kind of heterogeneous-network service separate system, comprising: the service request receiver module being linked in sequence, information detection module, shunting decision-making module and flow-dividing control module, wherein:

Service request receiver module, for receiving the service requesting information of multimode terminal, then solicited message detecting module carries out information detection;

Information detection module, carries out information detection for every link to heterogeneous network, periodically obtains every link buffer area queuing long message and data packet number, is sent to shunting decision-making module;

Shunting decision-making module, for according to the every link buffer area queuing long message periodically getting and data packet number, determines the service distributing ratio of every link, and result is sent to flow-dividing control module;

Flow-dividing control module, for according to the service distributing ratio of described every link, periodically carries out service distributing.

Further, described shunting decision-making module comprises:

Parameter adjustment submodule, for by every link queuing long message l _nwith the long threshold value L of link queuing comparison, determine the service speed μ of every link _n, n is the number of links in heterogeneous network;

Packet extracts submodule, for according to the data packet number mn of every the link periodically getting, extracts the data packet number m of every link when queuing up long threshold value L _nL;

Delay computing submodule, for according to the service speed μ of every link _nwith the data packet number m when queuing up long threshold value L _l, the packet that obtains every link postpones D _n;

Decision-making submodule, for the principle postponing according to minimization system, determines the service distributing ratio of every link.

Further, described parameter adjustment submodule comprises:

Parameter recommendation unit, for utilizing following formula to determine the service speed μ of every link _n:

${\mathrm{\&mu;}}_n=\left\{\begin{array}{cc}{\mathrm{\&mu;}}_{\mathrm{nl}}& {\mathrm{ifl}}_n<L\\ {\mathrm{\&mu;}}_{\mathrm{nm}}& \mathrm{otherwise}\end{array}\right.;$

μ wherein _nlfor the service speed higher limit of this link, μ _nmservice speed lower limit for this link.

Further, described Delay computing submodule comprises:

Long-term evolving network unit, for utilizing formula

the packet that obtains long-term evolving network link postpones, wherein λ _nfor the arrival rate of this link, this link traffic stream obedience parameter is λ _npoisson flow;

WLAN retransmission unit, for utilizing formula

obtaining long-term evolving network postpones through the packet of WLAN conversion link:

Wherein

for the packet delay of considering that WLAN disturbs flow to cause,

wherein λ ' is the parameter of WLAN interference flow Poisson flow, μ ' _narrival rate for this link after consideration WLAN interference flow;

the average data packet delay causing for long-term evolving network:

$\begin{array}{c}\begin{array}{c}{D}_n^{\mathrm{LTE}}={(\frac{1-{\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nl}}}\right)}^{m_{\mathrm{nL}}+1}}{1-\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nL}}}\right)}+\frac{{\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nl}}}\right)}^{m_{\mathrm{nL}}}\&CenterDot;\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{um}}}}{1-\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nm}}}\right)})}^{-1}\&CenterDot;\{\frac{\left(\frac{1}{{\mathrm{\&mu;}}_{\mathrm{nl}}}\right)(1+(m_{\mathrm{nL}}-1){\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nl}}}\right)}^{m_{\mathrm{nL}}}-m_{\mathrm{nL}}{\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nl}}}\right)}^{m_{\mathrm{nL}}-1})}{{(1-\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nl}}}\right))}^2}\\ +\frac{{\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nl}}}\right)}^{m_{\mathrm{nL}}-1}(m_{\mathrm{nL}}-(m_{\mathrm{nL}}-1)\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nm}}}\right))}{{(1-\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nm}}}\right))}^2}\}\end{array}\end{array}$

λ wherein _nfor the arrival rate of this link in long-term evolving network.

Further,

Described decision-making submodule comprises: optimize unit, the principle for postponing according to minimization system, solves following optimization problem:

make

and meet 0≤λ in long-term evolving network link _n≤ μ _n, long-term evolving network is 0≤λ in WLAN conversion link _n<min (μ _n, μ _n')-λ ', wherein λ is the parameter of total service traffics Poisson flow, vector

$\stackrel{\&RightArrow;}{\mathrm{\&lambda;}}=({\mathrm{\&lambda;}}_1,{\mathrm{\&lambda;}}_2,\&CenterDot;\&CenterDot;\&CenterDot;,{\mathrm{\&lambda;}}_n);$

Described flow-dividing control module comprises: shunting submodule, and for according to the service distributing ratio of resulting every link

periodically carry out service distributing.

(3) beneficial effect

Visible, in a kind of heterogeneous-network service shunt method and system proposing in the present invention, not only can periodically obtain every business feedback in link buffer area, detected information can also be incorporated in theoretical model, thereby the dynamic bind of the shunting decision-making technique of realization theory and engineering practice, has promoted accuracy and the initiative of shunt method.

In addition, the present invention has introduced random disturbances possible in network, according to the periodically variable network information and parameter, considers the disturbed condition in network, according to effective Diffluence Algorithm, draw rational shunting decision-making, further strengthened the adaptability to engineering practice.

Accompanying drawing explanation

In order to be illustrated more clearly in the embodiment of the present invention or technical scheme of the prior art, to the accompanying drawing of required use in embodiment or description of the Prior Art be briefly described below, apparently, accompanying drawing in the following describes is some embodiments of the present invention, for those of ordinary skills, do not paying under the prerequisite of creative work, can also obtain according to these accompanying drawings other accompanying drawing.

Fig. 1 is the basic procedure schematic diagram of embodiment of the present invention heterogeneous-network service shunt method;

Fig. 2 is the schematic flow sheet of a preferred embodiment of the invention heterogeneous-network service shunt method;

Fig. 3 is the schematic diagram of embodiment of the present invention heterogeneous network shunting transmitting scene;

Fig. 4 is embodiment of the present invention heterogeneous network shunting configuration diagram;

Fig. 5 is the basic structure schematic diagram of embodiment of the present invention heterogeneous-network service separate system;

Fig. 6 is the structural representation of a preferred embodiment of the invention heterogeneous-network service separate system;

Fig. 7 is the structural representation of a preferred embodiment of the invention heterogeneous-network service separate system;

Fig. 8 is the sequential chart of embodiment of the present invention heterogeneous-network service separate system agent structure.

Embodiment

For making object, technical scheme and the advantage of the embodiment of the present invention clearer, below in conjunction with the accompanying drawing in the embodiment of the present invention, technical scheme in the embodiment of the present invention is clearly and completely described, obviously, described embodiment is the present invention's part embodiment, rather than whole embodiment.Embodiment based in the present invention, those of ordinary skills, not making the every other embodiment obtaining under creative work prerequisite, belong to the scope of protection of the invention.

First the embodiment of the present invention proposes a kind of heterogeneous-network service shunt method, referring to Fig. 1, comprising:

Step 101: the service requesting information that receives multimode terminal.

Step 102: every link in heterogeneous network is carried out to information detection, periodically obtain every link buffer area queuing long message and data packet number.

Step 103: according to the every link buffer area queuing long message periodically getting and data packet number, determine the service distributing ratio of every link.

Step 104: according to the service distributing ratio of described every link, periodically carry out service distributing.

Visible, in a kind of heterogeneous-network service shunt method proposing in the embodiment of the present invention, not only can periodically obtain every business feedback in link buffer area, detected information can also be incorporated in theoretical model, thereby the dynamic bind of the shunting decision-making technique of realization theory and engineering practice, has promoted accuracy and the initiative of shunt method.

In one embodiment of the invention, preferably, according to the every link buffer area queuing long message periodically getting and data packet number, determine that the service distributing ratio of every link can comprise the steps:

By every link queuing long message l _nwith the long threshold value L of link queuing comparison, determine the service speed μ of every link _n, n is the number of links in heterogeneous network;

According to the data packet number m of every the link periodically getting _n, extract the data packet number m of every link when queuing up long threshold value L _nL;

According to the service speed μ of every link _nwith the data packet number m when queuing up long threshold value L _nL, the packet that obtains every link postpones D _n;

According to the principle of minimization system delay, determine the service distributing ratio of every link.

In another embodiment of the present invention, when using queueing theory to carry out modeling to heterogeneous network separate system, service traffics represent with arrival rate λ, and network availability bandwidth utilizes service speed μ to represent.Ratio between λ and μ is referred to as service intensity ρ, and the size of this value directly has relation with the long l that queues up.That is, when network state is not good, μ value is less, causes service intensity large, and team leader l is also long; Otherwise team leader l is less.This explanation, can judge that whether current network state is good by surveying the size of the long l of queuing of business in each Long Term Evolution (LTE) link, according to different network state revision μ values, thereby realize the improvement to queuing system, make separate system change and accomplish dynamically adapting network state.

To queuing up in long concrete detection process, can periodically send information detection instruction, according to the feedback information of buffer area in each LTE link, obtain every link queuing long message l _n.As every link queuing long message l _nwhen known, need to judge that whether network state corresponding to this team leader be good, and according to different l _nadjudicate rational μ value.Preferably, can profit determine with the following method the service speed μ of every link _n:

${\mathrm{\&mu;}}_n=\left\{\begin{array}{cc}{\mathrm{\&mu;}}_{\mathrm{nl}}& {\mathrm{ifl}}_n<L\\ {\mathrm{\&mu;}}_{\mathrm{nm}}& \mathrm{otherwise}\end{array}\right.;$

μ wherein _nlfor the service speed higher limit of this link, μ _nmservice speed lower limit for this link.

Above formula is determined the long threshold value L that queues up based on experience value, if survey the long l of queuing obtaining _nbe less than this threshold value, illustrate that current network state is good, μ _ncapping value μ _nl, otherwise explanation current network state is bad, takes off limit value μ _nm.By this way, can be so that Queueing Theory Method has adaptability to a certain degree to network time-varying characteristics.Analytical method that it should be noted that above formula is also a kind of approximate, and completely accurately description can increase the complexity of computing greatly, is unfavorable for engineering application.

In another embodiment of the present invention, preferably, can profit determine that with the following method the packet of every link postpones D _n:

For long-term evolving network link, the packet that produces postpone for via LTE directly the packet of transmission postpone, this part packet postpones can be in the hope of by Li Teer law:

$D_n=\frac{1}{{\mathrm{\&mu;}}_n-{\mathrm{\&lambda;}}_n},$

λ wherein _nfor the arrival rate of this link, this link traffic stream obedience parameter is λ _npoisson flow, be also the probability density function f (t) that adjacent two data bag arrives system time interval t, be parameter lambda _nnegative exponent distribute:

$f\left(t\right)={\mathrm{\&lambda;}}_n\&CenterDot;e^{{-\mathrm{\&lambda;}}_{n}t}.$

And for LTE network when the wlan network conversion link, the packet producing postpones to consider the service speed of LTE and wlan network, is equivalent to same Business Stream through the delay of twice service and queuing:

The delay wherein causing for wlan network,

wherein λ ' is the parameter of WLAN interference flow Poisson flow,

arrival rate for this link after consideration WLAN interference flow.

And the delay causing for LTE network becomes during due to its service speed, so the expression formula of average retardation can be:

$\begin{array}{c}\begin{array}{c}{D}_n^{\mathrm{LTE}}={(\frac{1-{\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nl}}}\right)}^{m_{\mathrm{nL}}+1}}{1-\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nL}}}\right)}+\frac{{\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nl}}}\right)}^{m_{\mathrm{nL}}}\&CenterDot;\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{um}}}}{1-\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nm}}}\right)})}^{-1}\&CenterDot;\{\frac{\left(\frac{1}{{\mathrm{\&mu;}}_{\mathrm{nl}}}\right)(1+(m_{\mathrm{nL}}-1){\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nl}}}\right)}^{m_{\mathrm{nL}}}-m_{\mathrm{nL}}{\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nl}}}\right)}^{m_{\mathrm{nL}}-1})}{{(1-\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nl}}}\right))}^2}\\ +\frac{{\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nl}}}\right)}^{m_{\mathrm{nL}}-1}(m_{\mathrm{nL}}-(m_{\mathrm{nL}}-1)\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nm}}}\right))}{{(1-\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nm}}}\right))}^2}\}\end{array}\end{array}$

λ wherein _nfor the arrival rate of this link in LTE network.

Thus, the delay of all links all can be calculated.In order to determine the suitable shunt ratio of Business Stream, for Queueing Theory Method, the main target of optimization that the minimum latency of can take is queuing system.In another embodiment of the present invention, the principle that can postpone according to minimization system, solves following optimization problem:

$\underset{\underset{\mathrm{\&lambda;}}{\&RightArrow;}}{\min }\underset{i=1}{\overset{n}{\mathrm{\&Sigma;}}}{D}_i,$

Make

and meet 0≤λ in LTE network link _n≤ μ _n, LTE network is 0≤λ in wlan network conversion link _n<min (μ _n, μ _n')-λ ', wherein λ is the parameter of total service traffics Poisson flow, vector

This constraints is that assurance business arrival rate can not surpass system service speed, is reflected in real network, total the traffic carrying capacity that the available total bandwidth of system can not be less than, even if otherwise also cannot guarantee user's business experience by fractionating system.

Can be in the hope of the solution of the optimum service distributing ratio under current network state by solving above-mentioned optimization problem

, preferably, can be according to the service distributing ratio of resulting every link

, periodically carry out service distributing.Also be basis in next cycle, adopt this shunt ratio to carry out service distributing, until carry out again the calculating of next cycle and redistribute.

With in a heterogeneous network shunting transmitting scene, the concrete grammar of shunting is example, describes the implementation procedure of the embodiment of the present invention in detail, as Fig. 2 below.

Figure 3 shows that the basic scene of heterogeneous network minute flow transmission, comprise minute streaming server, LTE base station and LTE network, WLAN AP node and a wlan network, a plurality of multimode terminals and interfered with terminal.The process of shunting for this heterogeneous network shunting transmitting scene comprises:

Step 201: multimode terminal user 1 is to a minute streaming server application business.

Step 202: minute streaming server sends and confirms solicited message to user 1.

Now, shunting discovering server user 1 can access LTE network and wlan network simultaneously, starts to carry out service distributing.

Step 203: a minute streaming server carries out information detection to every link in heterogeneous network.

Step 204: a minute streaming server periodically obtains every link buffer area queuing long message and data packet number.

In the embodiment of the present invention, when using queueing theory to carry out modeling to heterogeneous network separate system, service traffics represent with arrival rate λ, and network availability bandwidth utilizes service speed μ to represent.Ratio between λ and μ is referred to as service intensity ρ, and the size of this value directly has relation with the long l that queues up.That is, when network state is not good, μ value is less, causes service intensity large, and team leader l is also long; Otherwise team leader l is less.This explanation, can judge that whether current network state is good by surveying the size of the long l of queuing of business in each LTE link, according to different network state revision μ values, thereby realize the improvement to queuing system, make separate system change and accomplish dynamically adapting network state.

Step 205: by every link queuing long message ln and the long threshold value L of link queuing comparison, determine the service speed μ of every link _n.

In this step, as every link queuing long message l _nwhen known, need to judge that whether network state corresponding to this team leader be good, and according to different l _nadjudicate rational μ value:

${\mathrm{\&mu;}}_n=\left\{\begin{array}{cc}{\mathrm{\&mu;}}_{\mathrm{nl}}& {\mathrm{ifl}}_n<L\\ {\mathrm{\&mu;}}_{\mathrm{nm}}& \mathrm{otherwise}\end{array}\right.;$

μ wherein _nlfor the service speed higher limit of this link, μ _nmservice speed lower limit for this link.

Above formula is determined the long threshold value L that queues up based on experience value, if survey the long l of queuing obtaining _nbe less than this threshold value, illustrate that current network state is good, μ _ncapping value μ _nl, otherwise explanation current network state is bad, takes off limit value μ _nm.By this way, can be so that Queueing Theory Method has adaptability to a certain degree to network time-varying characteristics.

Step 206: extract the data packet number m of every link when queuing up long threshold value L _nL.

In the embodiment of the present invention, for be minute flow transmission of single kind business, the type of service in system only has a kind of.In this step, when queuing up the long threshold value L of being, the data packet length of this class of business that order is obtained is l _d, data packet number m _nLcan be directly by L divided by l _dobtain.

Step 207: according to the service speed μ of every link _nwith the data packet number m when queuing up long threshold value L _nL, the packet that obtains every link postpones D _n.

For long-term evolving network link, the packet that the packet producing postpones for directly transmitting via LTE postpones, and this part packet postpones D ₁can be in the hope of by Li Teer law:

$D_1=\frac{1}{{\mathrm{\&mu;}}_1-{\mathrm{\&lambda;}}_1},$

λ wherein ₁arrival rate for this link.

And for LTE network when the wlan network conversion link, the packet producing postpones D ₂to D _nneed to consider the service speed of LTE and wlan network, be equivalent to same Business Stream through the delay of twice service and queuing:

The delay wherein causing for wlan network,

wherein λ ' is the parameter of WLAN interference flow Poisson flow,

arrival rate for this link after consideration WLAN interference flow.

And the delay causing for LTE network becomes during due to its service speed, so the expression formula of average retardation can be:

$\begin{array}{c}\begin{array}{c}{D}_n^{\mathrm{LTE}}={(\frac{1-{\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nl}}}\right)}^{m_{\mathrm{nL}}+1}}{1-\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nL}}}\right)}+\frac{{\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nl}}}\right)}^{m_{\mathrm{nL}}}\&CenterDot;\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{um}}}}{1-\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nm}}}\right)})}^{-1}\&CenterDot;\{\frac{\left(\frac{1}{{\mathrm{\&mu;}}_{\mathrm{nl}}}\right)(1+(m_{\mathrm{nL}}-1){\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nl}}}\right)}^{m_{\mathrm{nL}}}-m_{\mathrm{nL}}{\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nl}}}\right)}^{m_{\mathrm{nL}}-1})}{{(1-\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nl}}}\right))}^2}\\ +\frac{{\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nl}}}\right)}^{m_{\mathrm{nL}}-1}(m_{\mathrm{nL}}-(m_{\mathrm{nL}}-1)\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nm}}}\right))}{{(1-\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nm}}}\right))}^2}\}\end{array}\end{array}$

λ wherein _nfor the arrival rate of this link in LTE network.

Step 208: according to the principle of minimization system delay, determine the service distributing ratio of every link.

The principle postponing according to minimization system, can determine by solving following optimization problem the service distributing ratio of every link:

$\underset{\mathrm{\&lambda;}}{\underset{\&RightArrow;}{\min }}\underset{i=1}{\overset{n}{\mathrm{\&Sigma;}}}{D}_i$

s.t.

$\underset{i=1}{\overset{n}{\mathrm{\&Sigma;}}}{\mathrm{\&lambda;}}_i=\mathrm{\&lambda;}$

0≤λ _n≤μ _n

0≤λ _n<min(μ _n,μ _n')-λ' i=2,3,…,n

Wherein constraints guarantees that business arrival rate can not surpass system service speed, total reaction is the traffic carrying capacity that the available total bandwidth of system can not be less than in real network, even if otherwise also cannot guarantee user's business experience by separate system.

Step 209: according to the service distributing ratio of every link, periodically carry out service distributing.

Can be in the hope of the solution of the optimum service distributing ratio under current network state by the optimization problem solving above

,

in next cycle, adopt this shunt ratio to carry out service distributing, until carry out again the calculating of next cycle and redistribute.Final resulting shunting configuration diagram as shown in Figure 4.

So far, completed the process of embodiment of the present invention heterogeneous-network service shunt method.

In addition, it should be noted that, it is a kind of preferred implementation procedure of heterogeneous-network service shunt method of the present invention that above-mentioned all flow processs based on Fig. 2 are described, in the actual realization of heterogeneous-network service shunt method of the present invention, can on the basis of flow process shown in Fig. 1, carry out as required random variation, can be to select the arbitrary steps in Fig. 2 to realize, the sequencing of each step also can be adjusted etc. as required.

The embodiment of the present invention also provides a kind of heterogeneous-network service separate system, as shown in Figure 5, comprising:

Service request receiver module 501, for receiving the service requesting information of multimode terminal, then solicited message detecting module carries out information detection;

Information detection module 502, carries out information detection for every link to heterogeneous network, periodically obtains every link buffer area queuing long message and data packet number, is sent to shunting decision-making module;

Shunting decision-making module 503, for according to the every link buffer area queuing long message periodically getting and data packet number, determines the service distributing ratio of every link, and result is sent to flow-dividing control module;

Flow-dividing control module 504, for according to the service distributing ratio of described every link, periodically carries out service distributing.

In one embodiment of the invention, preferably, shunting decision-making module 503 can comprise:

Parameter adjustment submodule 601, as shown in Figure 6, for by every link queuing long message l _nwith the long threshold value L of link queuing comparison, determine the service speed μ of every link _n, n is the number of links in heterogeneous network;

Packet extracts submodule 602, for according to the data packet number m of every the link periodically getting _n, extract the data packet number m of every link when queuing up long threshold value L _nL;

Delay computing submodule 603, for according to the service speed μ of every link _nwith the data packet number m when queuing up long threshold value L _l, the packet that obtains every link postpones D _n;

Decision-making submodule 604, for the principle postponing according to minimization system, determines the service distributing ratio of every link.

In another embodiment of the present invention, when using queueing theory to carry out modeling to heterogeneous network separate system, service traffics represent with arrival rate λ, and network availability bandwidth utilizes service speed μ to represent.Ratio between λ and μ is referred to as service intensity ρ, and the size of this value directly has relation with the long l that queues up.That is, when network state is not good, μ value is less, causes service intensity large, and team leader l is also long; Otherwise team leader l is less.This explanation, can judge that whether current network state is good by surveying the size of the long l of queuing of business in each LTE link, according to different network state revision μ values, thereby realize the improvement to queuing system, make separate system change and accomplish dynamically adapting network state.

To queuing up in long concrete detection process, can periodically send information detection instruction, according to the feedback information of buffer area in each LTE link, obtain every link queuing long message l _n.As every link queuing long message l _nwhen known, need to judge that whether network state corresponding to this team leader be good, and according to different l _nadjudicate rational μ value.Preferably, parameter adjustment submodule 601 can comprise:

Parameter recommendation unit 701, as Fig. 7, for utilizing following formula to determine the service speed μ of every link _n:

${\mathrm{\&mu;}}_n=\left\{\begin{array}{cc}{\mathrm{\&mu;}}_{\mathrm{nl}}& {\mathrm{ifl}}_n<L\\ {\mathrm{\&mu;}}_{\mathrm{nm}}& \mathrm{otherwise}\end{array}\right.;$

μ wherein _nlfor the service speed higher limit of this link, μ _nmservice speed lower limit for this link.

Above formula is determined the long threshold value L that queues up based on experience value, if survey the long l of queuing obtaining _nbe less than this threshold value, illustrate that current network state is good, μ _ncapping value μ _nl, otherwise explanation current network state is bad, takes off limit value μ _nm.By this way, can be so that Queueing Theory Method has adaptability to a certain degree to network time-varying characteristics.Analytical method that it should be noted that above formula is also a kind of approximate, and completely accurately description can increase the complexity of computing greatly, is unfavorable for engineering application.

In another embodiment of the present invention, preferably, Delay computing submodule 603 can comprise:

Long-term evolving network unit 702, for utilizing formula

the packet that obtains long-term evolving network link postpones, wherein λ _nfor the arrival rate of this link, this link traffic stream obedience parameter is λ _npoisson flow;

WLAN retransmission unit 703, for utilizing formula

obtaining long-term evolving network postpones through the packet of WLAN conversion link:

Wherein

for the packet delay of considering that WLAN disturbs flow to cause,

wherein λ ' is the parameter of WLAN interference flow Poisson flow,

arrival rate for this link after consideration WLAN interference flow;

the average data packet delay causing for long-term evolving network:

$\begin{array}{c}\begin{array}{c}{D}_n^{\mathrm{LTE}}={(\frac{1-{\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nl}}}\right)}^{m_{\mathrm{nL}}+1}}{1-\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nL}}}\right)}+\frac{{\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nl}}}\right)}^{m_{\mathrm{nL}}}\&CenterDot;\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{um}}}}{1-\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nm}}}\right)})}^{-1}\&CenterDot;\{\frac{\left(\frac{1}{{\mathrm{\&mu;}}_{\mathrm{nl}}}\right)(1+(m_{\mathrm{nL}}-1){\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nl}}}\right)}^{m_{\mathrm{nL}}}-m_{\mathrm{nL}}{\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nl}}}\right)}^{m_{\mathrm{nL}}-1})}{{(1-\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nl}}}\right))}^2}\\ +\frac{{\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nl}}}\right)}^{m_{\mathrm{nL}}-1}(m_{\mathrm{nL}}-(m_{\mathrm{nL}}-1)\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nm}}}\right))}{{(1-\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nm}}}\right))}^2}\}\end{array}\end{array}$

λ wherein _nfor the arrival rate of this link in long-term evolving network.

Thus, the delay of all links all can be calculated.In order to determine the suitable shunt ratio of Business Stream, for Queueing Theory Method, the main target of optimization that the minimum latency of can take is queuing system.In another embodiment of the present invention, decision-making submodule 604 can comprise: optimize unit 704, the principle for postponing according to minimization system, solves following optimization problem:

make

and meet 0≤λ in long-term evolving network link _n≤ μ _n, long-term evolving network is 0≤λ in WLAN conversion link _n<min (μ _n, μ ' _n)-λ ', wherein λ is the parameter of total service traffics Poisson flow, vector

And flow-dividing control module 504 can comprise: shunting submodule 605, for according to the service distributing ratio of resulting every link

, periodically carry out service distributing.Also be basis in next cycle, adopt this shunt ratio to carry out service distributing, until carry out again the calculating of next cycle and redistribute.

It should be noted that, the structure of each embodiment of the heterogeneous-network service separate system shown in above-mentioned Fig. 6, Fig. 7 can be carried out combination in any use.The sequential chart of whole system agent structure as shown in Figure 8.

Visible, the embodiment of the present invention has following beneficial effect:

In a kind of heterogeneous-network service shunt method and system proposing in the embodiment of the present invention, not only can periodically obtain every business feedback in link buffer area, detected information can also be incorporated in theoretical model, thereby the dynamic bind of the shunting decision-making technique of realization theory and engineering practice, has promoted accuracy and the initiative of shunt method.

In addition, the embodiment of the present invention has been introduced random disturbances possible in network, according to the periodically variable network information and parameter, considers the disturbed condition in network, according to effective Diffluence Algorithm, draw rational shunting decision-making, further strengthened the adaptability to engineering practice.

Finally it should be noted that: above embodiment only, in order to technical scheme of the present invention to be described, is not intended to limit; Although the present invention is had been described in detail with reference to previous embodiment, those of ordinary skill in the art is to be understood that: its technical scheme that still can record aforementioned each embodiment is modified, or part technical characterictic is wherein equal to replacement; And these modifications or replacement do not make the essence of appropriate technical solution depart from the spirit and scope of various embodiments of the present invention technical scheme.

Claims (10)

1. a heterogeneous-network service shunt method, is characterized in that, comprising:

Receive the service requesting information of multimode terminal;

Every link in heterogeneous network is carried out to information detection, periodically obtain every link buffer area queuing long message and data packet number;

According to the every link buffer area queuing long message periodically getting and data packet number, determine the service distributing ratio of every link;

According to the service distributing ratio of described every link, periodically carry out service distributing.

2. heterogeneous-network service shunt method according to claim 1, is characterized in that, described according to the every link buffer area queuing long message periodically getting and data packet number, determines that the service distributing ratio of every link comprises:

By every link queuing long message l _nwith the long threshold value L of link queuing comparison, determine the service speed μ of every link _n, n is the number of links in heterogeneous network;

According to the data packet number m of every the link periodically getting _n, extract the data packet number m of every link when queuing up long threshold value L _nL;

According to the service speed μ of every link _nwith the data packet number m when queuing up long threshold value L _nL, the packet that obtains every link postpones D _n;

According to the principle of minimization system delay, determine the service distributing ratio of every link.

3. heterogeneous-network service shunt method according to claim 2, is characterized in that, described by every link queuing long message l _nwith the long threshold value L of link queuing comparison, determine the service speed μ of every link _ncomprise:

Profit is determined the service speed μ of every link with the following method _n:

${\mathrm{\&mu;}}_n=\left\{\begin{array}{cc}{\mathrm{\&mu;}}_{\mathrm{nl}}& {\mathrm{ifl}}_n<L\\ {\mathrm{\&mu;}}_{\mathrm{nm}}& \mathrm{otherwise}\end{array}\right.;$

μ wherein _nlfor the service speed higher limit of this link, μ _nmservice speed lower limit for this link.

4. heterogeneous-network service shunt method according to claim 3, is characterized in that,

Described according to the service speed μ of every link _nwith the data packet number m when queuing up long threshold value L _l, the packet that obtains every link postpones D _ncomprise:

When this link is long-term evolving network link,

λ wherein _nfor the arrival rate of this link, this link traffic stream obedience parameter is λ _npoisson flow;

When this link is long-term evolving network during through WLAN conversion link,

$D_n=D_n^{\mathrm{WLAN}}+D_n^{\mathrm{LTE}},$

Wherein

for the packet delay of considering that WLAN disturbs flow to cause,

wherein λ ' is the parameter of WLAN interference flow Poisson flow, μ ' _narrival rate for this link after consideration WLAN interference flow;

the average data packet delay causing for long-term evolving network:

$\begin{array}{c}\begin{array}{c}{D}_n^{\mathrm{LTE}}={(\frac{1-{\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nl}}}\right)}^{m_{\mathrm{nL}}+1}}{1-\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nL}}}\right)}+\frac{{\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nl}}}\right)}^{m_{\mathrm{nL}}}\&CenterDot;\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{um}}}}{1-\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nm}}}\right)})}^{-1}\&CenterDot;\{\frac{\left(\frac{1}{{\mathrm{\&mu;}}_{\mathrm{nl}}}\right)(1+(m_{\mathrm{nL}}-1){\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nl}}}\right)}^{m_{\mathrm{nL}}}-m_{\mathrm{nL}}{\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nl}}}\right)}^{m_{\mathrm{nL}}-1})}{{(1-\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nl}}}\right))}^2}\\ +\frac{{\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nl}}}\right)}^{m_{\mathrm{nL}}-1}(m_{\mathrm{nL}}-(m_{\mathrm{nL}}-1)\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nm}}}\right))}{{(1-\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nm}}}\right))}^2}\}\end{array}\end{array}$

λ wherein _nfor the arrival rate of this link in long-term evolving network.

5. heterogeneous-network service shunt method according to claim 4, is characterized in that:

The described principle postponing according to minimization system, determine that the service distributing ratio of every link comprises:

The principle postponing according to minimization system, solves following optimization problem:

make

and meet 0≤λ in long-term evolving network link _n≤ μ _n, long-term evolving network is 0≤λ in WLAN conversion link _n<min (μ _n, μ ' _n)-λ ', wherein λ is the parameter of total service traffics Poisson flow, vector

Described according to the service distributing ratio of described every link, periodically carry out service distributing and comprise: according to the service distributing ratio of resulting every link

, periodically carry out service distributing.

6. a heterogeneous-network service separate system, is characterized in that, comprising: the service request receiver module being linked in sequence, information detection module, shunting decision-making module and flow-dividing control module, wherein:

Service request receiver module, for receiving the service requesting information of multimode terminal, then solicited message detecting module carries out information detection;

Information detection module, carries out information detection for every link to heterogeneous network, periodically obtains every link buffer area queuing long message and data packet number, is sent to shunting decision-making module;

Shunting decision-making module, for according to the every link buffer area queuing long message periodically getting and data packet number, determines the service distributing ratio of every link, and result is sent to flow-dividing control module;

Flow-dividing control module, for according to the service distributing ratio of described every link, periodically carries out service distributing.

7. heterogeneous-network service separate system according to claim 6, is characterized in that, described shunting decision-making module comprises:

Parameter adjustment submodule, for by every link queuing long message ln and the long threshold value L of link queuing comparison, determines the service speed μ of every link _n, n is the number of links in heterogeneous network;

Packet extracts submodule, for according to the data packet number m of every the link periodically getting _n, extract the data packet number m of every link when queuing up long threshold value L _nL;

Delay computing submodule, for according to the service speed μ of every link _nwith the data packet number m when queuing up long threshold value L _l, the packet that obtains every link postpones D _n;

Decision-making submodule, for the principle postponing according to minimization system, determines the service distributing ratio of every link.

8. heterogeneous-network service separate system according to claim 7, is characterized in that, described parameter adjustment submodule comprises:

Parameter recommendation unit, for utilizing following formula to determine the service speed μ of every link _n:

${\mathrm{\&mu;}}_n=\left\{\begin{array}{cc}{\mathrm{\&mu;}}_{\mathrm{nl}}& {\mathrm{ifl}}_n<L\\ {\mathrm{\&mu;}}_{\mathrm{nm}}& \mathrm{otherwise}\end{array}\right.;$

μ wherein _nlfor the service speed higher limit of this link, μ _nmservice speed lower limit for this link.

9. heterogeneous-network service separate system according to claim 8, is characterized in that, described Delay computing submodule comprises:

Long-term evolving network unit, for utilizing formula the packet that obtains long-term evolving network link postpones, wherein λ _nfor the arrival rate of this link, this link traffic stream obedience parameter is λ _npoisson flow;

WLAN retransmission unit, for utilizing formula

obtaining long-term evolving network postpones through the packet of WLAN conversion link:

Wherein

for the packet delay of considering that WLAN disturbs flow to cause, wherein λ ' is the parameter of WLAN interference flow Poisson flow, μ ' _narrival rate for this link after consideration WLAN interference flow;

the average data packet delay causing for long-term evolving network:

$\begin{array}{c}\begin{array}{c}{D}_n^{\mathrm{LTE}}={(\frac{1-{\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nl}}}\right)}^{m_{\mathrm{nL}}+1}}{1-\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nL}}}\right)}+\frac{{\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nl}}}\right)}^{m_{\mathrm{nL}}}\&CenterDot;\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{um}}}}{1-\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nm}}}\right)})}^{-1}\&CenterDot;\{\frac{\left(\frac{1}{{\mathrm{\&mu;}}_{\mathrm{nl}}}\right)(1+(m_{\mathrm{nL}}-1){\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nl}}}\right)}^{m_{\mathrm{nL}}}-m_{\mathrm{nL}}{\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nl}}}\right)}^{m_{\mathrm{nL}}-1})}{{(1-\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nl}}}\right))}^2}\\ +\frac{{\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nl}}}\right)}^{m_{\mathrm{nL}}-1}(m_{\mathrm{nL}}-(m_{\mathrm{nL}}-1)\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nm}}}\right))}{{(1-\left(\frac{{\mathrm{\&lambda;}}_n}{{\mathrm{\&mu;}}_{\mathrm{nm}}}\right))}^2}\}\end{array}\end{array}$ λ wherein _nfor the arrival rate of this link in long-term evolving network.

10. heterogeneous-network service separate system according to claim 9, is characterized in that:

Described decision-making submodule comprises: optimize unit, the principle for postponing according to minimization system, solves following optimization problem:

make

and meet 0≤λ in long-term evolving network link _n≤ μ _n, long-term evolving network is 0≤λ in WLAN conversion link _n<min (μ _n, μ ' _n)-λ ', wherein λ is the parameter of total service traffics Poisson flow, vector

$\stackrel{\&RightArrow;}{\mathrm{\&lambda;}}=({\mathrm{\&lambda;}}_1,{\mathrm{\&lambda;}}_2,\&CenterDot;\&CenterDot;\&CenterDot;,{\mathrm{\&lambda;}}_n);$

Described flow-dividing control module comprises: shunting submodule, and for according to the service distributing ratio of resulting every link periodically carry out service distributing.

Images