CN102572987B - Network selection method orienting to heterogeneous wireless network environment - Google Patents

Abstract

The invention provides a network selection method orienting to a heterogeneous wireless network environment. The network selection method specifically comprises the following steps: (1) regarding all users accessing to a same network and single users rejected in network access as a coalition, and establishing a multi-service utility function; (2) establishing a coalition game model according to the acquired utility function in (1) and a set of all users; (3) introducing user enjoyment preferences and constructing the coalition game model into an enjoyment coalition game model; (4) establishing a conversion rule based on the enjoyment preferences, and carrying out distributed type coalition formation by each user according to the rule; (5) converging the coalition formation in a stable coalition structure which gives a network selection result; and (6) periodically updating a coalition formation process so as to adapt to the dynamic changes of the heterogeneous wireless environment. The method disclosed by the invention has the advantages of distributiveness, dynamic property, environment adaptiveness and the like, and can be used for acquiring preferable total utility of a heterogeneous system and preferable user acceptance sum, wherein the acquired total utility approaches the maximal value of the total utility of the heterogeneous system.

Description

A kind of network selecting method towards heterogeneous wireless network environment

Technical field

The present invention relates to wireless communication technology field, be specifically related to a kind of network selecting method of heterogeneous wireless network.

Background technology

Along with the develop rapidly of the communication technology, access network presents the heterogeneous wireless network environment that multiple wireless access technology coexists.In order to overcome the restriction of resource in radio communication, meet the access demand of any time, any place, any business, there is provided the IP multimedia service of different QoS, two-forty, the heterogeneous wireless network (heterogeneous wireless networks) realizing syncretization has become the inexorable trend of wireless communication technology development.

Under wireless isomer environment, one is possessed to the multimode terminal user of access multiple network, the key realizing seamless service link is that how to choose optimum network in multiple coverage area accesses.This is because different candidate network is at its framework, and signal conditioning, all there is difference the aspects such as service quality guarantee, and the network choosing non-optimal may cause worse service experience or higher service cost.For this reason, network is selected to be the key problem in heterogeneous wireless network.

In recent years, network selecting method is with the study hotspot of mechanism, on the one hand, in order to adapt to the dynamic of wireless environment height, requires that network selects mechanism also to have higher dynamic; On the other hand, by means of low complex degree advantage and do not rely on any Central Control Facility that each network is coordinated, network select move towards distributed from centralization just gradually, from network terminal decision-making move towards user side drive.

Summary of the invention

The object of the invention is to propose a kind of network selecting method towards heterogeneous wireless network environment, the visual angle of coalition formation is incorporated into heterogeneous network field by the method first, have distributed, the feature of mobilism, reduce the complexity that network is selected, simultaneously can by the process of self-organization of coalition formation, the complexity change of the adaptation wireless environment of mobilism, promote heterogeneous system total utility and maximum receiving total number of users, and the maximum of extremely convergence heterogeneous system total utility.

Towards a network selecting method for heterogeneous wireless network environment, comprise following steps:

Step 1) t accessed all users of consolidated network or be not labeled as alliance S by the unique user of network insertion _k, set up each alliance S according to the type of service that user asks _kalliance total utility function v (S _k);

Step 2) set up Game with Coalitions model (N (t), the v (S of transferable utility _k)), the user that N (t) is t request service gathers, according to fairness allocation criteria by alliance total utility function v (S _k) distribute to alliance S _kinterior user, obtains the individual revenue function φ of each user _n(S _k);

Step 3) according to user n ∈ N (t) individual revenue function φ _n(S _k) designing user n is to alliance S _kpreference function f _n(S _k), according to preference function f _n(S _k) for user n is to the alliance S of any two its possibility subordinates _α, S _βset up complete, anti-body, a transferable binary preference relation S _α>= _ns _β, and then by Game with Coalitions model (N (t), v (S _k)) be reconstructed into enjoyment Game with Coalitions model (N (t),>= _n);

Step 4) foundation enjoyment Game with Coalitions model (N (t),>= _n), user n is shifted to the alliance of a preference more from the alliance of current affiliation, and iteration transposition is until user n can not be shifted to new alliance more successively;

Step 5) user n selects in step 4) network corresponding to final alliance after iteration transposition be as target access network.

Further, also comprise step 6) periodically repeat step 4) with step 5) alliance's transpose process, with the dynamic change of self adaptation heterogeneous wireless environment.

Further, described alliance total utility function v (S _k) the characterizing consumer service revenue of serving to objective network requested service and service cost.

Towards a network selecting method for heterogeneous wireless network environment, comprise following steps:

T a) is accessed all users of consolidated network or is not labeled as alliance S by the unique user of network insertion by step _k, the type of service of asking according to user sets up the individual utility function v of each user _n(S _k);

Step b) set up Game with Coalitions model (N (t), the V (S of non-transferable utility _k)), the user that N (t) is t request service gathers, V (S _k) be by alliance S _kthe individual utility function v of interior each user _n(S _k) vector that forms;

Step c) according to user n ∈ N (t) individual utility function v _n(S _k) designing user n is to alliance S _kpreference function f _n(S _k), according to preference function f _n(S _k) for user n is to the alliance S of any two its possibility subordinates _α, S _βset up one complete, anti-body with transferable binary preference relation S _α>= _ns _βbe designated as>= _n, and then by Game with Coalitions model (N (t), v (S _k)) be reconstructed into enjoyment Game with Coalitions model (N (t),>= _n);

Steps d) foundation enjoyment Game with Coalitions model (N (t),>= _n), user n changes to the alliance of a preference more from the alliance of current affiliation, and iteration transposition is until user n can not be shifted to new alliance more successively;

Step e) user n selects in steps d) network corresponding to final alliance after iteration transposition be as target access network.

Further, also comprise step f) periodically repeat steps d) with step e) alliance's transpose process, with the dynamic change of self adaptation heterogeneous wireless environment.

Further, described individual utility function v _n(S _k) the characterizing consumer service revenue of serving to objective network requested service and service cost.

Technique effect of the present invention is embodied in:

A kind of network selecting method towards heterogeneous wireless network environment of the present invention, first with the network select permeability of the time of coalition formation close examination heterogeneous network, there is born dynamic and distributed nature, by organizing themselves into stable coalition structure, can adaptive wireless system complex environment change, the movement of such as user, the life of calling is gone out, the switching etc. of type of service, simultaneously, compare traditional network selecting method based on received signal strength (RSNS), greatly can improve the maximum user Tong Na number of system, the overall utility of elevator system, and level off to the maximum of system total utility.

Accompanying drawing explanation

Fig. 1 is a kind of network selecting method flow chart towards heterogeneous wireless network environment of the present invention;

Fig. 2 is a heterogeneous wireless network system example schematic diagram;

Fig. 3 is multi-services revenue function curve chart, and Fig. 3 is a) multi-services revenue function curve chart when bandwidth is 0-10000Kbps, Fig. 3 b) be the amplification situation of Fig. 3 a) when bandwidth is 0-900Kbps;

Fig. 4 is the network selection result schematic diagram under 35 user distributions;

Fig. 5 is that system total utility value compares schematic diagram;

Fig. 6 is in order to tackle the dynamic change of heterogeneous system and the mobilism network selection procedures schematic diagram carried out;

Network selection result schematic diagram after Fig. 7 dynamic change;

The comparison schematic diagram of the maximum accommodation number of users of Fig. 8 heterogeneous system.

Embodiment

In order to more clearly embody the present invention's concrete implementation of a kind of network selecting method towards heterogeneous wireless network environment and advantage, be described in further details below in conjunction with drawings and the specific embodiments.Obviously, described embodiment is only a part of embodiment of the present invention, and the every other embodiment that those of ordinary skill in the art obtain under the prerequisite not paying creative work all belongs to the scope of protection of the invention.

Fig. 2 gives embodiment a heterogeneous wireless network system used, comprises a following K=4 network: 1 worldwide interoperability for microwave access WiMAX network, is labeled as network 1, i.e. Internet Service Provider SP ₁(without loss of generality, the present invention suppose each network manage by the Internet Service Provider), 1 WLAN WLAN based on IEEE802.11 agreement, is labeled as network 3, i.e. SP ₃, and the Wideband Code Division Multiple Access (WCDMA) cellular network WCDMA that 2 adjacent, be labeled as network 2, SP respectively ₂with network 4, SP ₄, different network k has different covering radius D _kwith maximum bandwidth B _k, Fig. 2 significantly can find out that this heterogeneous wireless environment exists multiple coverage area, and table 1 gives the concrete network parameter values that the present embodiment adopts.In the coverage of whole heterogeneous system the coverage of network 1 (i.e. in), the network user of random distribution and dynamic change needs to choose suitable objective network and accesses.Be alliance S user's aggregated label of access consolidated network k _k, wherein 1≤k≤K, temporarily being stoped the user of access to be considered as unmarried alliance S by system _k, wherein k>K.Any active ues being request service that " user " considers in the present invention.

The network parameter of table 1 heterogeneous wireless system

The present invention below sets up user utility function, and in this example, the present invention will directly set up alliance total utility function v (S _k): without loss of generality, the present invention considers each SP _kthree kinds of business J={voice (voice) can be provided, data (data), video (video) }, each user n is to a kind of business service j ∈ of each network request J, each SP _kcan be used for this user R _n,kbandwidth, the service revenue function u that like this can be defined as follows _j(R _n,k):

$u_{\mathrm{data}}\left(R_{n,k}\right)=\frac{{\log }_2((R_{n,k}+15)/15)}{1.21}$

Fig. 3 gives the service revenue function curve of above-mentioned 3 kinds of business.Can see, definition and the service feature of this revenue function are closely related, characterize the Consumer's Experience of different business under different bandwidth condition or service satisfactory level.Like this, to the user coalitions S of any one access network k _k, wherein 1≤k≤K, its service revenue summation can be expressed as:

$U\left(S_j\right)=\underset{n\∈S_k}{\mathrm{\Σ}}{u}_j\left(R_{n,k}\right),\∀j\∈J,k=\mathrm{1,2},...,K$

Next, consider that service pays.Each service provider receives user's calling, all collects a cost value, further, takes a kind of differential pricing strategy, namely different SP _kdifferent cost value is collected to different service calls, such as, WCDMA network is for the service call of high priority, such as audio call, can collect low expense, similar, wlan network prefers data service, these preferences are embodied on can being fixed a price by differentiation, therefore, and the price matrix P of the 3 × K that can be defined as follows:

$P=\left(\begin{array}{ccc}{p}_1^{\mathrm{voice}}& ...& p_K^{\mathrm{voice}}\\ p_1^{\mathrm{data}}& ...& p_K^{\mathrm{data}}\\ p_1^{\mathrm{video}}& ...& p_K^{\mathrm{video}}\end{array}\right)$

Wherein represent network k or Internet Service Provider SP _kto the expense that the calling that is j of each business from user is collected.Concrete, the price matrix that the present embodiment adopts is:

$P_{3\×4}=\left(\begin{array}{cccc}0.04& 0.02& 0.08& 0.02\\ 2.5& 3.5& 2.0& 3.5\\ 3.5& 4.0& 3.0& 4.0\end{array}\right)$

Finally, signal receiving strength (Received Signal Strength, the RSS) impact on value of utility function is considered.Signal receiving strength is one of decision criteria generally adopted during heterogeneous network network is selected, and it can embody the quality of reception of network signal, and obviously, user is more ready to access the network providing higher RSS.Because wireless networks different in heterogeneous system is in the difference of maximum transmission power and threshold level, for this reason, can list of references, W.Shen and Q.A.Zeng, " Cost-function-based networkselection strategy in integrated wireless and mobile networks, " IEEETransactions on Vehicular Technology, vol.57, no.6, pp.3778-3788, November2008. (syncretization is wireless with mobile network under based on the network selection policies of cost function, IEEE Vehicle Technology transactions, 57 volumes the 6th phase, in November, 2008) define a kind of relative signal receiving intensity (RelativeReceived Signal Strength, RRSS), the RRSS of network k is defined as follows:

${\mathrm{RRSS}}_k=1-\frac{\log \left(d_{n,k}\right)}{\log \left(D_k\right)},$ Wherein k=1,2 ..., K

Wherein d _n,kuser n to SP _kdistance.For embodiment, lower RRSS obviously can reduce business service income or the satisfactory level of terminal use, for this reason, can define the subtraction function of a RRSS, as the Part II of cost function.To any one user coalitions S _k, wherein 1≤k≤K, this cost function is defined as:

$\begin{array}{c}C\left(S_k\right)=\underset{n\∈S_k}{\mathrm{\Σ}}(1-{\mathrm{RRSS}}_k(d_{n,k}+1))\\ =\frac{\log \left(\underset{n\∈S_k}{\mathrm{\Π}}(d_{n,k}+1)\right)}{\log \left(D_k\right)}\end{array}$

So, to any one user coalitions S _k, it is as follows that the present invention can provide its alliance's total utility function:

This utility function characterizes any one user coalitions S _kin the total utility value that t obtains.Wherein, Section 1 represents that the value of utility that the user coalitions of access network k obtains is that service revenue deducts cost function; Section 2 represents user not by network insertion, and when namely becoming unmarried alliance, its value of utility is 0, and Section 3 represents if alliance S _kmiddle existence is in the user outside network k covering radius, then this alliance can not be formed.Notice, this utility function is based on multi-services income or satisfactory level, and access pays, and received signal strength cost function three part directly establishes the total utility of an alliance, without loss of generality, consider that the utility function model of other parameters and form is also feasible.

So far, a transferable utility Game with Coalitions model (TU coalitional game) (N (t), v (S is established _k)), wherein effectiveness v (S _k) also referred to as the characteristic function (characteristic function) of betting model, for the Game with Coalitions of transferable utility, characterize alliance S _ktotal utility value can divide with any regular between its member.Therefore, next need to define a kind of fairness allocation criteria, to give each user individual income.Common allocation criteria comprises the maximum-minimize fair allocat solution based on Nucleolus, and Sharply Value distributes, and in the present embodiment, takes the relatively simple equality fairness scheme of one, namely to each user n ∈ S _k, its individual income is:

${\mathrm{\φ}}_n\left(S_k\right)=\frac{1}{\left|S_k\right|}(v\left(S_k\right)-\underset{m\∈S_k}{\mathrm{\Σ}}v\left(\right\{m\left\}\right))+v\left(\right\{n\left\}\right)$

Because user accesses different SP _kdifferent characteristic function values can be obtained, according to the definition of Game with Coalitions opinion, SP _kneed the special user being considered as this betting model, therefore, alliance S _k(1≤k≤K) is equivalent to { SP _k, S _k, and alliance S _k(k>K) also { SP can be characterized by ₀, S _k.Finally, in order to prevent SP _kbeat between different alliances, that is, a SP in an alliance, can only be had _k, have v ({ SP _k1, SP _k2, S _k)=0, to all k1, k2=1,2 ..., K.

Notice, at defined above-mentioned (N (t), v (S _k)) in Game with Coalitions model, due to alliance's cost and the existence to SP constraint, its Major Leagues, namely to consist of the situation of an alliance non-existent often for all users and SP, in other words, this model does not have superadditivity, therefore, this model belongs to coalition formation game (coalition formation game), namely often in all participants, final rock-steady structure is the little alliance that some are separated, and namely coalition structure H (also can be called alliance's segmentation, i.e. H={S ₁..., S _r, r is total number of the alliance that user forms, and comprises alliance and each unmarried alliance of access network).

Next, introduce enjoyment to arrange.In the present embodiment, the binary preference relation be defined as follows>= _n:

$S_{\mathrm{\α}}{\≥}_n{S}_{\mathrm{\β}}\⇔f_n\left(S_a\right)\≥f_n\left(S_{\mathrm{\β}}\right)$

Wherein two alliances comprising user n, i.e. n ∈ S _αand n ∈ S _β.F _nbe the user preference function defined any one user n, it characterizes user to the preference numerical value being subordinated to certain alliance.Above-mentioned relation refers to, user n is more ready to become S _αmember, or at least, it is to S _αwith S _βhobby be identical.

In the present embodiment, to an arbitrary user n and belonging alliance S thereof _kmake n ∈ S _k, the increment preference function f be defined as follows _n:

f _n(S _k)＝|S _k|·φn(S _k)-(|S _k|-1)·φ _m(Sk\{n})

Wherein user m ∈ S _k{ n}, φ _n(S _k), φ _m(S _kn}) be the individual financial value obtained for each user in alliance.

So far, above-mentioned set up network selects coalition formation game (N (t), v (S _k)) be reconfigured as the game of an enjoyment coalition formation (N (t),>= _n).

Next, based on above-mentioned enjoyment coalition formation game definition enjoyment rule of conversion, distributed coalition formation is carried out: a given current user coalitions structure or segmentation H={S ₁..., S _r, any one user n determine leave it current belonging to alliance S _h(n)=S _k, k ∈ 1 ..., r} also adds another alliance and if only if S _g∪ n}>= _ns _h(n).Like this, { S _k, S _g} → { S _k{ n}, S _g∪ n}}, and H → H', H' represent a new coalition structure enjoyment ex-situ operations often occurring and produces.Note,>= _nrepresent strict preference relation, look back preference relation defined above and preference function, namely

(|S _g|+1)·φ _n(S _g∪{n})-|S _g|·φm(S _g)＞

|S _k|·φ _n(S _k)-(|S _k|-1)·φ _m(S _k\{n})

According to the description of step 5, each user n replaces according to this transposition criterion successively iteration, until do not have user again to replace, namely coalition structure converges on final structure H _f, and the H that the present embodiment obtains _fthat Nash is stable.Notice, the rule of conversion that the present embodiment provides can ensure that each independent ex-situ operations can produce the new coalition structure that a past do not accessed, and can ensure iteration convergence.

Obvious, H _fgive the result that network is selected.Wherein, unmarried alliance S _k(k>K) i.e. { SP ₀, S _kfor temporarily to be gathered by the user that heterogeneous system blocks, that is non-access user; Alliance S _k(1≤k≤K) i.e. { SP _k, S _kin user, then select network k to access as objective network.Fig. 4 gives a kind of 35 random distribution users the stable H that the final distributed ad-hoc under the heterogeneous network system given by the present embodiment becomes _f, i.e. network selection result.Wherein circular, triangle and the request of square representative of consumer respectively be voice, data and video traffic, hollow is denoted as the user of access No. 1 WiMAX network, the solid user being denoted as access No. 2 WCDMA networks, cross is denoted as the user of access No. 3 wlan networks, and other are the user of access No. 4 WCDMA networks.Non-access user is there is not in this example.And H _funder the total utility value of 4 alliances be respectively v (S ₁)=21.3238, v (S ₂)=9.2979, v (S ₃)=18.7035, and v (S ₄)=6.8542, therefore, system total utility value is 56.1794, and the optimal solution now equaled under these 35 network user's distribution scenario (although the acquisition of optimal solution is np complete problem, adopts a kind of centralization searching method to obtain optimal solution, but this optimal solution only obtains when the total number of users of alliance is less, when user is too much, due to Space Explosion, search optimal solution will be very difficult, time complexity is very high, until may obtain hardly).In order to verify that method that the present invention takes can make system total utility and level off to the maximum of system total utility, simultaneously than W.Shenand Q.A.Zeng, " Cost-function-based network selection strategy in integratedwireless and mobile networks, " IEEE Transactions on Vehicular Technology, vol.57, no.6, pp.3778-3788, November2008. (syncretization is wireless with mobile network under based on the network selection policies of cost function, IEEE Vehicle Technology transactions, 57 volumes the 6th phase, in November, 2008) in introduce single RSNS system of selection better, the present invention carries out 10000 Monte Carlo Experiments, namely customer location is dispensed at random, Random assignment customer service type of call, by adding users quantity gradually, total utility value (only providing within 40 at the number of users) optimal solution that records centerization search provides, the relation of the total utility value that the present invention provides and the total utility value three that RSNS method provides, Fig. 5 gives the comparison of three, can see, the method that the present invention provides is far superior to RSNS method, simultaneously and level off to optimal solution.

According to step 6, above-mentioned coalition formation process is upgraded with cycle Г, namely each user according to current coalition structure (namely going up the coalition structure of cycle convergence), iteration carries out above-mentioned enjoyment transpose process successively again, until converge on new Stable coalitions structure.The object of this step is the complexity change of mobilism adaptive wireless environment, and such as, node motion is to new position, and after re-starting enjoyment transposition, possible node can leave current alliance, enters the new alliance of more preference.Game, each time ex-situ operations are selected for heterogeneous network proposed by the invention, is all equivalent to primary network handoff procedure.In order to verify the method proposed by the invention adaptivity to dynamic environment, the network selection result provided with Fig. 4 is designated as moment t ₀, at [t ₀, t ₀+ Δ Г], in the time interval of wherein Δ ∈ [1,25], consider that user 34 is vehicle-mounted user, from t ₀(0.6, the 0.2) uniform motion in moment is to (-0.6,0.2) position, and but change his type of service twice in this process, meanwhile, a pedestrian user 15 is from (-0.11,-0.05) uniform motion is to (-0.11,0.07), further, when Δ=6, user 5 terminated call, no longer be considered as any active ues, when Δ=16, a user 36 initiates audio access request to system.Fig. 6 gives the [t under above-mentioned complicated dynamic case ₀, t ₀+ Δ Г] in how to carry out mobilism network reselection based on each associated user of method proposed by the invention and select and switch with the above-mentioned change of self adaptation, Fig. 7 gives Δ=25, namely above-mentioned change terminate after network selection result, network selection result before it and Fig. 4 and dynamic change is contrasted, can see, for most of static subscriber, their objective network does not change during this period, thus prove that method proposed by the invention can avoid network frequently to switch, save unnecessary handover overhead.

Finally, based on the Monte Carlo Experiment scene of Fig. 5, Fig. 8 gives the change of method proposed by the invention to the maximum accommodation number of users of heterogeneous system (total number of users that namely can access) and comparable situation.Can see, method proposed by the invention can improve accommodation number of users compared to previously described RSNS method.Especially, after Virtual network operator upgrades their network, such as, in Fig. 8 SP ₁the WiMAX network of former table 1 upgraded to based on 802.16m, maximum bandwidth is B ₁after the network of=30Mbps, the maximum accommodation number of users of system can obtain significantly, further promotes.

Claims (6)

1. towards a network selecting method for heterogeneous wireless network environment, it is characterized in that, comprise following steps:

Step 1) t accessed all users of consolidated network or be not labeled as alliance S by the unique user of network insertion _k, set up each alliance S according to the type of service that user asks _kalliance total utility function v (S _k);

Step 2) set up Game with Coalitions model (N (t), the v (S of transferable utility _k)), the user that N (t) is t request service gathers, according to fairness allocation criteria by alliance total utility function v (S _k) distribute to alliance S _kinterior user, obtains the individual revenue function φ of each user _n(S _k);

Step 3) according to user n ∈ N (t) individual revenue function φ _n(S _k) designing user n is to alliance S _kpreference function f _n(S _k), according to preference function f _n(S _k) for user n is to the alliance S of any two its subordinates _α, S _βset up complete, anti-body, a transferable binary preference relation S _α>= _ns _β, and then by Game with Coalitions model (N (t), v (S _k)) be reconstructed into enjoyment Game with Coalitions model (N (t),>= _n);

Step 4) foundation enjoyment Game with Coalitions model (N (t),>= _n), user n is shifted to the alliance of a preference more from the alliance of current affiliation, and iteration transposition is until user n can not be shifted to new alliance more successively;

Step 5) user n selects in step 4) network corresponding to final alliance after iteration transposition be as target access network.

2. the network selecting method towards heterogeneous wireless network environment according to claim 1, it is characterized in that, also comprise step 6) periodically repeat step 4) with step 5) alliance's transpose process, with the dynamic change of self adaptation heterogeneous wireless environment.

3. the network selecting method towards heterogeneous wireless network environment according to claim 1, is characterized in that, described alliance total utility function v (S _k) the characterizing consumer service revenue of serving to objective network requested service and service cost.

4. towards a network selecting method for heterogeneous wireless network environment, it is characterized in that, comprise following steps:

T a) is accessed all users of consolidated network or is not labeled as alliance S by the unique user of network insertion by step _k, the type of service of asking according to user sets up the individual utility function v of each user _n(S _k);

Step b) set up Game with Coalitions model (N (t), the V (S of non-transferable utility _k)), the user that N (t) is t request service gathers, V (S _k) be by alliance S _kthe individual utility function v of interior each user _n(S _k) vector that forms;

Step c) according to user n ∈ N (t) individual utility function v _n(S _k) designing user n is to alliance S _kpreference function f _n(S _k), according to preference function f _n(S _k) for user n is to the alliance S of any two its subordinates _α, S _βset up one complete, anti-body with transferable binary preference relation S _α>=S _βbe designated as>= _n, and then by Game with Coalitions model (N (t), v (S _k)) be reconstructed into enjoyment Game with Coalitions model (N (t),>= _n);

Steps d) foundation enjoyment Game with Coalitions model (N (t),>= _n), user n changes to the alliance of a preference more from the alliance of current affiliation, and iteration transposition is until user n can not be shifted to new alliance more successively;

Step e) user n selects in steps d) network corresponding to final alliance after iteration transposition be as target access network.

5. the network selecting method towards heterogeneous wireless network environment according to claim 4, it is characterized in that, also comprise step f) periodically repeat steps d) with step e) alliance's transpose process, with the dynamic change of self adaptation heterogeneous wireless environment.

6. the network selecting method towards heterogeneous wireless network environment according to claim 4, is characterized in that, described individual utility function v _n(S _k) the characterizing consumer service revenue of serving to objective network requested service and service cost.