CN103826234A - TA re-planning method and system - Google Patents

Abstract

The invention discloses a TA re-planning method and system. The method comprises the steps that current user parameters of cells and user flowing parameters between the cells are obtained; a TA re-planning model is established through the obtained current user parameters of the cells and user flowing parameters between the cells, wherein the target function of the TA re-planning model comprises an updating expenditure minimum function, a paging signaling expenditure minimum function and a re-planning cost minimum function; based on original TA planning information of the cells, the TA re-planning model is resolved, and TA re-planning information of the cells is obtained; and according to the obtained TA re-planning information of the cells, TA re-planning is carried out on the cells. According to the method and system, position zone updating cost, paging cost and reconstruction TA planning cost are ranked as three targets, the mutual contradiction and the restriction of the three targets can be well reflected, a pareto solution set which can well reflect a network environment can be obtained, and the TA re-planning method and system are suitable for a large network.

Description

TA weight-normality is drawn method and system

Technical field

The present invention relates to communication technical field, particularly relate to a kind of TA weight-normality and draw method and system.

Background technology

In mobile communication system, because the band of position relates to user's ambulant management, so terminal equipment has all designed the band of position, as the lane place in GSM/UMTS (Location Area) and Route Area (Routing Area).For core net can only not need whole ICQ at limited range call terminal, for the mobile terminal of idle condition, core net need to be known terminal position roughly.4G epoch mobile wireless technology evolution standard TD-LTE, has applied similar band of position concept, and this band of position is called tracking area (Tracking Area, TA).Similarly, EPC, to the user in idle condition and connection status, will manage the TA of its registration, and user also can change the TA log-on message in EPC in the time there is TA change, and TA each other can not be overlapping.Too little when TA planning, the frequency that TA upgrades improves greatly, has also just improved the load of network signal process.Meanwhile, TA also can not plan too greatly, can expand like this paging domain of UE, and paging domain is the Radio Resource of conference waste system too.Therefore, TA planning is according to network actual state, obtains optimal balance between the two.

In actual applications, the planning of TA arranges in the time of network initial deployment, seeks the compromise of network renewal signaling consumption and radio call signaling consumption.And network environment is real-time change, passing in time, customer location and mobility model all can change.The TA of initial plan by As time goes on gradually no longer adapting to current customer location and mobility model, therefore, is necessary in use, the TA having disposed is carried out to weight-normality and draw.Current heavy planning technology is combined location area updating and ICQ cost as a cost target function using identical weights conventionally.

But above-mentioned heavy planning technology carries out weight-normality while drawing for large scale network, be difficult to the optimal solution set that provides complete, thereby cause signaling consumption and upgrade expense excessive, be not suitable for large scale network.

Summary of the invention

Based on this, be necessary to be difficult to the optimal solution set that provides complete, the problem that is not suitable for large scale network for the heavy planning technology of above-mentioned TA, provide a kind of TA weight-normality to draw method and system.

The heavy planing method of a kind of TA, comprises the following steps:

Obtain active user's parameter of each community and user's flow parameter of each minizone;

Build the heavy plan model of TA by obtaining active user's parameter of Ge community and user's flow parameter of each minizone, wherein, the target function of the heavy plan model of described TA comprises that upgrading expense minimum function, call signaling expense minimum function and weight-normality draws Least-cost function;

Based on the former TA planning information of each community, the heavy plan model of described TA is solved, obtain the heavy planning information of TA of each community;

Carrying out TA weight-normality according to the heavy planning information Dui Ge of the TA community that obtains Ge community draws.

The heavy planning system of a kind of TA, comprising:

Acquisition module, for obtaining active user's parameter of each community and user's flow parameter of each minizone;

MBM, for building the heavy plan model of TA by obtaining active user's parameter of Ge community and user's flow parameter of each minizone, wherein, the target function of the heavy plan model of described TA comprises that upgrading expense minimum function, call signaling expense minimum function and weight-normality draws Least-cost function;

Parsing module, for the former TA planning information based on each community, solves the heavy plan model of described TA, obtains the heavy planning information of TA of each community;

Planning module, draws for carrying out TA weight-normality according to the heavy planning information Dui Ge of the TA community that obtains Ge community.

Above-mentioned TA weight-normality is drawn method and system, build the heavy plan model of TA by obtaining active user's parameter of Ge community and user's flow parameter of each minizone, the target function of the heavy plan model of described TA comprises that upgrading expense minimum function, call signaling expense minimum function and weight-normality draws Least-cost function, based on the former TA planning information of each community, the heavy plan model of described TA is solved, and carry out TA weight-normality according to the heavy planning information Dui Ge of the TA community that obtains Ge community and draw.Location area updating cost, paging cost, reconstruct TA planning cost three are listed as three targets by the heavy plan model of described TA, more can embody the conflicting and restriction between three, the pareto disaggregation that can obtain more reflecting network environment, can better be applicable to large scale network.

Accompanying drawing explanation

Fig. 1 is the schematic flow sheet of heavy planing method the first execution mode of TA of the present invention;

Fig. 2 is the schematic flow sheet of heavy planing method the second execution mode of TA of the present invention;

Fig. 3 is the schematic flow sheet of the heavy planing method of TA of the present invention the 3rd execution mode;

Fig. 4 is the structural representation of heavy planning system the first execution mode of TA of the present invention.

Embodiment

Refer to Fig. 1, Fig. 1 is the schematic flow sheet of heavy planing method the first execution mode of TA of the present invention.

The heavy planing method of described TA of present embodiment comprises the following steps:

Step 101, obtains active user's parameter of each community and user's flow parameter of each minizone.

Step 102, build the heavy plan model of TA by obtaining active user's parameter of Ge community and user's flow parameter of each minizone, wherein, the target function of the heavy plan model of described TA comprises that upgrading expense minimum function, call signaling expense minimum function and weight-normality draws Least-cost function.

Step 103, based on the former TA planning information of each community, solves the heavy plan model of described TA, obtains the heavy planning information of TA of each community.

Step 104, carries out TA weight-normality according to the heavy planning information Dui Ge of the TA community that obtains Ge community and draws.

The heavy planing method of TA described in present embodiment, build the heavy plan model of TA by obtaining active user's parameter of Ge community and user's flow parameter of each minizone, the target function of the heavy plan model of described TA comprises that upgrading expense minimum function, call signaling expense minimum function and weight-normality draws Least-cost function, based on the former TA planning information of each community, the heavy plan model of described TA is solved, and carry out TA weight-normality according to the heavy planning information Dui Ge of the TA community that obtains Ge community and draw.Location area updating cost, paging cost, reconstruct TA planning cost three are listed as three targets by the heavy plan model of described TA, more can embody the conflicting and restriction between three, the pareto disaggregation that can obtain more reflecting network environment, can better be applicable to large scale network.

Wherein, for step 101, user's flow parameter of described each minizone preferably can be user's flow parameter of the each minizone in preset period of time, described preset period of time is the predefined timing statistics that obtains each inter-cell user amount of flow, can be set according to the actual conditions in planning region by those skilled in the art, can be one day that obtains before time of active user's parameter, also can be the week before the time of obtaining active user's parameter.

Preferably, described community is user concentrated area, in 3G system, no longer include the differentiation of circuit region and packet zone, user's position management method is not taked the register method of lane place (LA) and Route Area (RA) yet, but unified " tracking area " concept (TA) that adopts, by carrying out location management to the corresponding mark of user assignment, a tracking area is made up of some base stations or some communities conventionally.

Preferably, active user's parameter of described each community is preferably active user's number of each community, and described user's flow parameter preferably can comprise user's number that flows, can real-time statistics described in preset period of time, i community is to the mobile number of users in j community.Can also obtain described user's flow parameter by the usual other technologies means of those skilled in the art.

In one embodiment, the step of obtaining active user's parameter of each community and user's flow parameter of each minizone described in is further comprising the steps of:

Whether the variable quantity that judges active user's parameter of each community and user's flow parameter of each minizone exceedes threshold value.

If so, build the heavy plan model of TA by obtaining active user's parameter of Ge community and user's flow parameter of each minizone.

If not, stopping TA weight-normality draws.

Wherein, described threshold value is the maximum magnitude of the maximum magnitude of number of users flow parameter variable quantity or the difference of active user's parameter and original subscriber's ginseng.

In another embodiment, before described preset period of time, those skilled in the art can adopt the plan model Dui Ge community that comprises following target function to carry out initial TA planning:

$\underset{v}{\min }(f_1\left(v\right),f_2\left(v\right)).;$

$f_1\left(v\right)=\underset{i\∈C}{\mathrm{\Σ}}\underset{j\∈C:j\≠i}{\mathrm{\Σ}}{c}^u{w}_{i,j}(1-m_{i,j}\left(v\right)).;$

$f_2\left(v\right)=\underset{i\∈C}{\mathrm{\Σ}}\underset{j\∈C:j\≠i}{\mathrm{\Σ}}\mathrm{\α}{c}^p{u}_i{m}_{i,j}\left(v\right).;$

Wherein, u _ifor the active user's number in cell i, w _i,jfor cell i is to community j user's flow parameter, can be added up by active user's switching.C ^pbe the cost of a paging operation, c ^ufor once upgrading the cost of operation, α is the call density factor.

Wherein, by above-mentioned plan model, initial plan is carried out in Ke Duige community, obtains before described preset period of time the former TA planning information of each community.

For step 102, while building the heavy plan model of TA, can in real time the corresponding relation obtaining between active user's parameter of Ge community and user's flow parameter of each minizone, the paging parameters that sets in advance or store, undated parameter, each community and unknown TA be built to the heavy plan model of described TA.Also can replace in advance user's flow parameter of active user's parameter He Ge community of Ge to be obtained community with a label, according to arranging or corresponding relation structure preset model storage, between paging parameters, undated parameter, each community and unknown TA, then will obtain preset model described in active user's parameter of Ge community and user's flow parameter substitution of each minizone in this step, construct the heavy plan model of told TA.

It is in one embodiment, described that by obtaining active user's parameter of Ge community and user's flow parameter of each minizone, to build the step of the heavy plan model of TA further comprising the steps of:

According to the real network environment of each community, cost and paging density factor that a paging operation is set are respectively that paging parameters, the cost of once upgrading operation are the corresponding relation between undated parameter and each community and unknown TA.

With the corresponding relation between the paging parameters, undated parameter, each community and the unknown TA that arrange with obtain active user's parameter of Ge community and user's flow parameter of each minizone, establishing target function comprises as described in minor function TA weight plan model:

$\underset{v}{\min }{f}_1\left(v\right)=\underset{i\∈C}{\mathrm{\Σ}}\underset{j\∈C:j\≠i}{\mathrm{\Σ}}{c}^u{w}_{i,j}(1-m_{i,j}\left(v\right)),$

$\underset{v}{\min }{f}_2\left(v\right)=\underset{i\∈C}{\mathrm{\Σ}}\underset{j\∈C:j\≠i}{\mathrm{\Σ}}\mathrm{\α}{c}^p{u}_i{m}_{i,j}\left(v\right),$ $\underset{v}{\min }{f}_3\left(v\right)=\underset{i\∈C}{\mathrm{\Σ}}{u}_i{g}_i(v,v^0).;$

Wherein, c ^ube the cost of a paging operation, α is the call density factor, c ^pfor once upgrading the cost of operation, w _i,jfor user's flow parameter small area i is to user's flow parameter of community j, u _ifor active user's parameter of cell i, f ₁(v) for upgrading expense, f ₂(v) be call signaling expense, f ₃(v) the planning cost of attaching most importance to, v _ifor the Digital ID of the current affiliated tracking area in i community, v _jfor the Digital ID of the current affiliated tracking area in j community, for the Digital ID of the former affiliated tracking area in i community, m _i,j(v) whether belong to same community, g for identifying two different districts _i(v, v ⁰) whether change for tracking area under identifying same community.

In the present embodiment, cell ID can be set, making set of cells is C={1,2,3 ..., N}, arranges TA mark, makes TAs set for T _a=1,2,3 ..., T}, note vector v={ v ₁, v ₂, v ₃..., v _nbe the TA distribution of community, wherein, v _irepresent the mark of the TA under cell i, also available Binary Zero, 1 matrix M (v) expression vector v.

It is in another embodiment, described that by obtaining active user's parameter of Ge community and user's flow parameter of each minizone, to build the step of the heavy plan model of TA further comprising the steps of:

Comprise that by obtaining active user's parameter of Ge community and user's flow parameter substitution of each minizone the preset model of following target function constructs the heavy plan model of described TA:

$\underset{v}{\min }{f}_1\left(v\right)=\underset{i\∈C}{\mathrm{\Σ}}\underset{j\∈C:j\≠i}{\mathrm{\Σ}}{c}^u{w}_{i,j}(1-m_{i,j}\left(v\right)),$

$\underset{v}{\min }{f}_2\left(v\right)=\underset{i\∈C}{\mathrm{\Σ}}\underset{j\∈C:j\≠i}{\mathrm{\Σ}}\mathrm{\α}{c}^p{u}_i{m}_{i,j}\left(v\right),$

$\underset{v}{\min }{f}_3\left(v\right)=\underset{i\∈C}{\mathrm{\Σ}}{u}_i{g}_i(v,v^0).;$

Wherein, c ^ube the cost of a paging operation, α is the call density factor, c ^pfor once upgrading the cost of operation, w _i,jfor user's flow parameter small area i is to user's flow parameter of community j, u _ifor active user's number of cell i, f ₁(v) for upgrading expense, f ₂(v) be call signaling expense, f ₃(v) the planning cost of attaching most importance to, v _ifor the Digital ID of the current affiliated tracking area in i community, v _jfor the Digital ID of the current affiliated tracking area in j community,

for the Digital ID of the former affiliated tracking area in i community, m _i,j(v) whether belong to same community, g for identifying two different districts _i(v, v ⁰) whether change for tracking area under identifying same community.

For step 103, can solve described returning of TA model by the usual various optimized algorithms of those skilled in the art.

For step 104, by optimized algorithm, the heavy plan model of described TA is solved, can obtain a disaggregation, comprise multiple heavy programme.

In one embodiment, in the time that to obtain returning of the TA information of each community be multiple, it is further comprising the steps of that the step that TA weight-normality draws is carried out in the heavy planning information Dui Ge of the TA community that described basis is obtained Ge community:

From the heavy planning information of multiple TA, choose the heavy planning information of TA that a kind of default compromise that meets three target functions requires.

According to the heavy planning information of the TA selecting, Dui Ge community is carried out TA weight-normality and is drawn.

In the present embodiment, can draw cost threshold value to upgrading cost threshold value, paging cost threshold value and weight-normality according to predefined, choose final heavy planning information.

Refer to Fig. 2, Fig. 2 is the schematic flow sheet of heavy planing method the second execution mode of TA of the present invention.

The heavy planing method of TA of present embodiment and the difference of the first execution mode are: the described former TA planning information based on each community, the heavy plan model of described TA is solved, and the step of obtaining the heavy planning information of TA of each community comprises the following steps:

Step 201, estimates respectively maximum and the minimum value of three target functions of the heavy plan model of described TA.

Step 202, according to maximum and the minimum value of three target functions of the heavy plan model of described TA estimating, three target functions to the heavy plan model of described TA carry out normalizing operation, construct and comprise that three goal constraint optimizations of following three target functions draw model:

$\underset{v}{\min }{h}_1\left(v\right)=\frac{f_1\left(v\right)-\underset{\‾}{f_1}}{\stackrel{\‾}{f_1}-\underset{\‾}{f_1}},$

$\underset{v}{\min }{h}_2\left(v\right)=\frac{f_2\left(v\right)-\underset{\‾}{f_2}}{\stackrel{\‾}{f_2}-\underset{\‾}{f_2}},$

$\underset{v}{\min }{h}_3\left(v\right)=\frac{f_3\left(v\right)-\underset{\‾}{f_3}}{\stackrel{\‾}{f_3}-\underset{\‾}{f_3}}.;$

Wherein, f ₁ , f ₂ , f ₃ and be respectively maximum and the minimum value of three target functions of the heavy plan model of described TA, h ₁(v), h ₂and h (v) ₃(v) be three target functions.

Step 203, based on the former TA planning information of each community, draws model to described three goal constraint optimizations and solves, and obtains the heavy planning information of TA of each community.

The heavy planing method of TA of present embodiment, can avoid excessive causing of single target function too to be emphasized, makes three target functions of reaction that analytic solutions are more balanced.

In one embodiment, the described former TA planning information based on each community, draws model to described three goal constraint optimizations and solves, and the step of the heavy planning information of TA of obtaining each community is further comprising the steps of:

By the evolution algorithm of min max principle and spherical coordinate transformation, described three goal constraint optimizations are drawn to model and solve, obtain the heavy planning information of TA of each community.

Wherein, solve described TA with the evolution algorithm of min max principle and spherical coordinate transformation and weigh plan model, before solving, overemphasized for avoiding each target function to change excessive certain target that causes, can before solving, introduce standardized technique, three target functions are carried out respectively to standardization.The span of each target function after conversion is [0,1], and is positioned at the first quartile of coordinate system.

Refer to Fig. 3, Fig. 3 is the schematic flow sheet of the heavy planing method of TA of the present invention the 3rd execution mode.

The heavy planing method of TA of present embodiment and the difference of the first execution mode and the second execution mode are: described by the evolution algorithm of min max principle and spherical coordinate transformation, described three goal constraint optimizations are drawn to model and solve, the step of obtaining the heavy planning information of TA of each community comprises the following steps:

Step 301, three target functions described three goal constraint optimizations being drawn to model by following formula carry out spherical coordinate transformation:

$\left\{\begin{array}{c}{h}_1=r\cos {\mathrm{\θ}}_1,\\ h_2=r\sin {\mathrm{\θ}}_1\cos {\mathrm{\θ}}_2,\\ h_3=r\sin {\mathrm{\θ}}_1\sin {\mathrm{\θ}}_2.\end{array}\right.$

Wherein, h ₁, h ₂and h ₃be respectively three target functions, the radius that r is spherical coordinate, θ ₁and θ ₂for quadrant angle.

Step 302, arranges numerical value K, draws K as follows from zero point of reference frame to first quartile ²the equally distributed ray of bar:

$\left\{\begin{array}{c}{\mathrm{\θ}}_{1,l_1}=\frac{l_1\mathrm{\π}}{2(k+1)},l_1=\mathrm{1,2},...,k,\\ {\mathrm{\θ}}_{2,l_2}=\frac{l_2\mathrm{\π}}{2(k+1)},l_1=\mathrm{1,2},...,k.\end{array}\right.;$

Step 303, obtains weight coefficient as follows according to described ray:

$\left\{\begin{array}{c}{q}_{1,j}=\frac{1}{\cos {\mathrm{\θ}}_{1,j}}\sqrt[3]{\underset{i=1}{\overset{2}{\mathrm{\Π}}}{\sin }^2{\mathrm{\θ}}_{i,j}\cos {\mathrm{\θ}}_{i,j}},\\ q_{2,j}=\frac{\cos {\mathrm{\θ}}_{1,j}}{\sin {\mathrm{\θ}}_{1,j}\cos {\mathrm{\θ}}_{2,j}}{w}_{1,j},\\ q_{3,j}=\frac{\cos {\mathrm{\θ}}_{2,j}}{\sin {\mathrm{\θ}}_{2,j}}{w}_{2,j}.\end{array}\right.;$

Step 304, according to described weight coefficient and described ray, arranges K ²individual fitness function as follows:

$F_j\left(v\right)=\underset{1\≤i\≤3}{\max }\left\{w_{i,j}{h}_i\left(v\right)\right\},j=\mathrm{1,2},...,k^2.;$

Wherein, F _j(v) be fitness function.

The heavy planing method of TA described in present embodiment, based on the evolution algorithm of min max principle and spherical coordinate transformation, can obtain approximate uniform pareto disaggregation, for real network TA weight-normality is provided by the comprehensive information that provides, be applicable to the large-scale network planning, improved network whole efficiency.

In one embodiment, based on above-mentioned given fitness function, TA is heavy, and planning process is specific as follows:

Step 401, according to the actual conditions of network, setting is once upgraded the cost parameter of operation and the cost parameter of a paging operation and is: c ^uand c ^p, switch statistical estimate by each community active user and go out m _i,j, the average called density factor α of given community user.

Step 402, by m _i,jconcrete numerical value substitution preset model, construct described returning of TA model, described returning of TA model is carried out to standardization, can obtain described three goal constraint optimizations and draw model.

Step 403, the Population Size that the evolution algorithm of min max principle and spherical coordinate transformation is set is that S, fitness function number k, crossover probability and variation probability are respectively p _cand p _m, maximum genetic algebra is that Q, initial TAs population number are that B (0), first initial algebra are r=0.

Step 404, calculates respectively θ _i,jand q _i,j, by the q calculating _i,j, define fitness function, calculate the fitness function value of all TAs population at individuals.

Step 405, utilizes each fitness function, chooses best l individuality, each individuality of choosing is chosen at random to one from current population and match, and utilizes single-point to intersect and produces l to new individuality.

Step 406, the random number that produces S value [0,1] scope is designated as z ₁, z ₂..., z _s.If z _i≤ p _m, i individuality makes a variation.

Step 407, utilizes fitness function, is each fitness letter

number is chosen best individuality, forms corresponding set, and all set are merged together, and forms TA planning population B (t+1) in the next generation.

Step 408, if r=Q, all noninferior solutions are elected optimum pareto optimal solution as, algorithm is ended, otherwise, make r=r+1, again perform step 406 to step 408, again calculate respectively θ _i,jand q _i,j, by the q calculating _i,j, define fitness function, calculate the fitness function value of all TAs population at individuals.

Step 409 according to network actual state and relevant demand information, is chosen a solution to the good compromise of three target functions from optimum pareto front interface, and the scheme of drawing as TAs weight-normality arranges.

Refer to Fig. 4, Fig. 4 is the structural representation of heavy planning system the first execution mode of TA of the present invention.

The described TA weight-normality of present embodiment is drawn system acquisition module 100, MBM 200, parsing module 300 and planning module 400, wherein:

Acquisition module 100, for obtaining active user's parameter of each community and user's flow parameter of each minizone.

MBM 200, for building the heavy plan model of TA by obtaining active user's parameter of Ge community and user's flow parameter of each minizone, wherein, the target function of the heavy plan model of described TA comprises that upgrading expense minimum function, call signaling expense minimum function and weight-normality draws Least-cost function.

Parsing module 300, for the former TA planning information based on each community, solves the heavy plan model of described TA, obtains the heavy planning information of TA of each community.

Planning module 400, draws for carrying out TA weight-normality according to the heavy planning information Dui Ge of the TA community that obtains Ge community.

The heavy planning system of TA described in present embodiment, build the heavy plan model of TA by obtaining active user's parameter of Ge community and user's flow parameter of each minizone, the target function of the heavy plan model of described TA comprises that upgrading expense minimum function, call signaling expense minimum function and weight-normality draws Least-cost function, based on the former TA planning information of each community, the heavy plan model of described TA is solved, and carry out TA weight-normality according to the heavy planning information Dui Ge of the TA community that obtains Ge community and draw.Location area updating cost, paging cost, reconstruct TA planning cost three are listed as three targets by the heavy plan model of described TA, more can embody the conflicting and restriction between three, the pareto disaggregation that can obtain more reflecting network environment, can better be applicable to large scale network.

Wherein, for acquisition module 100, user's flow parameter of described each minizone preferably can be user's flow parameter of the each minizone in preset period of time, described preset period of time is the predefined timing statistics that obtains each inter-cell user amount of flow, can be set according to the actual conditions in planning region by those skilled in the art, can be one day that obtains before time of active user's parameter, also can be the week before the time of obtaining active user's parameter.

Preferably, described community is user concentrated area, in 3G system, the position management method that no longer includes the differentiation user of circuit region and packet zone is not taked the register method of lane place (LA) and Route Area (RA) yet, but unified " tracking area " concept (TA) that adopts, by carrying out location management to the corresponding mark of user assignment, a tracking area is made up of some base stations or some communities conventionally.

Preferably, active user's parameter of described each community is preferably active user's number of each community, and described user's flow parameter preferably can comprise user's number that flows, can real-time statistics described in preset period of time, i community is to the mobile number of users in j community.Can also obtain described user's flow parameter by the usual other technologies means of those skilled in the art.

In one embodiment, acquisition module 100 also can be used for:

Whether the converted quantity that judges active user's parameter of each community and user's flow parameter of each minizone exceedes threshold value.

If so, build the heavy plan model of TA by obtaining active user's parameter of Ge community and user's flow parameter of each minizone.

If not, stopping TA weight-normality draws.

Wherein, described threshold value is the maximum magnitude of the maximum magnitude of number of users flow parameter variable quantity or the difference of active user's parameter and original subscriber's ginseng.

In another embodiment, before described preset period of time, those skilled in the art can adopt the plan model Dui Ge community that comprises following target function to carry out initial TA planning:

$\underset{v}{\min }(f_1\left(v\right),f_2\left(v\right)).;$

$f_1\left(v\right)=\underset{i\∈C}{\mathrm{\Σ}}\underset{j\∈C:j\≠i}{\mathrm{\Σ}}{c}^u{w}_{i,j}(1-m_{i,j}\left(v\right)).;$

$f_2\left(v\right)=\underset{i\∈C}{\mathrm{\Σ}}\underset{j\∈C:j\≠i}{\mathrm{\Σ}}\mathrm{\α}{c}^p{u}_i{m}_{i,j}\left(v\right).;$

Wherein, u _ifor the active user's number in cell i, w _i,jfor cell i is to community j user's flow parameter, can be added up by active user's switching.C ^pbe the cost of a paging operation, c ^ufor once upgrading the cost of operation, α is the call density factor.

Wherein, by above-mentioned plan model, initial plan is carried out in Ke Duige community, obtains before described preset period of time the former TA planning information of each community.

For MBM 200, while building the heavy plan model of TA, can in real time the corresponding relation obtaining between active user's parameter of Ge community and user's flow parameter of each minizone, the paging parameters that sets in advance or store, undated parameter, each community and unknown TA be built to the heavy plan model of described TA.Also can replace in advance user's flow parameter of active user's parameter He Ge community of Ge to be obtained community with a label, according to arranging or corresponding relation structure preset model storage, between paging parameters, undated parameter, each community and unknown TA, then will obtain preset model described in active user's parameter of Ge community and user's flow parameter substitution of each minizone in this step, construct the heavy plan model of told TA.

In one embodiment, MBM 200 also can be used for:

According to the real network environment of each community, cost and paging density factor that a paging operation is set are respectively that paging parameters, the cost of once upgrading operation are the corresponding relation between undated parameter and each community and unknown TA.

With the corresponding relation between the paging parameters, undated parameter, each community and the unknown TA that arrange with obtain active user's parameter of Ge community and user's flow parameter of each minizone, establishing target function comprises as described in minor function TA weight plan model:

$\underset{v}{\min }{f}_1\left(v\right)=\underset{i\∈C}{\mathrm{\Σ}}\underset{j\∈C:j\≠i}{\mathrm{\Σ}}{c}^u{w}_{i,j}(1-m_{i,j}\left(v\right)),$

$\underset{v}{\min }{f}_2\left(v\right)=\underset{i\∈C}{\mathrm{\Σ}}\underset{j\∈C:j\≠i}{\mathrm{\Σ}}\mathrm{\α}{c}^p{u}_i{m}_{i,j}\left(v\right),$

$\underset{v}{\min }{f}_3\left(v\right)=\underset{i\∈C}{\mathrm{\Σ}}{u}_i{g}_i(v,v^0).;$

Wherein, c ^ube the cost of a paging operation, α is the call density factor, c ^pfor once upgrading the cost of operation, w _i,jfor user's flow parameter small area i is to user's flow parameter of community j, u _ifor active user's parameter of cell i, f ₁(v) for upgrading expense, f ₂(v) be call signaling expense, f ₃(v) the planning cost of attaching most importance to, v _ifor the Digital ID of the current affiliated tracking area in i community, v _jfor the Digital ID of the current affiliated tracking area in j community,

for the Digital ID of the former affiliated tracking area in i community, m _i,j(v) whether belong to same community, g for identifying two different districts _i(v, v ⁰) whether change for tracking area under identifying same community.

In the present embodiment, cell ID can be set, making set of cells is C={1,2,3 ..., N}, arranges TA mark, makes TAs set for T _a=1,2,3 ..., T}, note vector v={ v ₁, v ₂, v ₃..., v _nbe the TA distribution of community, wherein, v _irepresent the mark of the TA under cell i, also available Binary Zero, 1 matrix M (v) expression vector v.

In another embodiment, MBM 200 also can be further used for:

Comprise that by obtaining active user's parameter of Ge community and user's flow parameter substitution of each minizone the preset model of following target function constructs the heavy plan model of described TA:

$\underset{v}{\min }{f}_1\left(v\right)=\underset{i\∈C}{\mathrm{\Σ}}\underset{j\∈C:j\≠i}{\mathrm{\Σ}}{c}^u{w}_{i,j}(1-m_{i,j}\left(v\right)),$

$\underset{v}{\min }{f}_2\left(v\right)=\underset{i\∈C}{\mathrm{\Σ}}\underset{j\∈C:j\≠i}{\mathrm{\Σ}}\mathrm{\α}{c}^p{u}_i{m}_{i,j}\left(v\right),$

$\underset{v}{\min }{f}_3\left(v\right)=\underset{i\∈C}{\mathrm{\Σ}}{u}_i{g}_i(v,v^0).;$

Wherein, c ^ube the cost of a paging operation, α is the call density factor, c ^pfor once upgrading the cost of operation, w _i,jfor user's flow parameter small area i is to user's flow parameter of community j, u _ifor active user's number of cell i, f ₁(v) for upgrading expense, f ₂(v) be call signaling expense, f ₃(v) the planning cost of attaching most importance to, v _ifor the Digital ID of the current affiliated tracking area in i community, v _jfor the Digital ID of the current affiliated tracking area in j community,

for the Digital ID of the former affiliated tracking area in i community, m _i,j(v) whether belong to same community, g for identifying two different districts _i(v, v ⁰) whether change for tracking area under identifying same community.

For parsing module 300, can solve described returning of TA model by the usual various optimized algorithms of those skilled in the art.

For planning module 400, by optimized algorithm, the heavy plan model of described TA is solved, can obtain a disaggregation, comprise multiple heavy programme.

In one embodiment, in the time that to obtain returning of the TA information of each community be multiple, planning module 400 can be used for:

From the heavy planning information of multiple TA, choose the heavy planning information of TA that a kind of default compromise that meets three target functions requires.

According to the heavy planning information of the TA selecting, Dui Ge community is carried out TA weight-normality and is drawn.

In the present embodiment, can draw cost threshold value to upgrading cost threshold value, paging cost threshold value and weight-normality according to predefined, choose final heavy planning information.

The following stated is heavy planning system the second execution mode of TA of the present invention.

The heavy planning system of TA of present embodiment and the difference of the first execution mode are: parsing module 300 also can be used for:

Estimate respectively maximum and the minimum value of three target functions of the heavy plan model of described TA.

According to maximum and the minimum value of three target functions of the heavy plan model of described TA estimating, three target functions of the heavy plan model of described TA are carried out to normalizing operation, construct and comprise that three goal constraint optimizations of following three target functions draw model:

$\underset{v}{\min }{h}_1\left(v\right)=\frac{f_1\left(v\right)-\underset{\‾}{f_1}}{\stackrel{\‾}{f_1}-\underset{\‾}{f_1}},$

$\underset{v}{\min }{h}_2\left(v\right)=\frac{f_2\left(v\right)-\underset{\‾}{f_2}}{\stackrel{\‾}{f_2}-\underset{\‾}{f_2}},$

$\underset{v}{\min }{h}_3\left(v\right)=\frac{f_3\left(v\right)-\underset{\‾}{f_3}}{\stackrel{\‾}{f_3}-\underset{\‾}{f_3}}.;$

Wherein, f ₁ , f ₂ , f ₃ and

be respectively maximum and the minimum value of three target functions of the heavy plan model of described TA, h ₁(v), h ₂and h (v) ₃(v) be three target functions.

Based on the former TA planning information of each community, described three goal constraint optimizations are drawn to model and solve, obtain the heavy planning information of TA of each community.

The heavy planning system of TA of present embodiment, can avoid excessive causing of single target function too to be emphasized, makes three target functions of reaction that analytic solutions are more balanced.

In one embodiment, parsing module 300 also can be further used for:

By the evolution algorithm of min max principle and spherical coordinate transformation, described three goal constraint optimizations are drawn to model and solve, obtain the heavy planning information of TA of each community.

Wherein, solve described TA with the evolution algorithm of min max principle and spherical coordinate transformation and weigh plan model, before solving, overemphasized for avoiding each target function to change excessive certain target that causes, can before solving, introduce standardized technique, three target functions are carried out respectively to standardization.The span of each target function after conversion is [0,1], and is positioned at the first quartile of coordinate system.

The following stated is the heavy planning system of TA of the present invention the 3rd execution mode.

The heavy planning system of TA of present embodiment and the difference of the first execution mode and the second execution mode are: parsing module 300 also can be further used for:

Three target functions described three goal constraint optimizations being drawn to model by following formula carry out spherical coordinate transformation:

$\left\{\begin{array}{c}{h}_1=r\cos {\mathrm{\θ}}_1,\\ h_2=r\sin {\mathrm{\θ}}_1\cos {\mathrm{\θ}}_2,\\ h_3=r\sin {\mathrm{\θ}}_1\sin {\mathrm{\θ}}_2.\end{array}\right.$

Wherein, h ₁, h ₂and h ₃be respectively three target functions, the radius that r is spherical coordinate, θ ₁and θ ₂for quadrant angle.

Numerical value K is set, draws K as follows from zero point of reference frame to first quartile ²the equally distributed ray of bar:

$\left\{\begin{array}{c}{\mathrm{\θ}}_{1,l_1}=\frac{l_1\mathrm{\π}}{2(k+1)},l_1=\mathrm{1,2},...,k,\\ {\mathrm{\θ}}_{2,l_2}=\frac{l_2\mathrm{\π}}{2(k+1)},l_1=\mathrm{1,2},...,k.\end{array}\right.;$

Obtain weight coefficient as follows according to described ray:

$\left\{\begin{array}{c}{q}_{1,j}=\frac{1}{\cos {\mathrm{\θ}}_{1,j}}\sqrt[3]{\underset{i=1}{\overset{2}{\mathrm{\Π}}}{\sin }^2{\mathrm{\θ}}_{i,j}\cos {\mathrm{\θ}}_{i,j}},\\ q_{2,j}=\frac{\cos {\mathrm{\θ}}_{1,j}}{\sin {\mathrm{\θ}}_{1,j}\cos {\mathrm{\θ}}_{2,j}}{w}_{1,j},\\ q_{3,j}=\frac{\cos {\mathrm{\θ}}_{2,j}}{\sin {\mathrm{\θ}}_{2,j}}{w}_{2,j}.\end{array}\right.;$

According to described weight coefficient and described ray, K is set ²individual fitness function as follows:

$F_j\left(v\right)=\underset{1\≤i\≤3}{\max }\left\{w_{i,j}{h}_i\left(v\right)\right\},j=\mathrm{1,2},...,k^2.;$

Wherein, F _j(v) be fitness function.

The heavy planning system of TA described in present embodiment, based on the evolution algorithm of min max principle and spherical coordinate transformation, can obtain approximate uniform pareto disaggregation, for real network TA weight-normality is provided by the comprehensive information that provides, be applicable to the large-scale network planning, improved network whole efficiency.

The above embodiment has only expressed several execution mode of the present invention, and it describes comparatively concrete and detailed, but can not therefore be interpreted as the restriction to the scope of the claims of the present invention.It should be pointed out that for the person of ordinary skill of the art, without departing from the inventive concept of the premise, can also make some distortion and improvement, these all belong to protection scope of the present invention.Therefore, the protection range of patent of the present invention should be as the criterion with claims.

Claims (10)

1. the heavy planing method of TA, is characterized in that, comprises the following steps:

Obtain active user's parameter of each community and user's flow parameter of each minizone;

Build the heavy plan model of TA by obtaining active user's parameter of Ge community and user's flow parameter of each minizone, wherein, the target function of the heavy plan model of described TA comprises that upgrading expense minimum function, call signaling expense minimum function and weight-normality draws Least-cost function;

Based on the former TA planning information of each community, the heavy plan model of described TA is solved, obtain the heavy planning information of TA of each community;

Carrying out TA weight-normality according to the heavy planning information Dui Ge of the TA community that obtains Ge community draws.

2. the heavy planing method of TA according to claim 1, is characterized in that, described in to obtain the step of active user's parameter of each community and user's flow parameter of each minizone further comprising the steps of:

Whether the variable quantity that judges active user's parameter of each community and user's flow parameter of each minizone exceedes threshold value;

If so, build the heavy plan model of TA by obtaining active user's parameter of Ge community and user's flow parameter of each minizone;

If not, stopping TA weight-normality draws.

3. the heavy planing method of TA according to claim 1, is characterized in that, described by obtaining active user's parameter of Ge community and user's flow parameter of each minizone, to build the step of the heavy plan model of TA further comprising the steps of:

According to the real network environment of each community, cost and paging density factor that a paging operation is set are respectively that paging parameters, the cost of once upgrading operation are the corresponding relation between undated parameter and each community and unknown TA;

With the corresponding relation between the paging parameters, undated parameter, each community and the unknown TA that arrange with obtain active user's parameter of Ge community and user's flow parameter of each minizone, establishing target function comprises as described in minor function TA weight plan model:

$\underset{v}{\min }{f}_1\left(v\right)=\underset{i\∈C}{\mathrm{\Σ}}\underset{j\∈C:j\≠i}{\mathrm{\Σ}}{c}^u{w}_{i,j}(1-m_{i,j}\left(v\right)),$

$\underset{v}{\min }{f}_2\left(v\right)=\underset{i\∈C}{\mathrm{\Σ}}\underset{j\∈C:j\≠i}{\mathrm{\Σ}}\mathrm{\α}{c}^p{u}_i{m}_{i,j}\left(v\right),$

$\underset{v}{\min }{f}_3\left(v\right)=\underset{i\∈C}{\mathrm{\Σ}}{u}_i{g}_i(v,v^0);$

Wherein, c ^ube the cost of a paging operation, α is the call density factor, c ^pfor once upgrading the cost of operation, w _i,jfor user's flow parameter small area i is to user's flow parameter of community j, u _ifor active user's parameter of cell i, f ₁(v) for upgrading expense, f ₂(v) be call signaling expense, f ₃(v) the planning cost of attaching most importance to, v _ifor the Digital ID of the current affiliated tracking area in i community, v _jfor the Digital ID of the current affiliated tracking area in j community,

for the Digital ID of the former affiliated tracking area in i community, m _i,j(v) whether belong to same community, g for identifying two different districts _i(v, v ⁰) whether change for tracking area under identifying same community.

4. the heavy planing method of TA according to claim 1, is characterized in that, described by obtaining active user's parameter of Ge community and user's flow parameter of each minizone, to build the step of the heavy plan model of TA further comprising the steps of:

Comprise that by obtaining active user's parameter of Ge community and user's flow parameter substitution of each minizone the preset model of following target function constructs the heavy plan model of described TA:

$\underset{v}{\min }{f}_1\left(v\right)=\underset{i\∈C}{\mathrm{\Σ}}\underset{j\∈C:j\≠i}{\mathrm{\Σ}}{c}^u{w}_{i,j}(1-m_{i,j}\left(v\right)),$

$\underset{v}{\min }{f}_2\left(v\right)=\underset{i\∈C}{\mathrm{\Σ}}\underset{j\∈C:j\≠i}{\mathrm{\Σ}}\mathrm{\α}{c}^p{u}_i{m}_{i,j}\left(v\right),$

$\underset{v}{\min }{f}_3\left(v\right)=\underset{i\∈C}{\mathrm{\Σ}}{u}_i{g}_i(v,v^0);$

Wherein, c ^ube the cost of a paging operation, α is the call density factor, c ^pfor once upgrading the cost of operation, w _i,jfor user's flow parameter small area i is to user's flow parameter of community j, u _ifor active user's parameter of cell i, f ₁(v) for upgrading expense, f ₂(v) be call signaling expense, f ₃(v) the planning cost of attaching most importance to, v _ifor the Digital ID of the current affiliated tracking area in i community, v _jfor the Digital ID of the current affiliated tracking area in j community, for the Digital ID of the former affiliated tracking area in i community, m _i,j(v) whether belong to same community, g for identifying two different districts _i(v, v ⁰) whether change for tracking area under identifying same community.

5. according to the heavy planing method of the TA described in any one in claim 1 to 4, it is characterized in that, the described former TA planning information based on each community, solves the heavy plan model of described TA, and the step of obtaining the heavy planning information of TA of each community comprises the following steps:

Estimate respectively maximum and the minimum value of three target functions of the heavy plan model of described TA;

According to maximum and the minimum value of three target functions of the heavy plan model of described TA estimating, three target functions of the heavy plan model of described TA are carried out to normalizing operation, construct and comprise that three goal constraint optimizations of following three target functions draw model:

$\underset{v}{\min }{h}_1\left(v\right)=\frac{f_1\left(v\right)-\underset{\‾}{f_1}}{\stackrel{\‾}{f_1}-\underset{\‾}{f_1}},$

$\underset{v}{\min }{h}_2\left(v\right)=\frac{f_2\left(v\right)-\underset{\‾}{f_2}}{\stackrel{\‾}{f_2}-\underset{\‾}{f_2}},$

$\underset{v}{\min }{h}_3\left(v\right)=\frac{f_3\left(v\right)-\underset{\‾}{f_3}}{\stackrel{\‾}{f_3}-\underset{\‾}{f_3}};$

Wherein, f ₁ , f ₂ , f ₃ and

be respectively maximum and the minimum value of three target functions of the heavy plan model of described TA, h ₁(v), h ₂and h (v) ₃(v) be three target functions;

Based on the former TA planning information of each community, described three goal constraint optimizations are drawn to model and solve, obtain the heavy planning information of TA of each community.

6. the heavy planing method of TA according to claim 5, is characterized in that, the described former TA planning information based on each community, draws model to described three goal constraint optimizations and solve, and the step of the heavy planning information of TA of obtaining each community is further comprising the steps of:

By the evolution algorithm of min max principle and spherical coordinate transformation, described three goal constraint optimizations are drawn to model and solve, obtain the heavy planning information of TA of each community.

7. the heavy planing method of TA according to claim 6, it is characterized in that, the described evolution algorithm that passes through min max principle and spherical coordinate transformation, draws model to described three goal constraint optimizations and solves, and the step of obtaining the heavy planning information of TA of each community comprises the following steps:

Three target functions described three goal constraint optimizations being drawn to model by following formula carry out spherical coordinate transformation:

$\left\{\begin{array}{c}{h}_1=r\cos {\mathrm{\θ}}_1,\\ h_2=r\sin {\mathrm{\θ}}_1\cos {\mathrm{\θ}}_2,\\ h_3=r\sin {\mathrm{\θ}}_1\sin {\mathrm{\θ}}_2\end{array}\right.;$

Wherein, h ₁, h ₂and h ₃be respectively three target functions, the radius that r is spherical coordinate, θ ₁and θ ₂for quadrant angle;

Numerical value K is set, draws K as follows from zero point of reference frame to first quartile ²the equally distributed ray of bar:

$\left\{\begin{array}{c}{\mathrm{\θ}}_{1,l_1}=\frac{l_1\mathrm{\π}}{2(k+1)},l_1=\mathrm{1,2},...,k,\\ {\mathrm{\θ}}_{2,l_2}=\frac{l_2\mathrm{\π}}{2(k+1)},l_1=\mathrm{1,2},...,k\end{array}\right.;$

Obtain weight coefficient as follows according to described ray:

$\left\{\begin{array}{c}{q}_{1,j}=\frac{1}{\cos {\mathrm{\θ}}_{1,j}}\sqrt[3]{\underset{i=1}{\overset{2}{\mathrm{\Π}}}{\sin }^2{\mathrm{\θ}}_{i,j}\cos {\mathrm{\θ}}_{i,j}},\\ q_{2,j}=\frac{\cos {\mathrm{\θ}}_{1,j}}{\sin {\mathrm{\θ}}_{1,j}\cos {\mathrm{\θ}}_{2,j}}{w}_{1,j},\\ q_{3,j}=\frac{\cos {\mathrm{\θ}}_{2,j}}{\sin {\mathrm{\θ}}_{2,j}}{w}_{2,j}\end{array}\right.;$

According to described weight coefficient and described ray, K is set ²individual fitness function as follows:

$F_j\left(v\right)=\underset{1\≤i\≤3}{\max }\left\{w_{i,j}{h}_i\left(v\right)\right\},j=\mathrm{1,2},...,k^2;$

Wherein, F _j(v) be fitness function.

8. the heavy planing method of TA according to claim 5, is characterized in that, in the time that to obtain returning of the TA information of Ge community be multiple, it is further comprising the steps of that the step that TA weight-normality draws is carried out in the heavy planning information Dui Ge of the TA community that described basis is obtained Ge community:

From the heavy planning information of multiple TA, choose the heavy planning information of TA that a kind of default compromise that meets three target functions requires;

According to the heavy planning information of the TA selecting, Dui Ge community is carried out TA weight-normality and is drawn.

9. the heavy planning system of TA, is characterized in that, comprising:

Acquisition module, for obtaining active user's parameter of each community and user's flow parameter of each minizone;

MBM, for building the heavy plan model of TA by obtaining active user's parameter of Ge community and user's flow parameter of each minizone, wherein, the target function of the heavy plan model of described TA comprises that upgrading expense minimum function, call signaling expense minimum function and weight-normality draws Least-cost function;

Parsing module, for the former TA planning information based on each community, solves the heavy plan model of described TA, obtains the heavy planning information of TA of each community;

Planning module, draws for carrying out TA weight-normality according to the heavy planning information Dui Ge of the TA community that obtains Ge community.

10. the heavy planning system of TA according to claim 9, is characterized in that, described acquisition module also for:

Whether the variable quantity that judges active user's parameter of each community and user's flow parameter of each minizone exceedes threshold value;

If so, build the heavy plan model of TA by obtaining active user's parameter of Ge community and user's flow parameter of each minizone;

If not, stopping TA weight-normality draws.

Images