CN102625365B - Method for mobility robustness optimization (MRO) and mobility load balancing (MLB) and base station - Google Patents

Abstract

The embodiment of the invention provides a method for mobility robustness optimization (MRO) and mobility load balancing (MLB) and a base station, which relate to the field of mobile communication and can be used for performing joint optimization on MRO and MLB functions synchronously to increase the overall performance of a system. The method comprises the steps of: setting frequencies fd, fMLB, fSPSA (Simultaneous Perturbation Stochastic Approximation), fMRO and fSAT (Saturation Threshold) according to a frequency setting rule; sending updated Rho c, HOF (Handover Failure), HPP (Handover Ping-Pong) and CDR (Call Drop Rate) to a corresponding controller at the frequency fd; updating Ocs (Cell-Specific Offset) according to updated Rho c by an MLB controller at the frequency fMLB; updating Hys* (Hysteresis) and TTT* (Time to Trigger) by an SPSA controller at the frequency fSPSA according to the updated HOF, HPP and CDR; updating Hys and TTTmro by an MRO controller at the frequency fMRO according to the updated HOF, HPP and CDR as well as the updated Hys* and TTT*; and updating TTT by an SAT controller at the frequency fSAT according to the updated Rhoc and the updated TTTmro.

Description

Method and base station for mobile robustness optimization and mobile load balancing

Technical Field

The present invention relates to the field of mobile communications, and in particular, to a method and a base station for mobility robustness optimization and mobility load balancing.

Background

In the LTE (Long Term Evolution) Project of 3GPP (3rd Generation Partnership Project), the number of network parameters is huge and the complexity is high, and if manual operation is adopted for configuring and managing the network parameters like the conventional network, the problems of high labor cost, high error rate and the like may occur. In order to reduce manual intervention in configuring and managing a base station (eNB) as much as possible, 3GPP proposes a Self-Organizing network (SON), which introduces an automation mechanism to reduce manual intervention, thereby reducing cost.

MRO (Mobility Robustness Optimization) and MLB (Mobility Load Balancing) are two important functions in ad hoc networks. The MLB changes the switching conditions by adjusting the switching parameter Ocs (the specific offset of the cell to the serving cell) to switch part of the users in the overloaded cell to the low-load neighboring cell in advance or delay the switching of the users in the neighboring cell to the overloaded cell, thereby balancing the load among the cells; the MRO changes the handover condition by adjusting the handover parameters TTT (Time to Trigger, Trigger Time) and hysteris (Hysteresis), thereby optimizing the network and avoiding or reducing radio link failure and unnecessary handover related to handover.

In the prior art, a base station independently optimizes the function of MRO or MLB, when the system triggers MLB to adjust Ocs for load balancing, a large number of radio link failures and unnecessary handovers occur, and when triggering MRO to reversely adjust TTT and Hys, the problem of system load imbalance occurs. That is, when the MRO and MLB independently adjust the parameters to optimize their own performance targets, optimization conflicts are caused, and the overall performance of the system is reduced.

Disclosure of Invention

The embodiment of the invention provides a method and a base station for mobile robustness optimization and mobile load balancing, which can synchronously carry out joint optimization on MRO and MLB functions and improve the comprehensive performance of a system.

In order to achieve the above purpose, the embodiment of the invention adopts the following technical scheme:

a method for mobile robustness optimization and mobile load balancing, comprising:

setting parameter update frequency f according to frequency setting principle_dMLB controller driving frequency f for mobile load balancing_MLBOptimizing MRO controller drive frequency f by moving robustness_MROSAT driving frequency f of saturation threshold controller_SATAnd synchronous disturbance random approximation algorithm SPSA controller driveFrequency f_SPSA；

At a frequency f_dInputting updated cell load ρ to the MLB controller and SAT_cInputting updated handover failure rate HOF, handover ping-pong effect HPP and call drop rate CDR to the MRO controller and the SPSA controller;

MLB controller at frequency f_MLBAccording to the updated rho_cUpdating the specific offset Ocs of the user to the serving cell;

SPSA controller with frequency f_SPSAUpdating hysteresis value perturbation value Hys according to the updated HOF, HPP and CDR^*And time-to-trigger disturbance value TTT^*And inputting the updated Hys to the MRO controller^*And TTT^*；

MRO controller at frequency f_MROAccording to the updated HOF, HPP and CDR and the updated Hys^*And TTT^*Updating hysteresis value Hys and optimizing time to trigger TTT_mroAnd inputting an updated TTT to the SAT_mro；

SAT at frequency f_SATAccording to the updated rho_cAnd the updated TTT_mroThe time to trigger TTT is updated.

A base station, comprising:

a setting unit for setting the parameter update frequency f according to the frequency setting principle_dMLB controller driving frequency f for mobile load balancing_MLBOptimizing MRO controller drive frequency f by moving robustness_MROSAT driving frequency f of saturation threshold controller_SATAnd synchronous disturbance random approximation algorithm SPSA controller driving frequency f_SPSA；

A transmitting unit for transmitting at a frequency f_dInputting updated cell load ρ to the MLB controller and SAT_cInputting updated HOF, switching ping-pong effect HPP and call drop rate CDR to the MRO controller and the SPSA controller;

MLB controller for controlling at frequency f_MLBAccording to the updated rho_cUpdating the specific offset Ocs of the user to the serving cell;

SPSA controller for controlling at a frequency f_SPSAUpdating hysteresis value perturbation value Hys according to the updated HOF, HPP and CDR^*And time-to-trigger disturbance value TTT^*And inputting the updated Hys to the MRO controller^*And TTT^*；

MRO controller for controlling the frequency f_MROAccording to the updated HOF, HPP and CDR and the updated Hys^*And TTT^*Updating hysteresis value Hys and optimizing time to trigger TTT_mroAnd inputting an updated TTT to the SAT_mro；

SAT for use at frequency f_SATAccording to the updated rho_cAnd the updated TTT_mroThe time to trigger TTT is updated.

The method and the base station for mobile robustness optimization and mobile load balancing provided by the technical scheme can be combined with the performance index rho of the MLB_cAnd MRO performance indexes HOF, HPP and CDR continuously update TTT, and MRO and MLB functions are synchronously optimized in a combined manner, so that the comprehensive performance of the system is improved.

Drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.

Fig. 1 is a schematic flowchart of a method for mobility robustness optimization and mobility load balancing according to an embodiment of the present invention;

fig. 2 is a block diagram of a base station according to an embodiment of the present invention.

Detailed Description

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

An embodiment of the present invention provides a method for mobility robustness optimization and mobility load balancing, as shown in fig. 1, the method includes the following steps:

101. setting parameter update frequency f according to frequency setting principle_dMLB controller drive frequency f_MLBMRO controller drive frequency f_MROSAT (Saturation threshold controller) driving frequency f_SATAnd SPSA (Simultaneous particle swarm storage Stochastic Approximation) controller driving frequency f_SPSA。

In this embodiment, the frequency setting principle includes: f is_MLBIs less than or equal to f_dSaid f_SATIs equal to f_dSaid f_MROIs greater than f_SPSASaid f_SPSAF or more_HPI。

Wherein, the f_HPIThe time for the performance of each controller to reach the stable state after the parameters Ocs, Hys and TTT are adjusted is the reciprocal of the time. For example, after adjusting the parameters Ocs, Hys and TTT, the performance of the MLB controller and the MRO controller fluctuates, if the time for the performance of the MLB controller to reach the steady state is 1s and the time for the MRO controller to reach the steady state is 5s, the time for the performance of the controllers to reach the steady state is 5s,therefore f_HPI＝1/5＝0.2Hz。

Optionally, in this embodiment, f is_HPIAt 0.2Hz, the base station may set the driving frequencies of the MLB controller, MRO controller, SAT, and SPSA controllers to: f. of_d＝20Hz，f_MLB＝2Hz，f_MRO＝2Hz，f_SAT＝20Hz，f_SPSA＝0.2Hz。

102. At a frequency f_dInputting updated cell load ρ to the MLB controller and SAT_cAnd inputting updated HOF (Handover failure rate), HPP (Handover Ping-Pong Ratio) and CDR (Call Drop Ratio) to the MRO controller and the SPSA controller.

In this embodiment, the base station will cycle by cycleTo statistically update the parameters, HOF, HPP and CDR. In the prior art, a base station calculates a performance index parameter ρ of an MLB controller according to parameters Ocs, Hys and TTT in each controller_cAnd the performance indicator parameters HOF, HPP and CDR of the MRO controller, the specific calculation methods thereof being well known to those skilled in the art and not described in detail herein. The base station calculates rho obtained by calculation every 0.05s_cInputting the data into the MLB controller and the SAT, and updating the rho values in the MLB controller and the SAT_cAnd inputting the counted parameters HOF, HPP and CDR into the MRO controller and the SPSA controller, and updating the three parameter values HOF, HPP and CDR in the MRO controller and the SPSA controller.

103. MLB controller at frequency f_MLBAccording to the updated rho_cAnd updating the Ocs.

In this embodiment, the base station inputs updated ρ to the MLB controller every 0.05s_cValue, MLB controller every otherAccording to updates in MLBρ_cAnd calculating the parameter Ocs, and updating the parameter Ocs in the MLB controller.

Optionally, in an initial state, that is, when the MLB controller is driven for the first time, the parameter Ocs in the MLB controller is a preset initial value Ocs₁。

When the MLB controller is driven for the ith (i is more than or equal to 1) time in a non-initial state, the Ocs is the Ocs_iSaid Ocs_iEvery 0.5s can be calculated according to the following formula:

<math> <mrow> <mi>OCS</mi> <mo>=</mo> <mrow> <mo>(</mo> <mi>r</mi> <mo>-</mo> <msub> <mi>&rho;</mi> <mi>c</mi> </msub> <mo>)</mo> </mrow> <msub> <mi>K</mi> <mi>P</mi> </msub> <mo>+</mo> <mfrac> <msub> <mi>a</mi> <mn>0</mn> </msub> <msub> <mi>b</mi> <mn>0</mn> </msub> </mfrac> <mi>r</mi> <mo>,</mo> </mrow> </math> <math> <mrow> <msub> <mi>K</mi> <mi>p</mi> </msub> <mo>=</mo> <msqrt> <mfrac> <msubsup> <mi>a</mi> <mn>0</mn> <mn>2</mn> </msubsup> <msubsup> <mi>b</mi> <mn>0</mn> <mn>2</mn> </msubsup> </mfrac> <mo>+</mo> <mfrac> <mi>Q</mi> <mi>&rho;R</mi> </mfrac> </msqrt> <mo>-</mo> <mfrac> <msub> <mi>a</mi> <mn>0</mn> </msub> <msub> <mi>b</mi> <mn>0</mn> </msub> </mfrac> <mo>,</mo> </mrow> </math> where r is a preset serving cell load target value, ρ_cServing cell load rho counted for current base station_cRho, Q and R are preset constants, a₀And b₀According to the rho_cAnd updated for the i-1 th timeOcs_i-1And (6) calculating.

Each base station serves a serving cell, r is a preset serving cell load target value, and if r is 0.5, the adjustment target of the MLB controller is represented to make the load of the serving cell half of the bearable load; r is 1, it means that the adjustment target of the controller is such that the load of the serving cell is full. If the load rho of the service cell counted by the current base station_cIf the load of the service cell is larger than r, the load of the service cell is adjusted to be smaller, and if the load of the service cell is counted by the current base station, rho is adjusted to be smaller_cLess than r will adjust the serving cell load in the larger direction. In this embodiment, the base station statistically updates the current load ρ every 0.05s_cThe MLB controller can track the current load rho in time_cDifference from the target load r, the current load ρ is made by changing the value of Ocs_cAnd continuously adjusting the load in the target direction to reach the preset target value r of the serving cell as much as possible. In addition, the specific value of r depends on the current network condition, and is the final target value of the load balancing of the serving cell.

Optionally, in this embodiment, the p is a function of the p_cAnd Ocs of i-1 th update_i-1Calculating to obtain a₀And b₀There are two methods, one is a BGF (Bounded-Gain-forming) algorithm, and one is a slope estimation algorithm.

Wherein, the BGF algorithm is as follows:

the first time the MLB controller is driven, the a₀And b₀Is a preset constant value a₁And b₁. At the moment, the Ocs is a preset initial value Ocs₁。

The a is in a non-initial state, i.e., the ith time the MLB controller is driven₀And b₀This can be calculated by the following method:

calculating a filtered signal matrix W_iW is as described_iThe following formula is satisfied: <math> <mrow> <msub> <mi>W</mi> <mi>i</mi> </msub> <mo>=</mo> <mrow> <mo>(</mo> <mn>1</mn> <mo>-</mo> <msub> <mi>&alpha;</mi> <mn>0</mn> </msub> <mo>)</mo> </mrow> <msub> <mi>W</mi> <mrow> <mi>i</mi> <mo>-</mo> <mn>1</mn> </mrow> </msub> <mo>+</mo> <msub> <mi>&alpha;</mi> <mn>0</mn> </msub> <mi>&Delta;t</mi> <msup> <mfenced open='[' close=']'> <mtable> <mtr> <mtd> <msub> <mi>&rho;</mi> <mi>c</mi> </msub> </mtd> </mtr> <mtr> <mtd> <msub> <mi>Ocs</mi> <mrow> <mi>i</mi> <mo>-</mo> <mn>1</mn> </mrow> </msub> </mtd> </mtr> </mtable> </mfenced> <mi>T</mi> </msup> <mo>,</mo> </mrow> </math> wherein, W₁＝[ρ_c Ocs₁]，α₀Is the low pass filter bandwidth constant, Δ t is the time step for invoking the BGF algorithm;

according to the W_iCalculating a gain matrix P_iSaid P is_iThe following formula is satisfied:wherein, P₁Is a preset initial value P₀；Is P_iThe derivative with respect to time t satisfies the following equation:λ_ias a forgetting factor, the following formula is satisfied:||P_i||₂the following formula is satisfied for the Euclidean norm of the gain matrix: <math> <mrow> <msub> <mrow> <mo>|</mo> <mo>|</mo> <msub> <mi>P</mi> <mi>i</mi> </msub> <mo>|</mo> <mo>|</mo> </mrow> <mn>2</mn> </msub> <mo>=</mo> <msqrt> <msub> <mi>&lambda;</mi> <mi>max</mi> </msub> <mrow> <mo>(</mo> <msubsup> <mi>P</mi> <mrow> <mi>i</mi> <mo>-</mo> <mn>1</mn> </mrow> <mi>T</mi> </msubsup> <msub> <mi>P</mi> <mrow> <mi>i</mi> <mo>-</mo> <mn>1</mn> </mrow> </msub> <mo>)</mo> </mrow> </msqrt> <mo>,</mo> </mrow> </math> is thatThe characteristic value of (2).

According to the W_iAnd P_iCalculating an estimated parameter vectorThe above-mentionedThe following formula is satisfied: <math> <mrow> <msub> <mover> <mi>a</mi> <mo>^</mo> </mover> <mi>i</mi> </msub> <mo>=</mo> <msub> <mover> <mi>a</mi> <mo>^</mo> </mover> <mrow> <mi>i</mi> <mo>-</mo> <mn>1</mn> </mrow> </msub> <mo>+</mo> <msub> <mover> <mi>a</mi> <mover> <mo>^</mo> <mo>&CenterDot;</mo> </mover> </mover> <mi>i</mi> </msub> <mi>&Delta;t</mi> <mo>,</mo> </mrow> </math> wherein, <math> <mrow> <msub> <mover> <mi>a</mi> <mo>^</mo> </mover> <mn>1</mn> </msub> <mo>=</mo> <mfenced open='[' close=']'> <mtable> <mtr> <mtd> <msub> <mi>&alpha;</mi> <mn>0</mn> </msub> <mo>-</mo> <msub> <mi>a</mi> <mn>1</mn> </msub> </mtd> </mtr> <mtr> <mtd> <msub> <mi>b</mi> <mn>1</mn> </msub> </mtd> </mtr> </mtable> </mfenced> <mo>,</mo> </mrow> </math> <math> <mrow> <msub> <mover> <mi>a</mi> <mover> <mo>^</mo> <mo>&CenterDot;</mo> </mover> </mover> <mi>i</mi> </msub> <mo>=</mo> <mo>-</mo> <msub> <mi>P</mi> <mi>i</mi> </msub> <msubsup> <mi>W</mi> <mi>i</mi> <mi>T</mi> </msubsup> <msub> <mi>e</mi> <mi>i</mi> </msub> <mo>,</mo> </mrow> </math> <math> <mrow> <msub> <mi>e</mi> <mi>i</mi> </msub> <mo>=</mo> <msub> <mi>W</mi> <mi>i</mi> </msub> <msub> <mover> <mi>a</mi> <mo>^</mo> </mover> <mrow> <mi>i</mi> <mo>-</mo> <mn>1</mn> </mrow> </msub> <mo>-</mo> <msub> <mi>&rho;</mi> <mi>c</mi> </msub> <mo>.</mo> </mrow> </math>

according to the aboveCalculating the a₀And b₀When the i-th driving MLB controller calculates the Ocs, the a₀Is equal to a_iSaid b is₀Is equal to b_i(ii) a Wherein, the a_iThe following formula is satisfied:b is_iThe following formula is satisfied:

the slope estimation algorithm is as follows:

the first time the MLB controller is driven, the a₀And b₀Is a preset constant value a₁And b₁. At the moment, the Ocs is a preset initial value Ocs₁。

The a is in a non-initial state, i.e., the ith time the MLB controller is driven₀And b₀This can be calculated by the following method:

calculating a filtered signal matrix W_iW is as described_iThe following formula is satisfied: <math> <mrow> <msub> <mi>W</mi> <mi>i</mi> </msub> <mo>=</mo> <mrow> <mo>(</mo> <mn>1</mn> <mo>-</mo> <msub> <mi>&alpha;</mi> <mn>0</mn> </msub> <mo>)</mo> </mrow> <msub> <mi>W</mi> <mrow> <mi>i</mi> <mo>-</mo> <mn>1</mn> </mrow> </msub> <mo>+</mo> <msub> <mi>&alpha;</mi> <mn>0</mn> </msub> <mi>&Delta;t</mi> <msup> <mfenced open='[' close=']'> <mtable> <mtr> <mtd> <msub> <mi>&rho;</mi> <mi>c</mi> </msub> </mtd> </mtr> <mtr> <mtd> <msub> <mi>Ocs</mi> <mrow> <mi>i</mi> <mo>-</mo> <mn>1</mn> </mrow> </msub> </mtd> </mtr> </mtable> </mfenced> <mi>T</mi> </msup> <mo>,</mo> </mrow> </math> wherein, W₁＝[ρ_c Ocs₁]，α₀Is the low pass filter bandwidth constant and Δ t is the time step to invoke the slope estimation algorithm.

According to the W_iCalculating an estimated parameter vectorThe above-mentionedThe following formula is satisfied: <math> <mrow> <msub> <mover> <mi>a</mi> <mo>^</mo> </mover> <mi>i</mi> </msub> <mo>=</mo> <msub> <mover> <mi>a</mi> <mo>^</mo> </mover> <mrow> <mi>i</mi> <mo>-</mo> <mn>1</mn> </mrow> </msub> <mo>+</mo> <msub> <mover> <mi>a</mi> <mover> <mo>^</mo> <mo>&CenterDot;</mo> </mover> </mover> <mi>i</mi> </msub> <mi>&Delta;t</mi> <mo>,</mo> </mrow> </math> wherein, <math> <mrow> <msub> <mover> <mi>a</mi> <mo>^</mo> </mover> <mn>1</mn> </msub> <mo>=</mo> <mfenced open='[' close=']'> <mtable> <mtr> <mtd> <msub> <mi>&alpha;</mi> <mn>0</mn> </msub> <mo>-</mo> <msub> <mi>a</mi> <mn>1</mn> </msub> </mtd> </mtr> <mtr> <mtd> <msub> <mi>b</mi> <mn>1</mn> </msub> </mtd> </mtr> </mtable> </mfenced> <mo>,</mo> </mrow> </math> <math> <mrow> <msub> <mover> <mi>a</mi> <mover> <mo>^</mo> <mo>&CenterDot;</mo> </mover> </mover> <mi>i</mi> </msub> <mo>=</mo> <mo>-</mo> <msub> <mi>P</mi> <mi>i</mi> </msub> <msubsup> <mi>W</mi> <mi>i</mi> <mi>T</mi> </msubsup> <msub> <mi>e</mi> <mi>i</mi> </msub> <mo>,</mo> </mrow> </math> <math> <mrow> <msub> <mi>e</mi> <mi>i</mi> </msub> <mo>=</mo> <msub> <mi>W</mi> <mi>i</mi> </msub> <msub> <mover> <mi>a</mi> <mo>^</mo> </mover> <mrow> <mi>i</mi> <mo>-</mo> <mn>1</mn> </mrow> </msub> <mo>-</mo> <msub> <mi>&rho;</mi> <mi>c</mi> </msub> <mo>,</mo> </mrow> </math> and P is a preset gain matrix.

According to the aboveCalculating the a₀And b₀When the i-th driving MLB controller calculates the Ocs, the a₀Is equal to a_iSaid b is₀Is equal to b_i(ii) a Wherein, the a_iThe following formula is satisfied:b is_iThe following formula is satisfied:

104. SPSA controller with frequency f_SPSAAccording toThe updated HOF, HPP and CDR and the updated TTT and Hys update hysteresis value perturbation values Hys^*And time-to-trigger disturbance value TTT^*And inputting the updated Hys to the MRO controller^*And TTT^*。

SPSA controller with periodThe Hys when the SPSA controller is driven for the jth time^*Is equal to Hys^* _jThe TTT^*Is equal to TTT^* _j；

Calculating Hys when the SPSA controller is driven for the 3n +1 th time^*And TTT^*Said Hys^*And TTT^*The following formula is satisfied:

<math> <mrow> <msup> <mi>TTT</mi> <mo>*</mo> </msup> <mo>=</mo> <msub> <msup> <mi>TTT</mi> <mo>*</mo> </msup> <mi>j</mi> </msub> <mo>=</mo> <msub> <msup> <mi>TTT</mi> <mo>*</mo> </msup> <mrow> <mn>3</mn> <mi>n</mi> <mo>+</mo> <mn>1</mn> </mrow> </msub> <mo>=</mo> <mi>TTT</mi> <mo>+</mo> <mi>c</mi> <msub> <msubsup> <mi>&Delta;</mi> <mi>k</mi> <mi>TTT</mi> </msubsup> <mi>n</mi> </msub> <mo>,</mo> </mrow> </math>

<math> <mrow> <msup> <mi>Hys</mi> <mo>*</mo> </msup> <mo>=</mo> <msub> <msup> <mi>Hys</mi> <mo>*</mo> </msup> <mi>j</mi> </msub> <mo>=</mo> <msub> <msup> <mi>Hys</mi> <mo>*</mo> </msup> <mrow> <mn>3</mn> <mi>n</mi> <mo>+</mo> <mn>1</mn> </mrow> </msub> <mo>=</mo> <mi>Hys</mi> <mo>+</mo> <msub> <msubsup> <mi>c&Delta;</mi> <mi>k</mi> <mi>KKK</mi> </msubsup> <mi>n</mi> </msub> <mo>;</mo> </mrow> </math> wherein n is an integer of 0 or more, andandand independently generating new variable values for two independent variables which are randomly generated, obey binomial distribution and take the value of +/-1 in each calculation.

Calculating loss parameters L + and TTT when the SPSA controller is driven for the 3n +2 times^*And Hys^*The L + and TTT^*And Hys^*The following formula is satisfied:

$L+=\frac{{\mathrm{HOFw}}_1+{\mathrm{HPPw}}_2+{\mathrm{CDRw}}_3}{w_1+w_2+w_3},$

<math> <mrow> <msup> <mi>TTT</mi> <mo>*</mo> </msup> <mo>=</mo> <msub> <msup> <mi>TTT</mi> <mo>*</mo> </msup> <mi>j</mi> </msub> <mo>=</mo> <msub> <msup> <mi>TTT</mi> <mo>*</mo> </msup> <mrow> <mn>3</mn> <mi>n</mi> <mo>+</mo> <mn>1</mn> </mrow> </msub> <mo>=</mo> <mi>TTT</mi> <mo>-</mo> <mi>c</mi> <msub> <msubsup> <mi>&Delta;</mi> <mi>k</mi> <mi>TTT</mi> </msubsup> <mi>n</mi> </msub> <mo>,</mo> </mrow> </math>

<math> <mrow> <msup> <mi>Hys</mi> <mo>*</mo> </msup> <mo>=</mo> <msub> <msup> <mi>Hys</mi> <mo>*</mo> </msup> <mi>j</mi> </msub> <mo>=</mo> <msub> <msup> <mi>Hys</mi> <mo>*</mo> </msup> <mrow> <mn>3</mn> <mi>n</mi> <mo>+</mo> <mn>1</mn> </mrow> </msub> <mo>=</mo> <mi>Hys</mi> <mo>-</mo> <msub> <msubsup> <mi>c&Delta;</mi> <mi>k</mi> <mi>KKK</mi> </msubsup> <mi>n</mi> </msub> <mo>;</mo> </mrow> </math>

calculating the inverse loss parameter L-, Hys when the SPSA controller is driven for the 3n +3 times^*And TTT^*The L-TTT^*And Hys^*The following formula is satisfied:

$L-=\frac{{\mathrm{HOFw}}_1+{\mathrm{HPPw}}_2+{\mathrm{CDRw}}_3}{w_1+w_2+w_3},$

${\mathrm{TTT}}^{*}={{\mathrm{TTT}}^{*}}_j={{\mathrm{TTT}}^{*}}_{3n+2}=\mathrm{TTT}-{a{g}_k^{\mathrm{TTT}}}_n,$

${\mathrm{Hys}}^{*}={{\mathrm{Hys}}^{*}}_j={{\mathrm{Hys}}^{*}}_{3n+2}=\mathrm{Hys}-{a{g}_k^{\mathrm{Hys}}}_n;$ wherein, theAndthe following formula is satisfied: <math> <mrow> <msub> <msubsup> <mi>ag</mi> <mi>k</mi> <mi>TTT</mi> </msubsup> <mi>n</mi> </msub> <mo>=</mo> <mfrac> <mrow> <mi>L</mi> <mo>+</mo> <mo>-</mo> <mi>L</mi> <mo>-</mo> </mrow> <msub> <msubsup> <mrow> <mn>2</mn> <mi>c&Delta;</mi> </mrow> <mi>k</mi> <mi>TTT</mi> </msubsup> <mi>n</mi> </msub> </mfrac> <mo>,</mo> </mrow> </math> <math> <mrow> <msub> <msubsup> <mi>ag</mi> <mi>k</mi> <mi>Hys</mi> </msubsup> <mi>n</mi> </msub> <mo>=</mo> <mfrac> <mrow> <mi>L</mi> <mo>+</mo> <mo>-</mo> <mi>L</mi> <mo>-</mo> </mrow> <msub> <msubsup> <mrow> <mn>2</mn> <mi>c&Delta;</mi> </mrow> <mi>k</mi> <mi>Hys</mi> </msubsup> <mi>n</mi> </msub> </mfrac> <mo>.</mo> </mrow> </math>

it should be noted here that the HOF, HPP and CDR are updated every 0.05s, and the TTT and Hys are updated every 0.5 s.

105. MRO controller at frequency f_MROAccording to the updated HOF, HPP and CDR and the updated Hys^*And TTT^*Updating hysteresis value Hys and optimizing time to trigger TTT_mroAnd inputting an updated TTT to the SAT_mro。

Optionally, the MRO controller cyclesAccording to the Hys^*And TTT^*Calculating a gamma value, the gamma value satisfying the following formula:

<math> <mrow> <mi>&gamma;</mi> <mo>=</mo> <mi>arctan</mi> <mrow> <mo>(</mo> <mfrac> <msup> <mi>Hys</mi> <mo>*</mo> </msup> <msup> <mrow> <mn>2</mn> <mi>TTT</mi> </mrow> <mo>*</mo> </msup> </mfrac> <mo>)</mo> </mrow> </mrow> </math>

if gamma is less than or equal to the preset lower limit value, Hys and TTT_mroThe following formula is satisfied:

TTT_mro＝TTT^*-y_mro，Hys＝Hys^*；

if gamma is greater than or equal to a preset lower limit value and less than or equal to a preset upper limit value, Hys and TTT_mroThe following formula is satisfied:

TTT_mro＝TTT^*-y_mro，Hys＝Hys^*-y_mro。

if gamma is greater than or equal to a preset upper limit value, Hys and TTT_mroThe following formula is satisfied:

TTT_mro＝TTT^*，Hys＝Hys^*-y_mro。

wherein, said y_mroAt a frequency f_MROCalculated according to the updated HOF, HPP and CDR, y_mroSatisfy the requirement of $y_{\mathrm{mro}}=\frac{{\mathrm{HOFw}}_1+{\mathrm{HPPw}}_2+{\mathrm{CDRw}}_3}{w_1+w_2+w_3},$ w₁、w₂、w₃Is a preset parameter weight value.

It should be noted that the preset lower limit and the preset upper limit may be set according to the specific situation of the actual scene, and are not fixed, and they function to make the ratio of Hys and TTT always stable within a proper range. In this embodiment, the preset lower limit value may be set to 20 °, and the preset upper limit value may be set to 80 °.

For a preset parameter weight value w₁、w₂、w₃In general, operators place more importance on HOF and CDR, i.e., w₁And w₃General ratio w₂Large, so that said y_mroIn most cases positive values. Thus, the MRO controller is paired with the TTT^*And Hys^*The adjustment of (c) is generally to remain unchanged or to a small direction, i.e. to a direction more favorable for the HOF and CDR.

106. SAT at frequency f_SATAccording to the updated rho_cAnd the updated TTT_mroThe time to trigger TTT is updated.

In a period ofAccording to rho_cCalculating a value of l, the value of l satisfying the following formula:

l＝1+(1-ρ_c)·15

if TTT_mroIf the value is less than the value of l, the TTT is equal to the TTT_mro；

If TTT_mroAnd if the value is larger than the value of l, the TTT is equal to l.

I.e. the SAT is updated every 0.05s for TTT according to the MRO controller performance index parameters HOF, HPP and CDR and the MLB controller performance index parameter p, as can be seen from steps 104, 105, 106_cAre jointly decided.

When the initial state of the base station, that is, when t is 0, the MLB controller, the SPSA controller, the MRO controller, and the SAT are sequentially driven, and the base station outputs the performance index parameter ρ to each controller_cHOF, HPP and CDR, updated every 0.05 s.

Starting from t being 0s, driving the SAT once every 0.05s, and performing step 106 to update an adjustment parameter TTT influencing the MRO; driving the MLB controller, the MRO controller and the SAT once every 0.5s in sequence, performing steps 103, 105 and 106, and updating the adjusting parameters Ocs, Hys and TTT influencing the MRO and the MLB; the MLB controller, the SPSA controller, the MRO controller, and the SAT are sequentially driven every 5s, and steps 103, 104, 105, and 106 are performed.

An embodiment of the present invention further provides a base station, where the base station is configured to complete the foregoing method, and as shown in fig. 2, the base station includes: setting unit 201, transmitting unit 202, MLB controller 203, SPSA controller 204, MRO controller 205, SAT 206.

The setting unit 201 is configured to set a parameter update frequency f according to a frequency setting principle_dMLB controller driving frequency f for mobile load balancing_MLBOptimizing MRO controller drive frequency f by moving robustness_MROSAT driving frequency f of saturation threshold controller_SATAnd synchronous disturbance random approximation algorithm SPSA controller drive frequency fspssa.

The frequency setting principle comprises: f is_MLBIs less than or equal to f_dSaid f_SATIs equal to f_dSaid f_MROIs greater than f_SPSASaid f_SPSAF or more_HPI(ii) a Wherein, the f_HPIThe time for the performance of each controller to reach the stable state after the parameters Ocs, Hys and TTT are adjusted is the reciprocal of the time. Optionally, at said f_HPIIn the case of 0.2Hz, the setting unit 201 may set the driving frequencies of the MLB controller, the MRO controller, the SAT, and the SPSA controller to: f. of_d＝20Hz，f_MLB＝2Hz，f_MRO＝2Hz，f_SAT＝20Hz，f_SPSA＝0.2Hz。

A transmitting unit 202 for transmitting at a frequency f_dInputting updated cell load ρ to the MLB controller and SAT_cInputting the updated HOF, the switching ping-pong effect HPP and the call drop rate CDR to the MRO controller and the SPSA controller.

The base station will cycle by cycleTo statistically update the parameters, HOF, HPP and CDR. In the prior art, a base station calculates a performance index parameter ρ of an MLB controller according to parameters Ocs, Hys and TTT in each controller_cAnd the performance indicator parameters HOF, HPP and CDR of the MRO controller, the specific calculation methods thereof being well known to those skilled in the art and not described in detail herein. The transmission unit 202 calculates the obtained ρ every 0.05s_cInputting the data into the MLB controller and the SAT, and updating the rho values in the MLB controller and the SAT_cAnd inputting the counted parameters HOF, HPP and CDR into the MRO controller and the SPSA controller, and updating the three parameter values HOF, HPP and CDR in the MRO controller and the SPSA controller. And the controllers jointly adjust the adjusting parameters Ocs, Hys and TTT of the MRO and MLB functions according to the continuously updated parameters, so that the system is comprehensive and optimal.

MLB controller 203 for controlling the frequency f_MLBAccording to the updated rho_cAnd updating the Ocs.

The transmitting unit 202 inputs the updated ρ to the MLB controller 203 every 0.05s_cValue, MLB controller 203 every otherAccording to the updated rho in MLB_cAnd calculating the parameter Ocs, and updating the parameter Ocs in the MLB controller.

Optionally, in an initial state, that is, when the MLB controller is driven for the first time, the parameter Ocs in the MLB controller is a preset initial value Ocs₁。

In a non-initial state, i.e., when the MLB controller 203 is driven i (i ≧ 1) th time, the Ocs is Ocs_iSaid Ocs_iEvery 0.5s can be calculated according to the following formula:

<math> <mrow> <mi>Ocs</mi> <mo>=</mo> <mrow> <mo>(</mo> <mi>r</mi> <mo>-</mo> <msub> <mi>&rho;</mi> <mi>c</mi> </msub> <mo>)</mo> </mrow> <msub> <mi>K</mi> <mi>P</mi> </msub> <mo>+</mo> <mfrac> <msub> <mi>a</mi> <mn>0</mn> </msub> <msub> <mi>b</mi> <mn>0</mn> </msub> </mfrac> <mi>r</mi> <mo>,</mo> </mrow> </math> <math> <mrow> <msub> <mi>K</mi> <mi>p</mi> </msub> <mo>=</mo> <msqrt> <mfrac> <msubsup> <mi>a</mi> <mn>0</mn> <mn>2</mn> </msubsup> <msubsup> <mi>b</mi> <mn>0</mn> <mn>2</mn> </msubsup> </mfrac> <mo>+</mo> <mfrac> <mi>Q</mi> <mi>&rho;R</mi> </mfrac> </msqrt> <mo>-</mo> <mfrac> <msub> <mi>a</mi> <mn>0</mn> </msub> <msub> <mi>b</mi> <mn>0</mn> </msub> </mfrac> <mo>,</mo> </mrow> </math> where r is a preset serving cell load target value, ρ_cServing cell load rho counted for current base station_cRho, Q and R are preset constants，a₀And b₀According to the rho_cAnd Ocs calculation.

r is a preset serving cell load target value, if r is 0.5, it means that the adjustment target of the MLB controller 203 is such that the load of the serving cell of the cell is half of the bearable load; r is 1, it means that the adjustment target of the controller is such that the load of the serving cell is full. If the load rho of the cell serving cell counted by the current base station_cIf the load of the cell serving cell is larger than r, the load of the cell serving cell is adjusted to be smaller, and if the load of the cell serving cell is counted by the current base station, rho_cLess than r will adjust the serving cell load in the larger direction. In this embodiment, the base station statistically updates the current load ρ every 0.05s_cThe MLB controller can track the current load rho in time_cDifference from the target load r, the current load ρ is made by changing the value of Ocs_cAnd continuously adjusting the load in the target direction to reach the preset target value r of the serving cell as much as possible. In addition, the specific value of r depends on the current network condition, and is the final target value of the load balancing of the serving cell.

Optionally, in this embodiment, the p is a function of the p_cAnd Ocs of i-1 th update_i-1Calculating to obtain a₀And b₀There are two methods, one is the BGF algorithm and one is the slope estimation algorithm.

Wherein, the BGF algorithm is as follows:

the first time the MLB controller 203 is driven, the a₀And b₀Is a preset constant value a₁And b₁. At the moment, the Ocs is a preset initial value Ocs₁。

The a is said at the non-initial state, i.e., the i-th time when the MLB controller 203 is driven₀And b₀This can be calculated by the following method:

calculating a filtered signal matrix W_iW is as described_iThe following formula is satisfied: <math> <mrow> <msub> <mi>W</mi> <mi>i</mi> </msub> <mo>=</mo> <mrow> <mo>(</mo> <mn>1</mn> <mo>-</mo> <msub> <mi>&alpha;</mi> <mn>0</mn> </msub> <mo>)</mo> </mrow> <msub> <mi>W</mi> <mrow> <mi>i</mi> <mo>-</mo> <mn>1</mn> </mrow> </msub> <mo>+</mo> <msub> <mi>&alpha;</mi> <mn>0</mn> </msub> <mi>&Delta;t</mi> <msup> <mfenced open='[' close=']'> <mtable> <mtr> <mtd> <msub> <mi>&rho;</mi> <mi>c</mi> </msub> </mtd> </mtr> <mtr> <mtd> <msub> <mi>Ocs</mi> <mrow> <mi>i</mi> <mo>-</mo> <mn>1</mn> </mrow> </msub> </mtd> </mtr> </mtable> </mfenced> <mi>T</mi> </msup> <mo>,</mo> </mrow> </math> wherein, W₁＝[ρ_c Ocs₁]，α₀Is the low pass filter bandwidth constant, Δ t is the time step for invoking the BGF algorithm;

according to the W_iCalculating a gain matrix P_iSaid P is_iThe following formula is satisfied:wherein, P₁Is a preset initial value P₀；Is P_iThe derivative with respect to time t satisfies the following equation:λ_ias a forgetting factor, the following formula is satisfied:||P_i||₂the following formula is satisfied for the Euclidean norm of the gain matrix: <math> <mrow> <msub> <mrow> <mo>|</mo> <mo>|</mo> <msub> <mi>P</mi> <mi>i</mi> </msub> <mo>|</mo> <mo>|</mo> </mrow> <mn>2</mn> </msub> <mo>=</mo> <msqrt> <msub> <mi>&lambda;</mi> <mi>max</mi> </msub> <mrow> <mo>(</mo> <msubsup> <mi>P</mi> <mrow> <mi>i</mi> <mo>-</mo> <mn>1</mn> </mrow> <mi>T</mi> </msubsup> <msub> <mi>P</mi> <mrow> <mi>i</mi> <mo>-</mo> <mn>1</mn> </mrow> </msub> <mo>)</mo> </mrow> </msqrt> <mo>,</mo> </mrow> </math> is thatThe characteristic value of (2).

According to the W_iAnd P_iCalculating an estimated parameter vectorThe above-mentionedThe following formula is satisfied: <math> <mrow> <msub> <mover> <mi>a</mi> <mo>^</mo> </mover> <mi>i</mi> </msub> <mo>=</mo> <msub> <mover> <mi>a</mi> <mo>^</mo> </mover> <mrow> <mi>i</mi> <mo>-</mo> <mn>1</mn> </mrow> </msub> <mo>+</mo> <msub> <mover> <mi>a</mi> <mover> <mo>^</mo> <mo>&CenterDot;</mo> </mover> </mover> <mi>i</mi> </msub> <mi>&Delta;t</mi> <mo>,</mo> </mrow> </math> wherein, <math> <mrow> <msub> <mover> <mi>a</mi> <mo>^</mo> </mover> <mn>1</mn> </msub> <mo>=</mo> <mfenced open='[' close=']'> <mtable> <mtr> <mtd> <msub> <mi>&alpha;</mi> <mn>0</mn> </msub> <mo>-</mo> <msub> <mi>a</mi> <mn>1</mn> </msub> </mtd> </mtr> <mtr> <mtd> <msub> <mi>b</mi> <mn>1</mn> </msub> </mtd> </mtr> </mtable> </mfenced> <mo>,</mo> </mrow> </math> <math> <mrow> <msub> <mover> <mi>a</mi> <mover> <mo>^</mo> <mo>&CenterDot;</mo> </mover> </mover> <mi>i</mi> </msub> <mo>=</mo> <mo>-</mo> <msub> <mi>P</mi> <mi>i</mi> </msub> <msubsup> <mi>W</mi> <mi>i</mi> <mi>T</mi> </msubsup> <msub> <mi>e</mi> <mi>i</mi> </msub> <mo>,</mo> </mrow> </math> <math> <mrow> <msub> <mi>e</mi> <mi>i</mi> </msub> <mo>=</mo> <msub> <mi>W</mi> <mi>i</mi> </msub> <msub> <mover> <mi>a</mi> <mo>^</mo> </mover> <mrow> <mi>i</mi> <mo>-</mo> <mn>1</mn> </mrow> </msub> <mo>-</mo> <msub> <mi>&rho;</mi> <mi>c</mi> </msub> <mo>.</mo> </mrow> </math>

according to the aboveCalculating the a₀And b₀When the i-th driving MLB controller calculates the Ocs, the a₀Is equal to a_iSaid b is₀Is equal to b_i(ii) a Wherein, the a_iThe following formula is satisfied:b is_iThe following formula is satisfied:

the slope estimation algorithm is as follows:

the first time the MLB controller 203 is driven, the a₀And b₀Is a preset constant value a₁And b₁. At the moment, the Ocs is a preset initial value Ocs₁。

The a is said at the non-initial state, i.e., the i-th time when the MLB controller 203 is driven₀And b₀This can be calculated by the following method:

calculating a filtered signal matrix W_iW is as described_iThe following formula is satisfied: <math> <mrow> <msub> <mi>W</mi> <mi>i</mi> </msub> <mo>=</mo> <mrow> <mo>(</mo> <mn>1</mn> <mo>-</mo> <msub> <mi>&alpha;</mi> <mn>0</mn> </msub> <mo>)</mo> </mrow> <msub> <mi>W</mi> <mrow> <mi>i</mi> <mo>-</mo> <mn>1</mn> </mrow> </msub> <mo>+</mo> <msub> <mi>&alpha;</mi> <mn>0</mn> </msub> <mi>&Delta;t</mi> <msup> <mfenced open='[' close=']'> <mtable> <mtr> <mtd> <msub> <mi>&rho;</mi> <mi>c</mi> </msub> </mtd> </mtr> <mtr> <mtd> <msub> <mi>Ocs</mi> <mrow> <mi>i</mi> <mo>-</mo> <mn>1</mn> </mrow> </msub> </mtd> </mtr> </mtable> </mfenced> <mi>T</mi> </msup> <mo>,</mo> </mrow> </math> wherein, W₁＝[ρ_c Ocs₁]，α₀Is the low pass filter bandwidth constant and Δ t is the time step to invoke the slope estimation algorithm.

According to the W_iCalculating an estimated parameter vectorThe above-mentionedThe following formula is satisfied: <math> <mrow> <msub> <mover> <mi>a</mi> <mo>^</mo> </mover> <mi>i</mi> </msub> <mo>=</mo> <msub> <mover> <mi>a</mi> <mo>^</mo> </mover> <mrow> <mi>i</mi> <mo>-</mo> <mn>1</mn> </mrow> </msub> <mo>+</mo> <msub> <mover> <mi>a</mi> <mover> <mo>^</mo> <mo>&CenterDot;</mo> </mover> </mover> <mi>i</mi> </msub> <mi>&Delta;t</mi> <mo>,</mo> </mrow> </math> wherein, <math> <mrow> <msub> <mover> <mi>a</mi> <mo>^</mo> </mover> <mn>1</mn> </msub> <mo>=</mo> <mfenced open='[' close=']'> <mtable> <mtr> <mtd> <msub> <mi>&alpha;</mi> <mn>0</mn> </msub> <mo>-</mo> <msub> <mi>a</mi> <mn>1</mn> </msub> </mtd> </mtr> <mtr> <mtd> <msub> <mi>b</mi> <mn>1</mn> </msub> </mtd> </mtr> </mtable> </mfenced> <mo>,</mo> </mrow> </math> <math> <mrow> <msub> <mover> <mi>a</mi> <mover> <mo>^</mo> <mo>&CenterDot;</mo> </mover> </mover> <mi>i</mi> </msub> <mo>=</mo> <mo>-</mo> <msub> <mi>P</mi> <mi>i</mi> </msub> <msubsup> <mi>W</mi> <mi>i</mi> <mi>T</mi> </msubsup> <msub> <mi>e</mi> <mi>i</mi> </msub> <mo>,</mo> </mrow> </math> <math> <mrow> <msub> <mi>e</mi> <mi>i</mi> </msub> <mo>=</mo> <msub> <mi>W</mi> <mi>i</mi> </msub> <msub> <mover> <mi>a</mi> <mo>^</mo> </mover> <mrow> <mi>i</mi> <mo>-</mo> <mn>1</mn> </mrow> </msub> <mo>-</mo> <msub> <mi>&rho;</mi> <mi>c</mi> </msub> <mo>,</mo> </mrow> </math> and P is a preset gain matrix.

According to the aboveCalculating the a₀And b₀When the i-th driving MLB controller 203 calculates the Ocs, the a₀Is equal to a_iSaid b is₀Is equal to b_i(ii) a Wherein, the a_iThe following formula is satisfied:b is_iThe following formula is satisfied:

SPSA controller 204 for controlling the frequency f_SPSAAccording to whatUpdating the HOF, HPP and CDR update hysteresis value perturbation value Hys^*And time-to-trigger disturbance value TTT^*And inputting the updated Hys to the MRO controller^*And TTT^*。

SPSA controller 204 cyclesThe Hys when the SPSA controller is driven for the jth time^*Is equal to Hys^* _jThe TTT^*Is equal to TTT^* _j；

Calculating Hys when the SPSA controller is driven for the 3n +1 th time^*And TTT^*Said Hys^*And TTT^*The following formula is satisfied:

<math> <mrow> <msup> <mi>TTT</mi> <mo>*</mo> </msup> <mo>=</mo> <msub> <msup> <mi>TTT</mi> <mo>*</mo> </msup> <mi>j</mi> </msub> <mo>=</mo> <msub> <msup> <mi>TTT</mi> <mo>*</mo> </msup> <mrow> <mn>3</mn> <mi>n</mi> <mo>+</mo> <mn>1</mn> </mrow> </msub> <mo>=</mo> <mi>TTT</mi> <mo>+</mo> <mi>c</mi> <msub> <msubsup> <mi>&Delta;</mi> <mi>k</mi> <mi>TTT</mi> </msubsup> <mi>n</mi> </msub> <mo>,</mo> </mrow> </math>

<math> <mrow> <msup> <mi>Hys</mi> <mo>*</mo> </msup> <mo>=</mo> <msub> <msup> <mi>Hys</mi> <mo>*</mo> </msup> <mi>j</mi> </msub> <mo>=</mo> <msub> <msup> <mi>Hys</mi> <mo>*</mo> </msup> <mrow> <mn>3</mn> <mi>n</mi> <mo>+</mo> <mn>1</mn> </mrow> </msub> <mo>=</mo> <mi>Hys</mi> <mo>+</mo> <msub> <msubsup> <mi>c&Delta;</mi> <mi>k</mi> <mi>KKK</mi> </msubsup> <mi>n</mi> </msub> <mo>;</mo> </mrow> </math> wherein n is an integer of 0 or more, andandand independently generating new variable values for two independent variables which are randomly generated, obey binomial distribution and take the value of +/-1 each time.

Calculating loss parameters L + and TTT when the SPSA controller is driven for the 3n +2 times^*And Hys^*The L + and TTT^*And Hys^*The following formula is satisfied:

$L+=\frac{{\mathrm{HOFw}}_1+{\mathrm{HPPw}}_2+{\mathrm{CDRw}}_3}{w_1+w_2+w_3},$

<math> <mrow> <msup> <mi>TTT</mi> <mo>*</mo> </msup> <mo>=</mo> <msub> <msup> <mi>TTT</mi> <mo>*</mo> </msup> <mi>j</mi> </msub> <mo>=</mo> <msub> <msup> <mi>TTT</mi> <mo>*</mo> </msup> <mrow> <mn>3</mn> <mi>n</mi> <mo>+</mo> <mn>1</mn> </mrow> </msub> <mo>=</mo> <mi>TTT</mi> <mo>-</mo> <mi>c</mi> <msub> <msubsup> <mi>&Delta;</mi> <mi>k</mi> <mi>TTT</mi> </msubsup> <mi>n</mi> </msub> <mo>,</mo> </mrow> </math>

<math> <mrow> <msup> <mi>Hys</mi> <mo>*</mo> </msup> <mo>=</mo> <msub> <msup> <mi>Hys</mi> <mo>*</mo> </msup> <mi>j</mi> </msub> <mo>=</mo> <msub> <msup> <mi>Hys</mi> <mo>*</mo> </msup> <mrow> <mn>3</mn> <mi>n</mi> <mo>+</mo> <mn>1</mn> </mrow> </msub> <mo>=</mo> <mi>Hys</mi> <mo>-</mo> <msub> <msubsup> <mi>c&Delta;</mi> <mi>k</mi> <mi>KKK</mi> </msubsup> <mi>n</mi> </msub> <mo>;</mo> </mrow> </math>

calculating the inverse loss parameter L-, Hys when the SPSA controller is driven for the 3n +3 times^*And TTT^*The L-TTT^*And Hys^*The following formula is satisfied:

$L-=\frac{{\mathrm{HOFw}}_1+{\mathrm{HPPw}}_2+{\mathrm{CDRw}}_3}{w_1+w_2+w_3},$

${\mathrm{TTT}}^{*}={{\mathrm{TTT}}^{*}}_j={{\mathrm{TTT}}^{*}}_{3n+2}=\mathrm{TTT}-{a{g}_k^{\mathrm{TTT}}}_n,$

${\mathrm{Hys}}^{*}={{\mathrm{Hys}}^{*}}_j={{\mathrm{Hys}}^{*}}_{3n+2}=\mathrm{Hys}-{a{g}_k^{\mathrm{Hys}}}_n;$ wherein, theAndthe following formula is satisfied: <math> <mrow> <msub> <msubsup> <mi>ag</mi> <mi>k</mi> <mi>TTT</mi> </msubsup> <mi>n</mi> </msub> <mo>=</mo> <mfrac> <mrow> <mi>L</mi> <mo>+</mo> <mo>-</mo> <mi>L</mi> <mo>-</mo> </mrow> <msub> <msubsup> <mrow> <mn>2</mn> <mi>c&Delta;</mi> </mrow> <mi>k</mi> <mi>TTT</mi> </msubsup> <mi>n</mi> </msub> </mfrac> <mo>,</mo> </mrow> </math> <math> <mrow> <msub> <msubsup> <mi>ag</mi> <mi>k</mi> <mi>Hys</mi> </msubsup> <mi>n</mi> </msub> <mo>=</mo> <mfrac> <mrow> <mi>L</mi> <mo>+</mo> <mo>-</mo> <mi>L</mi> <mo>-</mo> </mrow> <msub> <msubsup> <mrow> <mn>2</mn> <mi>c&Delta;</mi> </mrow> <mi>k</mi> <mi>Hys</mi> </msubsup> <mi>n</mi> </msub> </mfrac> <mo>,</mo> </mrow> </math> a is a preset arbitrary number.

It should be noted here that the HOF, HPP and CDR are updated every 0.05s, and the TTT and Hys are updated every 0.5 s.

MRO controller 205 for controlling the frequency f_MROAccording to the updated HOF, HPP and CDR and the updated Hys^*And TTT^*Updating hysteresis value Hys and optimizing time to trigger TTT_mroAnd inputting an updated TTT to the SAT_mro。

Optionally, the MRO controller cyclesAccording to the Hys^*And TTT^*Calculating a gamma value, the gamma value satisfying the following formula:

<math> <mrow> <mi>&gamma;</mi> <mo>=</mo> <mi>arctan</mi> <mrow> <mo>(</mo> <mfrac> <msup> <mi>Hys</mi> <mo>*</mo> </msup> <msup> <mrow> <mn>2</mn> <mi>TTT</mi> </mrow> <mo>*</mo> </msup> </mfrac> <mo>)</mo> </mrow> </mrow> </math>

if gamma is less than or equal to the preset lower limit value, Hys and TTT_mroThe following formula is satisfied:

TTT_mro＝TTT^*-y_mro，Hys＝Hys^*；

if gamma is greater than or equal to a preset lower limit value and less than or equal to a preset upper limit value, Hys and TTT_mroThe following formula is satisfied:

TTT_mro＝TTT^*-y_mro，Hys＝Hys^*-y_mro。

if gamma is greater than or equal to a preset upper limit value, Hys and TTT_mroThe following formula is satisfied:

TTT_mro＝TTT^*，Hys＝Hys^*-y_mro。

wherein, said y_mroAt a frequency f_MROCalculated according to the updated HOF, HPP and CDR, y_mroSatisfy the requirement of $y_{\mathrm{mro}}=\frac{{\mathrm{HOFw}}_1+{\mathrm{HPPw}}_2+{\mathrm{CDRw}}_3}{w_1+w_2+w_3},$ w₁、w₂、w₃Is a preset parameter weight value.

It should be noted that the preset lower limit and the preset upper limit may be set according to the specific situation of the actual scene, and are not fixed, and they function to make the ratio of Hys and TTT always stable within a proper range. In this embodiment, the preset lower limit value may be set to 20 °, and the preset upper limit value may be set to 80 °.

For a preset parameter weight value w₁、w₂、w₃In general, operators place more importance on HOF and CDR, i.e., w₁And w₃General ratio w₂Large, so that said y_mroIn most cases positive values. Thus, the MRO controller is paired with the TTT^*And Hys^*The adjustment of (c) is generally to remain unchanged or to a small direction, i.e. to a direction more favorable for the HOF and CDR.

SAT 206 for use at a frequency f_SATAccording to the updated rho_cAnd the updated TTT_mroThe time to trigger TTT is updated.

SAT 206 in cyclesAccording to rho_cCalculating a value of l, the value of l satisfying the following formula:

l＝1+(1-ρ_c)·15

if TTT_mroIf the value is less than the value of l, the TTT is equal to the TTT_mro；

If TTT_mroAnd if the value is larger than the value of l, the TTT is equal to l.

I.e. every 0.05s the SAT is updated with TTT, which is the performance index parameter HOF, HPP and CDR of the MRO controller and the performance index parameter p of the MLB controller_cAre jointly decided.

As shown in fig. 2, the transmission unit 202 updates the parameter ρ every 0.05s_cHOF, HPP and CDR are sent to each controller so that each controller updates the parameters every 0.05 s.

At time t-0, the MLB controller, the SPSA controller, the MRO controller, and the SAT are sequentially operated once. In the next time, the SAT runs once every 0.05s, and the adjusting parameter TTT influencing the MRO is updated and updated; the MLB controller runs once every 0.5s, and updates an adjusting parameter Ocs influencing the MLB; the MRO controller runs once every 0.5s, and updates adjustment parameters Hys and TTT influencing the MRO_mroAnd TTT_mroInputting to the SAT so that the SAT updates the TTT every 0.05 s; the SPSA updates Hys every 5s^*And TTT^*And combining said Hys^*And TTT^*Input to the MRO controller so that the MRO controller updates Hys and TTT_mro。

The embodiment of the invention provides a method and a base station for mobile robustness optimization and mobile load balancing, which can be combined with performance index parameters HOF, HPP and CDR of an MRO controller and performance index parameter rho of an MLB controller_cThe parameters Hys, TTT and Ocs are adjusted together, namely, two functions of mobile robustness optimization and mobile load balancing are adjusted jointly.

Those of ordinary skill in the art will understand that: all or part of the steps for implementing the method embodiments may be implemented by hardware related to program instructions, and the program may be stored in a computer readable storage medium, and when executed, the program performs the steps including the method embodiments; and the aforementioned storage medium includes: various media that can store program codes, such as ROM, RAM, magnetic or optical disks.

The above description is only for the specific embodiment of the present invention, but the scope of the present invention is not limited thereto, and any changes or substitutions that can be easily conceived by those skilled in the art within the technical scope of the present invention are included in the scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the appended claims.

Claims (12)

1. A method for mobile robustness optimization and mobile load balancing, comprising:

setting parameter update frequency f according to frequency setting principle_dMLB controller driving frequency f for mobile load balancing_MLBOptimizing MRO controller drive frequency f by moving robustness_MROSAT driving frequency f of saturation threshold controller_SATAnd synchronous disturbance random approximation algorithm SPSA controller driving frequency f_SPSA；

At a frequency f_dInputting updates to MLB controller and SATCell load of rho_cInputting updated handover failure rate HOF, handover ping-pong effect HPP and call drop rate CDR to the MRO controller and the SPSA controller;

MLB controller at frequency f_MLBAccording to the updated rho_cUpdating the specific offset Ocs of the user to the serving cell;

SPSA controller with frequency f_SPSAUpdating a hysteresis value perturbation value Hys and a trigger time perturbation value TTT according to the updated HOF, HPP and CDR, and inputting the updated Hys and TTT to the MRO controller;

MRO controller at frequency f_MROUpdating hysteresis values Hys and optimal time-to-trigger TTT based on the updated HOF, HPP and CDR and the updated Hys and TTT_mroAnd inputting an updated TTT to the SAT_mro；

SAT at frequency f_SATAccording to the updated rho_cAnd the updated TTT_mroUpdating the time to trigger TTT;

the frequency setting principle comprises:

f is_MLBIs less than or equal to f_dSaid f_SATIs equal to f_dSaid f_MROIs greater than f_SPSASaid f_SPSAF or more_HPI(ii) a Wherein, the f_HPIThe time for the performance of each controller to reach the stable state after the parameters Ocs, Hys and TTT are adjusted is the reciprocal of the time.

2. The method of claim 1, wherein the MLB controller is at a frequency f_MLBAccording to the updated rho_cThe updating Ocs specifically comprises the following steps:

when the MLB controller is driven for the ith time, the Ocs is updated to Ocs_i；

When i is equal to 1, the Ocs is a preset initial value Ocs₁；

When i is an integer greater than 1, the Ocs_iThe following formula is satisfied:

<math> <mrow> <msub> <mi>Ocs</mi> <mi>i</mi> </msub> <mo>=</mo> <mrow> <mo>(</mo> <mi>r</mi> <mo>-</mo> <msub> <mi>&rho;</mi> <mi>c</mi> </msub> <mo>)</mo> </mrow> <msub> <mi>K</mi> <mi>P</mi> </msub> <mo>+</mo> <mfrac> <msub> <mi>a</mi> <mn>0</mn> </msub> <msub> <mi>b</mi> <mn>0</mn> </msub> </mfrac> <mi>r</mi> <mo>,</mo> </mrow> </math> <math> <mrow> <msub> <mi>K</mi> <mi>p</mi> </msub> <mo>=</mo> <msqrt> <mfrac> <msubsup> <mi>a</mi> <mn>0</mn> <mn>2</mn> </msubsup> <msubsup> <mi>b</mi> <mn>0</mn> <mn>2</mn> </msubsup> </mfrac> <mo>+</mo> <mfrac> <mi>Q</mi> <mi>&rho;R</mi> </mfrac> </msqrt> <mo>-</mo> <mfrac> <msub> <mi>a</mi> <mn>0</mn> </msub> <msub> <mi>b</mi> <mn>0</mn> </msub> </mfrac> <mo>;</mo> </mrow> </math>

wherein r is a preset serving cell load target value; rho, Q and R are preset constants, a₀And b₀According to the updated rho_cAnd Ocs of i-1 th update_i-1And (6) calculating.

3. The algorithm of claim 2, wherein a is₀And b₀According to the updated rho_cAnd Ocs of i-1 th update_i-1The calculation specifically includes: calculating the a according to a BGF algorithm₀And b₀(ii) a Wherein the a is calculated according to the BGF algorithm₀And b₀The method comprises the following specific steps:

calculating a filtered signal matrix W_iW is as described_iThe following formula is satisfied:wherein, W₁=[ρ_c Ocs₁]，α₀Is the bandwidth constant of the low-pass filter, and Δ t is the time step for calling the BGF algorithm;

according to the W_iCalculating a gain matrix P_iSaid P is_iThe following formula is satisfied:wherein, P₁Is a preset initial value P₀；Is P_iThe derivative with respect to time t satisfies the following equation:λ_ifor the forgetting factor, the following formula is satisfied||P_i||₂The following formula is satisfied for the Euclidean norm of the gain matrix: is thatA characteristic value of (d);

according to the W_iAnd P_iCalculating an estimated parameter vectorThe above-mentionedThe following formula is satisfied: <math> <mrow> <msub> <mover> <mi>a</mi> <mo>^</mo> </mover> <mi>i</mi> </msub> <mo>=</mo> <msub> <mover> <mi>a</mi> <mo>^</mo> </mover> <mrow> <mi>i</mi> <mo>-</mo> <mn>1</mn> </mrow> </msub> <mo>+</mo> <msub> <mover> <mover> <mi>a</mi> <mo>^</mo> </mover> <mo>&CenterDot;</mo> </mover> <mi>i</mi> </msub> <mi>&Delta;t</mi> <mo>,</mo> </mrow> </math> wherein, <math> <mrow> <msub> <mover> <mi>a</mi> <mo>^</mo> </mover> <mn>1</mn> </msub> <mo>=</mo> <mfenced open='[' close=']'> <mtable> <mtr> <mtd> <msub> <mi>&alpha;</mi> <mn>0</mn> </msub> <mo>-</mo> <msub> <mi>a</mi> <mn>1</mn> </msub> </mtd> </mtr> <mtr> <mtd> <msub> <mi>b</mi> <mn>1</mn> </msub> </mtd> </mtr> </mtable> </mfenced> <mo>,</mo> </mrow> </math> <math> <mrow> <msub> <mover> <mover> <mi>a</mi> <mo>^</mo> </mover> <mo>&CenterDot;</mo> </mover> <mi>i</mi> </msub> <mo>=</mo> <mo>-</mo> <msub> <mi>P</mi> <mi>i</mi> </msub> <msubsup> <mi>W</mi> <mi>i</mi> <mi>T</mi> </msubsup> <msub> <mi>e</mi> <mi>i</mi> </msub> <mo>,</mo> </mrow> </math> <math> <mrow> <msub> <mi>e</mi> <mi>i</mi> </msub> <mo>=</mo> <msub> <mi>W</mi> <mi>i</mi> </msub> <msub> <mover> <mi>a</mi> <mo>^</mo> </mover> <mrow> <mi>i</mi> <mo>-</mo> <mn>1</mn> </mrow> </msub> <mo>-</mo> <msub> <mi>&rho;</mi> <mi>c</mi> </msub> <mo>;</mo> </mrow> </math>

according to the aboveCalculating the a₀And b₀When the i-th driving MLB controller calculates the Ocs, the a₀Is equal to a_iSaid b is₀Is equal to b_i(ii) a Wherein, the a_iThe following formula is satisfied:b is_iThe following formula is satisfied:

4. the algorithm of claim 2, wherein a is₀And b₀According to the rho_cAnd Ocs of i-1 th update_i-1The calculation specifically includes: calculating the a according to a slope estimation algorithm₀And b₀(ii) a Wherein the a is calculated according to the slope estimation algorithm₀And b₀The method comprises the following specific steps:

calculating a filtered signal matrix W_iW is as described_iThe following formula is satisfied:wherein, W₁=[ρ_c Ocs₁]，α₀Is the low pass filter bandwidth constant, Δ t is the time step for invoking the slope estimation algorithm;

according to the W_iCalculating an estimated parameter vectorThe above-mentionedThe following formula is satisfied: <math> <mrow> <msub> <mover> <mi>a</mi> <mo>^</mo> </mover> <mi>i</mi> </msub> <mo>=</mo> <msub> <mover> <mi>a</mi> <mo>^</mo> </mover> <mrow> <mi>i</mi> <mo>-</mo> <mn>1</mn> </mrow> </msub> <mo>+</mo> <msub> <mover> <mover> <mi>a</mi> <mo>^</mo> </mover> <mo>&CenterDot;</mo> </mover> <mi>i</mi> </msub> <mi>&Delta;t</mi> <mo>,</mo> </mrow> </math> wherein, <math> <mrow> <msub> <mover> <mi>a</mi> <mo>^</mo> </mover> <mn>1</mn> </msub> <mo>=</mo> <mfenced open='[' close=']'> <mtable> <mtr> <mtd> <msub> <mi>&alpha;</mi> <mn>0</mn> </msub> <mo>-</mo> <msub> <mi>a</mi> <mn>1</mn> </msub> </mtd> </mtr> <mtr> <mtd> <msub> <mi>b</mi> <mn>1</mn> </msub> </mtd> </mtr> </mtable> </mfenced> <mo>,</mo> </mrow> </math> <math> <mrow> <msub> <mover> <mover> <mi>a</mi> <mo>^</mo> </mover> <mo>&CenterDot;</mo> </mover> <mi>i</mi> </msub> <mo>=</mo> <mo>-</mo> <mi>P</mi> <msubsup> <mi>W</mi> <mi>i</mi> <mi>T</mi> </msubsup> <msub> <mi>e</mi> <mi>i</mi> </msub> <mo>,</mo> </mrow> </math> <math> <mrow> <msub> <mi>e</mi> <mi>i</mi> </msub> <mo>=</mo> <msub> <mi>W</mi> <mi>i</mi> </msub> <msub> <mover> <mi>a</mi> <mo>^</mo> </mover> <mrow> <mi>i</mi> <mo>-</mo> <mn>1</mn> </mrow> </msub> <mo>-</mo> <msub> <mi>&rho;</mi> <mi>c</mi> </msub> <mo>;</mo> </mrow> </math> the P is a preset gain matrix;

according to the aboveCalculating the a₀And b₀When the i-th driving MLB controller calculates the Ocs, the a₀Is equal to a_iSaid b is₀Is equal to b_i(ii) a Wherein, the a_iThe following formula is satisfied:b is_iThe following formula is satisfied:

5. the method of claim 1, wherein the SPSA controller is at a frequency f_SPSAUpdating a hysteresis value perturbation value Hys and a trigger time perturbation value TTT according to the updated HOF, HPP and CDR, and inputting the updated Hys and TTT to the MRO controller specifically comprises:

SPSA controller with frequency f_SPSAThe Hys is equal to Hys when the SPSA controller is driven for the j time_jSaid TTT is equal to TTT_j；

Then, when the SPSA controller is driven 3n +1 times, Hys and TTT are calculated, and satisfy the following equations:

<math> <mrow> <mi>TTT</mi> <mo>*</mo> <mo>=</mo> <msub> <mrow> <mi>TTT</mi> <mo>*</mo> </mrow> <mi>j</mi> </msub> <mo>=</mo> <msub> <mrow> <mi>TTT</mi> <mo>*</mo> </mrow> <mrow> <mn>3</mn> <mi>n</mi> <mo>+</mo> <mn>1</mn> </mrow> </msub> <mo>=</mo> <mi>TTT</mi> <mo>+</mo> <msub> <msubsup> <mi>c&Delta;</mi> <mi>k</mi> <mi>TTT</mi> </msubsup> <mi>n</mi> </msub> <mo>,</mo> </mrow> </math>

<math> <mrow> <mi>Hys</mi> <mo>*</mo> <mo>=</mo> <msub> <mrow> <mi>Hys</mi> <mo>*</mo> </mrow> <mi>j</mi> </msub> <mo>=</mo> <msub> <mrow> <mi>Hys</mi> <mo>*</mo> </mrow> <mrow> <mn>3</mn> <mi>n</mi> <mo>+</mo> <mn>1</mn> </mrow> </msub> <mo>=</mo> <mi>Hys</mi> <mo>+</mo> <msub> <msubsup> <mi>c&Delta;</mi> <mi>k</mi> <mi>KKK</mi> </msubsup> <mi>n</mi> </msub> <mo>;</mo> </mrow> </math> wherein n is an integer greater than or equal to 0, c is a preset arbitrary number, andandtwo independent variables which are randomly generated, obey binomial distribution and have the value of +/-1;

calculating loss parameters L +, TTT and Hys when the SPSA controller is driven for the 3n +2 times, wherein the L +, TTT and Hys satisfy the following formulas:

$L+=\frac{{\mathrm{HOFw}}_1+{\mathrm{HPPw}}_2+{\mathrm{CDRw}}_3}{w_1+w_2+w_3},$

<math> <mrow> <mi>TTT</mi> <mo>*</mo> <mo>=</mo> <msub> <mrow> <mi>TTT</mi> <mo>*</mo> </mrow> <mi>j</mi> </msub> <mo>=</mo> <msub> <mrow> <mi>TTT</mi> <mo>*</mo> </mrow> <mrow> <mn>3</mn> <mi>n</mi> <mo>+</mo> <mn>1</mn> </mrow> </msub> <mo>=</mo> <mi>TTT</mi> <mo>+</mo> <msub> <msubsup> <mi>c&Delta;</mi> <mi>k</mi> <mi>TTT</mi> </msubsup> <mi>n</mi> </msub> <mo>,</mo> </mrow> </math>

<math> <mrow> <mi>Hys</mi> <mo>*</mo> <mo>=</mo> <msub> <mrow> <mi>Hys</mi> <mo>*</mo> </mrow> <mi>j</mi> </msub> <mo>=</mo> <msub> <mrow> <mi>Hys</mi> <mo>*</mo> </mrow> <mrow> <mn>3</mn> <mi>n</mi> <mo>+</mo> <mn>1</mn> </mrow> </msub> <mo>=</mo> <mi>Hys</mi> <mo>+</mo> <msub> <msubsup> <mi>c&Delta;</mi> <mi>k</mi> <mi>KKK</mi> </msubsup> <mi>n</mi> </msub> <mo>;</mo> </mrow> </math>

calculating back loss parameters L-, Hys and TTT when the SPSA controller is driven for the 3n +3 times, wherein the L-, TTT and Hys satisfy the following formulas:

$L-=\frac{{\mathrm{HOFw}}_1+{\mathrm{HPPw}}_2+{\mathrm{CDRw}}_3}{w_1+w_2+w_3},$

$\mathrm{TTT}*={\mathrm{TTT}*}_j={\mathrm{TTT}*}_{3n+2}=\mathrm{TTT}-{{\mathrm{ag}}_k^{\mathrm{TTT}}}_n,$

$\mathrm{Hys}*={\mathrm{Hys}*}_j={\mathrm{Hys}*}_{3n+2}=\mathrm{Hys}-{{\mathrm{ag}}_k^{\mathrm{Hys}}}_n;$ wherein, theAndthe following formula is satisfied: <math> <mrow> <msub> <msubsup> <mi>ag</mi> <mi>k</mi> <mi>TTT</mi> </msubsup> <mi>n</mi> </msub> <mo>=</mo> <mfrac> <mrow> <mi>L</mi> <mo>+</mo> <mo>-</mo> <mi>L</mi> <mo>-</mo> </mrow> <msub> <msubsup> <mrow> <mn>2</mn> <mi>c&Delta;</mi> </mrow> <mi>k</mi> <mi>TTT</mi> </msubsup> <mi>n</mi> </msub> </mfrac> <mo>,</mo> </mrow> </math> <math> <mrow> <msub> <msubsup> <mi>ag</mi> <mi>k</mi> <mi>Hys</mi> </msubsup> <mi>n</mi> </msub> <mo>=</mo> <mfrac> <mrow> <mi>L</mi> <mo>+</mo> <mo>-</mo> <mi>L</mi> <mo>-</mo> </mrow> <msub> <msubsup> <mrow> <mn>2</mn> <mi>c&Delta;</mi> </mrow> <mi>k</mi> <mi>Hys</mi> </msubsup> <mi>n</mi> </msub> </mfrac> <mo>,</mo> </mrow> </math> a is a preset arbitrary number.

6. The method of claim 1, wherein the MRO controller is at a frequency f_MROUpdating hysteresis values Hys and optimal time-to-trigger TTT based on the updated HOF, HPP and CDR and the updated Hys and TTT_mroAnd inputting an updated TTT to the SAT_mroThe method specifically comprises the following steps:

at a frequency f_MROCalculating a gamma value from the Hys and TTT, the gamma value satisfying the following formula:

<math> <mrow> <mi>&gamma;</mi> <mo>=</mo> <mi>arctan</mi> <mrow> <mo>(</mo> <mfrac> <mrow> <mi>Hys</mi> <mo>*</mo> </mrow> <mrow> <mn>2</mn> <mi>TTT</mi> <mo>*</mo> </mrow> </mfrac> <mo>)</mo> </mrow> </mrow> </math>

if gamma is less than or equal to the preset lower limit value, Hys and TTT_mroThe following formula is satisfied:

TTT_mro=TTT*-y_mrohys = Hys ″; wherein, said y_mroAt a frequency f_MROCalculated according to the updated HOF, HPP and CDR, y_mroSatisfy the requirement ofw₁、w₂、w₃The parameter weight value is preset;

if gamma is greater than or equal to a preset lower limit value and less than or equal to a preset upper limit value, Hys and TTT_mroThe following formula is satisfied:

TTT_mro=TTT*-y_mro，Hys=Hys*-y_mro；

if gamma is greater than or equal to a preset upper limit value, Hys and TTT_mroThe following formula is satisfied:

TTT_mro=TTT*，Hys=Hys*-y_mro。

7. the method of claim 1 wherein the SAT is at a frequency f_SATAccording to the updated rho_cAnd the updated TTT_mroThe update trigger time TTT specifically includes:

at a frequency f_SATAccording to rho_cCalculating a value of l, the value of l satisfying the following formula:

l=1+(1-ρ_c)·15

if TTT_mroIf the value is less than the value of l, the TTT is equal to the TTT_mro；

If TTT_mroAnd if the value is larger than the value of l, the TTT is equal to l.

8. A base station, comprising:

a setting unit for setting the parameter update frequency f according to the frequency setting principle_dMLB controller driving frequency f for mobile load balancing_MLBOptimizing MRO controller drive frequency f by moving robustness_MROSAT driving frequency f of saturation threshold controller_SATAnd synchronous disturbance random approximation algorithm SPSA controller driving frequency f_SPSA；

A transmitting unit for transmitting at a frequency f_dInputting updated cell load ρ to the MLB controller and SAT_cInputting updated HOF, switching ping-pong effect HPP and call drop rate CDR to the MRO controller and the SPSA controller;

MLB controller for controlling at frequency f_MLBAccording to the updated rho_cUpdating the specific offset Ocs of the user to the serving cell;

SPSA controller for controlling at a frequency f_SPSAUpdating a hysteresis value perturbation value Hys and a trigger time perturbation value TTT according to the updated HOF, HPP and CDR, and inputting the updated Hys and TTT to the MRO controller;

MRO controller for controlling the frequency f_MROUpdating hysteresis values Hys and optimal time-to-trigger TTT based on the updated HOF, HPP and CDR and the updated Hys and TTT_mroAnd inputting an updated TTT to the SAT_mro；

SAT for use at frequency f_SATAccording to said updateρ_cAnd the updated TTT_mroUpdating the time to trigger TTT;

the frequency setting principle comprises: f is_MLBIs less than or equal to f_dSaid f_SATIs equal to f_dSaid f_MROIs greater than f_SPSASaid f_SPSAF or more_HPI(ii) a Wherein, the f_HPIThe time for the performance of each controller to reach the stable state after the parameters Ocs, Hys and TTT are adjusted is the reciprocal of the time.

9. The base station of claim 8, wherein the MLB controller is specifically configured to:

at a frequency f_MLBDriving an MLB controller;

when the MLB controller is driven for the ith time, the Ocs is updated to Ocs_i；

When i is equal to 1, the Ocs is a preset initial value Ocs₁；

When i is an integer greater than 1, the Ocs_iThe following formula is satisfied:

<math> <mrow> <msub> <mi>Ocs</mi> <mi>i</mi> </msub> <mo>=</mo> <mrow> <mo>(</mo> <mi>r</mi> <mo>-</mo> <msub> <mi>&rho;</mi> <mi>c</mi> </msub> <mo>)</mo> </mrow> <msub> <mi>K</mi> <mi>P</mi> </msub> <mo>+</mo> <mfrac> <msub> <mi>a</mi> <mn>0</mn> </msub> <msub> <mi>b</mi> <mn>0</mn> </msub> </mfrac> <mi>r</mi> <mo>,</mo> </mrow> </math> <math> <mrow> <msub> <mi>K</mi> <mi>p</mi> </msub> <mo>=</mo> <msqrt> <mfrac> <msubsup> <mi>a</mi> <mn>0</mn> <mn>2</mn> </msubsup> <msubsup> <mi>b</mi> <mn>0</mn> <mn>2</mn> </msubsup> </mfrac> <mo>+</mo> <mfrac> <mi>Q</mi> <mi>&rho;R</mi> </mfrac> </msqrt> <mo>-</mo> <mfrac> <msub> <mi>a</mi> <mn>0</mn> </msub> <msub> <mi>b</mi> <mn>0</mn> </msub> </mfrac> <mo>;</mo> </mrow> </math>

wherein r is a preset serving cell load target value; rho, Q and R are preset constants, a₀And b₀According to the updated rho_cAnd Ocs of i-1 th update_i-1Calculating to obtain;

wherein, the a₀And b₀According to the updated rho_cAnd Ocs of i-1 th update_i-1The obtaining of the calculation specifically comprises calculating the a according to a BGF algorithm₀And b₀The method comprises the following specific steps:

calculating a filtered signal matrix W_iW is as described_iThe following formula is satisfied:wherein, W₁=[ρ_c Ocs₁]，α₀Is the bandwidth constant of the low-pass filter, and Δ t is the time step for calling the BGF algorithm;

according to the W_iCalculating a gain matrix P_iSaid P is_iThe following formula is satisfied:wherein, P₁Is a preset initial value P₀；Is P_iThe derivative with respect to time t satisfies the following equation:λ_ias a forgetting factor, the following formula is satisfied:||P_i||₂the following formula is satisfied for the Euclidean norm of the gain matrix: is thatA characteristic value of (d);

according to the W_iAnd P_iCalculating an estimated parameter vectorThe above-mentionedThe following formula is satisfied: <math> <mrow> <msub> <mover> <mi>a</mi> <mo>^</mo> </mover> <mi>i</mi> </msub> <mo>=</mo> <msub> <mover> <mi>a</mi> <mo>^</mo> </mover> <mrow> <mi>i</mi> <mo>-</mo> <mn>1</mn> </mrow> </msub> <mo>+</mo> <msub> <mover> <mover> <mi>a</mi> <mo>^</mo> </mover> <mo>&CenterDot;</mo> </mover> <mi>i</mi> </msub> <mi>&Delta;t</mi> <mo>,</mo> </mrow> </math> wherein, <math> <mrow> <msub> <mover> <mi>a</mi> <mo>^</mo> </mover> <mn>1</mn> </msub> <mo>=</mo> <mfenced open='[' close=']'> <mtable> <mtr> <mtd> <msub> <mi>&alpha;</mi> <mn>0</mn> </msub> <mo>-</mo> <msub> <mi>a</mi> <mn>1</mn> </msub> </mtd> </mtr> <mtr> <mtd> <msub> <mi>b</mi> <mn>1</mn> </msub> </mtd> </mtr> </mtable> </mfenced> <mo>,</mo> </mrow> </math> <math> <mrow> <msub> <mover> <mover> <mi>a</mi> <mo>^</mo> </mover> <mo>&CenterDot;</mo> </mover> <mi>i</mi> </msub> <mo>=</mo> <mo>-</mo> <msub> <mi>P</mi> <mi>i</mi> </msub> <msubsup> <mi>W</mi> <mi>i</mi> <mi>T</mi> </msubsup> <msub> <mi>e</mi> <mi>i</mi> </msub> <mo>,</mo> </mrow> </math> <math> <mrow> <msub> <mi>e</mi> <mi>i</mi> </msub> <mo>=</mo> <msub> <mi>W</mi> <mi>i</mi> </msub> <msub> <mover> <mi>a</mi> <mo>^</mo> </mover> <mrow> <mi>i</mi> <mo>-</mo> <mn>1</mn> </mrow> </msub> <mo>-</mo> <msub> <mi>&rho;</mi> <mi>c</mi> </msub> <mo>;</mo> </mrow> </math>

according to the aboveCalculating the a₀And b₀When the i-th driving MLB controller calculates the Ocs, the a₀Is equal to a_iSaid b is₀Is equal to b_i(ii) a Wherein, the a_iThe following formula is satisfied:b is_iThe following formula is satisfied:

or, calculating the a according to the slope estimation algorithm₀And b₀The method comprises the following specific steps:

calculating a filtered signal matrix W_iW is as described_iThe following formula is satisfied:wherein, W₁=[ρ_c Ocs₁]，α₀Is the bandwidth constant of the low-pass filter, and Δ t is the time step for calling the BGF algorithm;

according to the W_iCalculating an estimated parameter vectorThe above-mentionedThe following formula is satisfied: <math> <mrow> <msub> <mover> <mi>a</mi> <mo>^</mo> </mover> <mi>i</mi> </msub> <mo>=</mo> <msub> <mover> <mi>a</mi> <mo>^</mo> </mover> <mrow> <mi>i</mi> <mo>-</mo> <mn>1</mn> </mrow> </msub> <mo>+</mo> <msub> <mover> <mover> <mi>a</mi> <mo>^</mo> </mover> <mo>&CenterDot;</mo> </mover> <mi>i</mi> </msub> <mi>&Delta;t</mi> <mo>,</mo> </mrow> </math> wherein, <math> <mrow> <msub> <mover> <mi>a</mi> <mo>^</mo> </mover> <mn>1</mn> </msub> <mo>=</mo> <mfenced open='[' close=']'> <mtable> <mtr> <mtd> <msub> <mi>&alpha;</mi> <mn>0</mn> </msub> <mo>-</mo> <msub> <mi>a</mi> <mn>1</mn> </msub> </mtd> </mtr> <mtr> <mtd> <msub> <mi>b</mi> <mn>1</mn> </msub> </mtd> </mtr> </mtable> </mfenced> <mo>,</mo> </mrow> </math> <math> <mrow> <msub> <mover> <mover> <mi>a</mi> <mo>^</mo> </mover> <mo>&CenterDot;</mo> </mover> <mi>i</mi> </msub> <mo>=</mo> <mo>-</mo> <mi>P</mi> <msubsup> <mi>W</mi> <mi>i</mi> <mi>T</mi> </msubsup> <msub> <mi>e</mi> <mi>i</mi> </msub> <mo>,</mo> </mrow> </math> <math> <mrow> <msub> <mi>e</mi> <mi>i</mi> </msub> <mo>=</mo> <msub> <mi>W</mi> <mi>i</mi> </msub> <msub> <mover> <mi>a</mi> <mo>^</mo> </mover> <mrow> <mi>i</mi> <mo>-</mo> <mn>1</mn> </mrow> </msub> <mo>-</mo> <msub> <mi>&rho;</mi> <mi>c</mi> </msub> <mo>,</mo> </mrow> </math> the P is a preset gain matrix;

according to the aboveCalculating the a₀And b₀When the i-th driving MLB controller calculates the Ocs, the a₀Is equal to a_iSaid b is₀Is equal to b_i(ii) a Wherein, the a_iThe following formula is satisfied:b is_iThe following formula is satisfied:

10. the base station of claim 8, wherein the SPSA controller is specifically configured to:

SPSA controller with frequency f_SPSAThe Hys is equal to Hys when the SPSA controller is driven for the j time_jSaid TTT is equal to TTT_j；

Then, when the SPSA controller is driven 3n +1 times, Hys and TTT are calculated, and satisfy the following equations:

<math> <mrow> <mi>TTT</mi> <mo>*</mo> <mo>=</mo> <msub> <mrow> <mi>TTT</mi> <mo>*</mo> </mrow> <mi>j</mi> </msub> <mo>=</mo> <msub> <mrow> <mi>TTT</mi> <mo>*</mo> </mrow> <mrow> <mn>3</mn> <mi>n</mi> <mo>+</mo> <mn>1</mn> </mrow> </msub> <mo>=</mo> <mi>TTT</mi> <mo>+</mo> <msub> <msubsup> <mi>c&Delta;</mi> <mi>k</mi> <mi>TTT</mi> </msubsup> <mi>n</mi> </msub> <mo>,</mo> </mrow> </math>

<math> <mrow> <mi>Hys</mi> <mo>*</mo> <mo>=</mo> <msub> <mrow> <mi>Hys</mi> <mo>*</mo> </mrow> <mi>j</mi> </msub> <mo>=</mo> <msub> <mrow> <mi>Hys</mi> <mo>*</mo> </mrow> <mrow> <mn>3</mn> <mi>n</mi> <mo>+</mo> <mn>1</mn> </mrow> </msub> <mo>=</mo> <mi>Hys</mi> <mo>+</mo> <msub> <msubsup> <mi>c&Delta;</mi> <mi>k</mi> <mi>KKK</mi> </msubsup> <mi>n</mi> </msub> <mo>;</mo> </mrow> </math> wherein n is an integer greater than or equal to 0, c is a preset arbitrary number, andandtwo independent variables which are randomly generated, obey binomial distribution and have the value of +/-1;

calculating loss parameters L +, TTT and Hys when the SPSA controller is driven for the 3n +2 times, wherein the L +, TTT and Hys satisfy the following formulas:

$L+=\frac{{\mathrm{HOFw}}_1+{\mathrm{HPPw}}_2+{\mathrm{CDRw}}_3}{w_1+w_2+w_3},$

<math> <mrow> <mi>TTT</mi> <mo>*</mo> <mo>=</mo> <msub> <mrow> <mi>TTT</mi> <mo>*</mo> </mrow> <mi>j</mi> </msub> <mo>=</mo> <msub> <mrow> <mi>TTT</mi> <mo>*</mo> </mrow> <mrow> <mn>3</mn> <mi>n</mi> <mo>+</mo> <mn>1</mn> </mrow> </msub> <mo>=</mo> <mi>TTT</mi> <mo>+</mo> <msub> <msubsup> <mi>c&Delta;</mi> <mi>k</mi> <mi>TTT</mi> </msubsup> <mi>n</mi> </msub> <mo>,</mo> </mrow> </math>

<math> <mrow> <mi>Hys</mi> <mo>*</mo> <mo>=</mo> <msub> <mrow> <mi>Hys</mi> <mo>*</mo> </mrow> <mi>j</mi> </msub> <mo>=</mo> <msub> <mrow> <mi>Hys</mi> <mo>*</mo> </mrow> <mrow> <mn>3</mn> <mi>n</mi> <mo>+</mo> <mn>1</mn> </mrow> </msub> <mo>=</mo> <mi>Hys</mi> <mo>+</mo> <msub> <msubsup> <mi>c&Delta;</mi> <mi>k</mi> <mi>KKK</mi> </msubsup> <mi>n</mi> </msub> <mo>;</mo> </mrow> </math>

calculating back loss parameters L-, Hys and TTT when the SPSA controller is driven for the 3n +3 times, wherein the L-, TTT and Hys satisfy the following formulas:

$L-=\frac{{\mathrm{HOFw}}_1+{\mathrm{HPPw}}_2+{\mathrm{CDRw}}_3}{w_1+w_2+w_3},$

$\mathrm{TTT}*={\mathrm{TTT}*}_j={\mathrm{TTT}*}_{3n+2}=\mathrm{TTT}-{{\mathrm{ag}}_k^{\mathrm{TTT}}}_n,$

$\mathrm{Hys}*={\mathrm{Hys}*}_j={\mathrm{Hys}*}_{3n+2}=\mathrm{Hys}-{{\mathrm{ag}}_k^{\mathrm{Hys}}}_n;$ wherein, theAndthe following formula is satisfied: <math> <mrow> <msub> <msubsup> <mi>ag</mi> <mi>k</mi> <mi>TTT</mi> </msubsup> <mi>n</mi> </msub> <mo>=</mo> <mfrac> <mrow> <mi>L</mi> <mo>+</mo> <mo>-</mo> <mi>L</mi> <mo>-</mo> </mrow> <msub> <msubsup> <mrow> <mn>2</mn> <mi>c&Delta;</mi> </mrow> <mi>k</mi> <mi>TTT</mi> </msubsup> <mi>n</mi> </msub> </mfrac> <mo>,</mo> </mrow> </math> <math> <mrow> <msub> <msubsup> <mi>ag</mi> <mi>k</mi> <mi>Hys</mi> </msubsup> <mi>n</mi> </msub> <mo>=</mo> <mfrac> <mrow> <mi>L</mi> <mo>+</mo> <mo>-</mo> <mi>L</mi> <mo>-</mo> </mrow> <msub> <msubsup> <mrow> <mn>2</mn> <mi>c&Delta;</mi> </mrow> <mi>k</mi> <mi>Hys</mi> </msubsup> <mi>n</mi> </msub> </mfrac> <mo>,</mo> </mrow> </math> a is a preset arbitrary number.

11. The base station of claim 8, wherein the MRO controller is specifically configured to:

at a frequency f_MROCalculating a gamma value from the Hys and TTT, the gamma value satisfying the following formula: <math> <mrow> <mi>&gamma;</mi> <mo>=</mo> <mi>arctan</mi> <mrow> <mo>(</mo> <mfrac> <mrow> <mi>Hys</mi> <mo>*</mo> </mrow> <mrow> <mn>2</mn> <mi>TTT</mi> <mo>*</mo> </mrow> </mfrac> <mo>)</mo> </mrow> </mrow> </math>

if gamma is less than or equal to the preset lower limit value, Hys and TTT_mroThe following formula is satisfied:

TTT_mro=TTT*-y_mrohys = Hys ″; wherein, said y_mroAt a frequency f_MROCalculated according to the updated HOF, HPP and CDR, y_mroSatisfy the requirement ofw₁、w₂、w₃The parameter weight value is preset;

if gamma is greater than or equal to a preset lower limit value and less than or equal to a preset upper limit value, Hys and TTT_mroThe following formula is satisfied:

TTT_mro=TTT*-y_mro，Hys=Hys*-y_mro；

if gamma is greater thanEqual to the preset upper limit value, Hys and TTT_mroThe following formula is satisfied:

TTT_mro=TTT*，Hys=Hys*-y_mro。

12. the base station of claim 8, wherein the SAT is specifically configured for

At a frequency f_SATAccording to rho_cCalculating a value of l, the value of l satisfying the following formula:

l=1+(1-ρ_c)·15

if TTT_mroIf the value is less than the value of l, the TTT is equal to the TTT_mro；

If TTT_mroAnd if the value is larger than the value of l, the TTT is equal to l.