# US 2018/0324712 A1 · Data driven management method and device of small cell network

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2018/0324712 A1
- Document ID: US20180324712A1
- Status: Published application
- Kind: A1
- Application number: 15/969,357
- Filed: 2018-05-02
- Published: 2018-11-08
- Earliest claimed: 2017-05-08
- Assignees: National Chiao Tung University
- Applicants: National Chiao Tung University
- Inventors: LI-CHUN WANG; SHAO-HUNG CHENG
- Web page: https://patentgrep.com/patent/US20180324712A1

## CPC

- H04W 52/244: {Interferences in heterogeneous networks, e.g. among macro and femto or pico cells or other sector / system interference [OSI]}
- H04W 16/22: Traffic simulation tools or models
- H04W 24/02: Arrangements for optimising operational condition
- H04W 52/245: {taking into account received signal strength}
- H04W 84/045: {using private Base Stations, e.g. femto Base Stations, home Node B}
- H04W 84/10: Small scale networks; Flat hierarchical networks
- H04W: WIRELESS COMMUNICATION NETWORKS (broadcast communication H04H; communication systems using wireless links for non-selective communication, e.g. wireless extensions H04M1/72)

## Abstract

A data driven management device of a small cell network is provided, wherein the device includes a data receiver, a data processor and a data configurator. The data receiver receives parameters configured and managed from each of small cells. The data processor is configured to pre-process the received parameters to remove the unnecessary parameters and to calculate at least one performance result based on the pre-processed parameters. The data processor is further configured to establish a model based on at least one of the pre-processed parameters. The data processor determines the optimized parameter based on the best performance result of the model to determine a decision result. The data configurator is connected to the data processor and configures and manages the small cells based on the decision result.

## Claims

### Claim 1

A data driven management device of a small cell network, wherein the data driven management device manages ultra-dense small cells within an network system and comprises:
a data receiver receiving configured and managed parameters from each of the small cells;
a data processor connected to the data receiver, comprising:
a data pre-processing module configured to pre-process the received parameters to remove unnecessary parameters and calculate at least one performance result based on the pre-processed parameters;
a performance predicting model establishing module configured to establish a model based on at least one of the pre-processed parameters; and
a self-optimizing parameter control module configured to determine an optimized parameter based on a best performance result of the performance predicting model to output a decision result; and
a data configurator connected to the data processor, and configuring and managing each of the small cells based on the decision result.

### Claim 2 (depends on 1)

The data driven management device of claim 1, wherein:
the at least one performance result comprises an energy efficiency parameter and a system throughput; and
the performance predicting model establishing module and the self-optimizing parameter control module are initiated when the energy efficiency is lower than or equal to a predetermined threshold.

### Claim 3 (depends on 2)

The data driven management device of claim 2, wherein:
the performance predicting model establishing module further detects the small cells which are not in service to any user devices; and
the data processor shuts down the small cells which are not in service to any of the user devices.

### Claim 4 (depends on 3)

The data driven management device of claim 3, wherein:
the performance predicting model establishing module calculates a reference index value of the active small cells and sorts the active small cells based on the reference index value; and
the performance predicting model establishing module selects and lowers power of the small cells which are corresponded to the reference index values ranked in top k in which k is a value from 1 to Q-S, to calculate the system throughput which is corresponded to each of the k value and obtain the performance predicting model of each of the k values and the system throughput which are corresponded thereto by a regression analysis method, wherein Q is a total number of the small cells, S is a total number of the small cells which are not in service to any of the user devices.

### Claim 5 (depends on 4)

The data driven management device of claim 4, wherein the self-optimizing parameter control module obtains an optimized k* value which is corresponded to the best system throughput within the performance predicting model and determines to lower the power of the sorted small cells ranked in top k*.

### Claim 6 (depends on 4)

The data driven management device of claim 4, wherein the reference index value of each of the active small cells is a reciprocal of a total signal strength value of the small cells, a total interference value, or a ratio of the interference value and a load sensing value obtained according to a ratio of the total interference value and the total signal strength value.

### Claim 7 (depends on 6)

The data driven management device of claim 6, wherein the total signal strength value of the each small cells is a sum of the received signal strength from the each user device within a coverage range of the cell; and the total interference value of the each small cells is the sum of the received interference signal from each served user device by the cell.

### Claim 8 (depends on 4)

The data driven management device of claim 4, wherein the regression analysis method is polynomial regression analysis method; and the small cell is a femtocell or a pico cell.

### Claim 9

A data driven management method of a small cell network for configuring and managing each of small cells within a network system, comprising:
receiving parameters configured and managed by each of the small cells;
pre-processing the received parameters to remove unnecessary parameters, and calculating at least one performance result based on the pre-processed parameters;
detecting each of the small cells which are not in service to any user devices;
calculating a reference index value of the active small cell, and sorting the active small cell based on the reference index value;
selecting and reducing power of the small cells which are corresponded to the reference indexes ranked in top k in which k is a value from 1 to Q-S, to calculate a system throughput which is corresponded to each of the k value and obtain a performance predicting model of each of the k values and the system throughput which are corresponded thereto by a regression analysis method, wherein Q is a total number of the small cells, S is a total number of the small cells which are not in service to any of the user devices; and
obtaining an optimized k* value which is corresponded to the best system throughput within the performance predicting model and determining to lower the power of the sorted small cells ranked in top k*.

### Claim 10 (depends on 9)

The data driven cell management method of claim 9, wherein the reference index value of each of the active small cells is a reciprocal of a total signal strength value of the small cells, a total interference value or a ratio of the interference value and a load sensing value obtained according to a ratio of the total reference value and the total signal strength value; the total signal strength value of the each small cells is a sum of the received signal strength from each user device within a coverage range of the cell; and the total interference value of the each small cells is a sum of the received interference signal from each served user device by the cell.
