# US 2013/0210479 A1 · Method and apparatus for improving radio frequency index in real time in distributed network

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

- Number: US 2013/0210479 A1
- Document ID: US20130210479A1
- Status: Published application
- Kind: A1
- Application number: 13/878,679
- Filed: 2011-06-16
- Published: 2013-08-15
- Earliest claimed: 2010-10-11
- Assignees: Zte Corporation
- Inventors: Xiajian Tang; Bojing Yang
- Web page: https://patentgrep.com/patent/US20130210479A1

## CPC

- H04W 52/52: using AGC [Automatic Gain Control] circuits or amplifiers
- H04W 52/143: {Downlink power control}
- H04B 17/13: of power amplifiers, e.g. gain or non-linearity
- H04B 17/373: Predicting channel quality {or other radio frequency [RF]} parameters
- H04B 17/3913: {Predictive models, e.g. based on neural network models}
- H04W 52/223: {predicting future states of the transmission}
- H04W: WIRELESS COMMUNICATION NETWORKS (broadcast communication H04H; communication systems using wireless links for non-selective communication, e.g. wireless extensions H04M1/72)
- H04B: TRANSMISSION

## Abstract

A real-time improvement method and a real-time improvement apparatus for a distributed network radio frequency performance are provided by the present invention. The method comprises: a data analysis unit receiving an input power, an output power and a link gain from a base station element node with a radio frequency module, and establishing a learning model of the base station element node according to the input power, the output power and the link gain; the data analysis unit obtaining an expected output power corresponding to the input power according to a correlation between a preset input power and the expected output power; in the learning model, regulating the link gain according to a difference value between the expected output power and the output power, until the difference value between the output power of the learning model and the expected output power reaches a preset range, and determining the link gain at this moment as a new link gain; and the data analysis unit sending the new link gain to the base station element node with a radio frequency module, and the base station element node regulating the link gain thereof according to the new link gain.

## Claims

### Claim 1

A real-time improvement method for a distributed network radio frequency performance, comprising the steps of:
a data analysis unit receiving an input power, an output power and a link gain from a base station element node with a radio frequency module, and establishing a learning model of the base station element node according to the input power, the output power and the link gain;
the data analysis unit obtaining an expected output power corresponding to the input power according to a correlation between a preset input power and the expected output power;
in the learning model, regulating the link gain according to a difference value between the expected output power and the output power, until the difference value between the output power of the learning model and the expected output power reaches a preset range, and determining the link gain at this moment as a new link gain; and
the data analysis unit sending the new link gain to the base station element node with a radio frequency module, and the base station element node regulating a link gain of the base station element node according to the new link gain.

### Claim 2 (depends on 1)

The real-time improvement method for a distributed network radio frequency performance according to claim 1, wherein the base station element node with the radio frequency module is a Radio Remote Unit (RRU) comprising a detection unit and a gain regulation unit; and
establishing the learning model of the base station element node comprises: the RRU detecting and collecting the input power, the output power and the link gain by way of the detection unit and sending the input power, the output power and the link gain to the data analysis unit; and the data analysis unit establishing the learning model of the RRU comprising an input layer, an intermediate layer and an output layer according to the input power, the output power and the link gain of the RRU.

### Claim 3 (depends on 2)

The real-time improvement method for a distributed network radio frequency performance according to claim 2, wherein determining the new link gain comprises:
if the difference value between the output power and the expected output power in the learning model is not within a preset range, then regulating the link gain in the learning model and updating the node weights of the input layer, the intermediate layer and the output layer in the learning model, until the difference value between the output power and the expected output power in the learning model is within the preset range, and determining the link gain at this moment as the new link gain.

### Claim 4 (depends on 3)

The real-time improvement method for a distributed network radio frequency performance according to claim 3, wherein regulating the link gain of the base station element node according to the new link gain comprises:
the data analysis unit sending the determined new link gain to the RRU, and the RRU initiating a gain regulation unit to regulate a downlink radio frequency gain according to a value of the new link gain.

### Claim 5 (depends on 1)

The real-time improvement method for a distributed network radio frequency performance according to claim 1, further comprising:
the data analysis unit collecting connection location relationship information and node production batch information about the base station element node with the radio frequency module.

### Claim 6 (depends on 5)

The real-time improvement method for a distributed network radio frequency performance according to claim 5, after determining the new link gain, further comprising:
the base station element node with the radio frequency module regulating the link gain according to the new link gain; and
the data analysis unit receiving a new output power of the base station element node, and comparing and analyzing a difference value between the new output power and the expected output power, and if the difference value is within a preset range, then the data analysis unit updating power sample information and a corresponding link gain parameter of the base station element node stored by the base station according to the connection location relationship about the base station element node.

### Claim 7 (depends on 6)

The real-time improvement method for a distributed network radio frequency performance according to claim 6, after if the difference value is within the preset range, then updating the power sample information and the corresponding link gain parameter of the base station element node stored by the base station according to the connection location relationship about the base station element node, further comprising:
a counting unit performing quality tracking on the base station element node by counting a change situation of the link gain parameter of the base station element node in each production batch.

### Claim 8

A real-time improvement apparatus for a distributed network radio frequency performance, comprising: a data analysis unit comprising a receiving module, a determination module, a regulation module and a sending module, wherein
the receiving module is configured to receive an input power, an output power and a link gain from a base station element node with a radio frequency module, and establish a learning model of the base station element node according to the input power, the output power and the link gain;
the determination module is configured to obtain an expected output power corresponding to the input power received by the receiving module according to a correlation between a preset input power and the expected output power;
the regulation unit, is configured to, in the learning model established by the receiving module, regulate the link gain according to a difference value between the expected output power and the output power obtained by the determination module, until the difference value between the output power of the learning model and the expected output power reaches a preset range, and the link gain at this moment is determined as a new link gain; and
the sending module is configured to send the new link gain determined by the regulation unit to the base station element node with a radio frequency module, which regulates a link gain of the base station element node according to the new link gain.

### Claim 9 (depends on 8)

The real-time improvement apparatus for a distributed network radio frequency performance according to claim 8, further comprising:
a counting unit, configured to perform quality tracking on the base station element node by counting a change situation of the link gain parameter of the base station element node in each production batch.
