# US 2017/0288778 A1 · Method and device for determining location of optical splitter

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

- Number: US 2017/0288778 A1
- Document ID: US20170288778A1
- Status: Published application
- Kind: A1
- Application number: 15/508,978
- Filed: 2015-11-02
- Published: 2017-10-05
- Earliest claimed: 2015-09-01
- Assignees: Soochow University
- Applicants: Soochow University
- Inventors: Gangxiang SHEN; Hao CHEN; Yongcheng LI
- Web page: https://patentgrep.com/patent/US20170288778A1

## CPC

- H04B 10/27: Arrangements for networking
- H04B 10/07: Arrangements for monitoring or testing transmission systems; Arrangements for fault measurement of transmission systems
- H04B 10/25891: {Transmission components (H04B10/40 takes precedence)}
- H04B 10/272: Star-type networks {or tree-type networks}
- H04Q 11/0067: {Provisions for optical access or distribution networks, e.g. Gigabit Ethernet Passive Optical Network (GE-PON), ATM-based Passive Optical Network (A-PON), PON-Ring}
- H04B 10/00: Transmission systems employing electromagnetic waves other than radio-waves, e.g. infrared, visible or ultraviolet light, or employing corpuscular radiation, e.g. quantum communication
- H04B: TRANSMISSION
- H04Q: SELECTING (switches, relays, selectors H01H; wireless communication networks H04W)

## Abstract

A method and a device for determining a location of an optical splitter are provided. With the method, optical network units in a PON network are classified into K clusters based on locations of the optical network units with a K-means clustering-based algorithm. Distances between multiple optical network units in the same cluster are small. If it is determined that the number of the optical network units in each of the clusters does not exceed a threshold, a central office is added into the K clusters to obtain K new clusters; and for each of the new clusters, a location of an optical splitter corresponding to the new cluster is determined, so as to ensure that a sum of distances between the location of the optical splitter and locations of all elements in the new cluster is minimized.

## Claims

### Claim 1

A method for determining a location of an optical splitter, comprising:
inputting a location of a central office and locations of optical network units in a PON network;
classifying, with a K-means clustering-based algorithm, the optical network units into K clusters based on the locations of the optical network units;
adding the central office to the K clusters to obtain K new clusters if it is determined that the number of the optical network units in each of the K clusters does not exceed a threshold; and
determining, for each of the K new clusters, a location of an optical splitter corresponding to the new cluster, wherein a sum of distances between the location of the optical splitter and the locations of the central office and the optical network units in the new cluster is minimized.

### Claim 2 (depends on 1)

The method according to claim 1, wherein if it is determined that there is a cluster having the number of optical network units exceeding the threshold among the K clusters, the method further comprises:
determining the cluster having the number of optical network units exceeding the threshold as a target cluster;
classifying, with the K-means clustering-based algorithm, the optical network units in the target cluster into a plurality of sub-clusters based on locations of the optical network units in the target cluster; and
detecting whether there is a sub-cluster having the number of optical network units exceeding the threshold; and returning to the process of determining the cluster having the number of optical network units exceeding the threshold as a target cluster, if it is detected that there is a sub-cluster having the number of optical network units exceeding the threshold.

### Claim 3 (depends on 1)

The method according to claim 1, wherein the determining a location of an optical splitter corresponding to the new cluster comprises:
determining, with the Weiszfeld algorithm, the location of the optical splitter corresponding to the new cluster based on the locations of the central office and the optical network units in the new cluster.

### Claim 4 (depends on 1)

The method according to claim 1, wherein in the classifying with a K-means clustering-based algorithm the optical network units into K clusters, a value of K is determined by:
creating a curve by taking the value of K as an argument on an X axis and taking a clustering metric value as a dependent variable on a Y axis, wherein the clustering metric value is defined as an average value of clustering index values of the K clusters classified with the K-means clustering-based algorithm; and
selecting a value of K corresponding to a point in the curve having a maximum rate of change in the clustering metric value, as a target value of K.

### Claim 5 (depends on 2)

The method according to claim 2, further comprising:
constituting a target set by the central office and the optical splitters corresponding to the sub-clusters obtained by dividing the target cluster;
determining, with the Weiszfeld algorithm, a location of a target point corresponding to the target set, wherein a sum of distances between the location of the target point and locations of all elements in the target set is minimized; and
determining a route from the location of the central office to the location of the target point as a target fiber conduit route, wherein the target fiber conduit route is shared by a plurality of optical fibers arranged between the central office and the optical splitters in the target set.

### Claim 6

A device for determining a location of an optical splitter, comprising:
a data receiving unit configured to receive a location of a central office and locations of optical network units in a PON network;
a cluster classifying unit configured to classify, with a K-means clustering-based algorithm, the optical network units into K clusters based on the locations of the optical network units;
a new cluster determining unit configured to add the central office to the K clusters to obtain K new clusters if it is determined that the number of the optical network units in each of the K clusters does not exceed a threshold; and
an optical splitter location calculating unit configured to determine, for each of the new clusters, a location of an optical splitter corresponding to the new cluster, wherein a sum of distances between the location of the optical splitter and locations of the central office and the optical network units in the new cluster is minimized.

### Claim 7 (depends on 6)

The device according to claim 6, further comprising:
a target cluster determining unit configured to determine, if it is determined that there is a cluster having the number of optical network units exceeding the threshold among the K clusters, the cluster having the number of optical network units exceeding the threshold as a target cluster;
a target cluster dividing unit configured to classify, with the K-means clustering-based algorithm, the optical network units in the target cluster into a plurality of sub-clusters based on locations of the optical network units in the target cluster; and
a sub-cluster judging unit configured to detect whether there is a sub-cluster having the number of optical network units exceeding the threshold; and return to perform the operation of the target cluster determining unit if it is detected that there is a sub-cluster having the number of optical network units exceeding the threshold.

### Claim 8 (depends on 6)

The device according to claim 6, wherein the optical splitter location calculating unit comprises:
a first optical splitter location calculating sub-unit configured to determine, with the Weiszfeld algorithm, a location of an optical splitter corresponding to the new cluster based on the locations of the central office and the optical network units in the new cluster.

### Claim 9 (depends on 6)

The device according to claim 6, wherein the cluster classifying unit comprises:
a curve creating unit configured to create a curve by taking a value of K as an argument on an X axis and taking a clustering metric value as a dependent variable on a Y axis, wherein the clustering metric value is defined as an average value of clustering index values of the K clusters classified with the K-means clustering-based algorithm; and
a K value selecting unit configured to select a value of K corresponding to a point in the curve having a maximum rate of change in the clustering metric value, as a target value of K.

### Claim 10 (depends on 7)

The device according to claim 7, further comprising:
a target set constituting unit configured to constitute a target set by the central office and the optical splitters corresponding to the sub-clusters obtained by dividing the target cluster;
a target point location determining unit configured to determine, with the Weiszfeld algorithm, a location of a target point corresponding to the target set, wherein a sum of distances between the location of the target point and locations of all elements in the target set is minimized; and
a target fiber conduit route determining unit configured to determine a route from the location of the central office to the location of the target point as a target fiber conduit route, wherein the target fiber conduit route is shared by a plurality of optical fibers arranged between the central office and the optical splitters in the target set.