# US 11,122,446 B2 · Method and system for managing a plurality of Wi-Fi access points considering backhauls and energy consumption using a cloud based adaptive software defined network

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

- Number: US 11,122,446 B2
- Document ID: US11122446B2
- Status: Granted
- Kind: B2
- Application number: 15/937,820
- Filed: 2018-03-27
- Published: 2021-04-29
- Granted: 2021-09-14
- Earliest claimed: 2017-03-27
- Assignees: Ambeent Inc.
- Applicants: Ambeent Wireless Bilişim ve Yazilim A.Ş
- Inventors: Mustafa Ergen; Hasan Şildir; Sinem Çöleri Ergen
- Web page: https://patentgrep.com/patent/US11122446B2

## CPC

- H04W 24/02: Arrangements for optimising operational condition
- H04W 52/267: {taking into account the information rate}
- G06N 5/01: Dynamic search techniques; Heuristics; Dynamic trees; Branch-and-bound
- G06N 20/00: Machine learning
- H04W 16/18: Network planning tools
- H04W 24/08: Testing, {supervising or monitoring} using real traffic
- H04W 28/06: Optimizing {the usage of the radio link}, e.g. header compression, information sizing {, discarding information (system modifying transmission characteristic according to link quality by modifying frame length H04L1/0007; dynamic adaptation of the packet size for flow control or congestion control H04L47/365)}
- H04W 52/245: {taking into account received signal strength}
- H04W 88/08: Access point devices
- H04W: WIRELESS COMMUNICATION NETWORKS (broadcast communication H04H; communication systems using wireless links for non-selective communication, e.g. wireless extensions H04M1/72)
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS

## Abstract

The invention provides a method and system for managing a plurality of Wi-Fi access points using a cloud based adaptive software defined network (A-SDN). The measurement data corresponding to a plurality of network variables are collected from one or more Wi-Fi access points and one or more client devices. Once the measurement data are collected, optimal operating channels and optimal transmit power for the one or more Wi-Fi access points are derived based on one or more network variables, backhaul limits corresponding to the one or more Wi-Fi access points and a download behavior of the one or more client devices using a real-time optimization. The real-time optimization includes deriving the optimal settings through heuristic solutions of an optimization problem formulated as a Mixed Integer Nonlinear Programming (MINLP) and a machine learning based optimization.

## Claims

### Claim 1

A method for managing a plurality of Wi-Fi access points using a cloud based adaptive software defined network (A-SDN), the method comprising:
collecting, by one or more processors, measurement data corresponding to a plurality of network variables from at least one Wi-Fi access point of the plurality of Wi-Fi access points and at least one client device of a plurality of client devices using a cloud based network application;
deriving, by one or more processors, a first optimal setting of a first decision variable for the at least one Wi-Fi-access point and a second optimal setting of a second decision variable for the at least one Wi-Fi-access point using a real-time optimization based on at least one network variable of the plurality of network variables, backhaul limits corresponding to the at least one Wi-Fi access point and a download behavior corresponding to the at least one client device, wherein the first decision variable is an operating channel selection corresponding to the at least one Wi-Fi access point and the second decision variable is a transmit power corresponding to the at least one Wi-Fi access point; and
wherein the real-time optimization comprises formulation of optimization problems through heuristic solutions and a machine learning based optimization;
wherein collecting the measurement data further comprises:
filtering, by one or more processors, the plurality of network variables to eliminate a channel and an environmental noise, wherein the filtering comprises eliminating outlier measurement data and extracting a representative pattern of the plurality of network variables through smoothing;
decomposing, by one or more processors, the representative pattern of the plurality of network variables into a set of correlated variables, wherein a number of the plurality of network variables is reduced;
selecting, by one or more processors, a subset of correlated variables from the set of correlated variables, wherein the subset of correlated variables has a relatively lower condition number of a covariance matrix; and
clustering, by one or more processors, the subset of correlated variables.

### Claim 2 (depends on 1)

The method according to claim 1, wherein collecting the measurement data comprises adapting, by one or more processors, a frequency of import of the measurement data corresponding to the plurality of network variables based on a transition profile of the plurality of network variables and environmental disturbance associated with the plurality of Wi-Fi access points.

### Claim 3 (depends on 1)

The method according to claim 1, wherein the subset of correlated variables is clustered based on one of a location of a network element, an access point property and a network element volatility.

### Claim 4 (depends on 1)

The method according to claim 1, wherein a network variable is one of an access point average access delay, an average access delay for each access category, an associated client device count, a channel utilization, a transmitted fragment count, a group addressed transmitted frame count, a failed count, a retry count, a multiple retry count, a frame duplicate count, a request to send (RTS) count, an RTS failure count, an acknowledge (ACK) failure count, a received fragment count, a group addresses received frame count, a frame check sequence (FCS) error count, a transmitted frame count, a received signal strength (RSSI) on a client device and a total data transfer.

### Claim 5 (depends on 1)

The method according to claim 1, wherein the real-time optimization is a two-step optimization process, wherein a first optimization step of the two-step optimization process is used for deriving the first optimal setting of the first decision variable for the at least one Wi-Fi access point; and
wherein the first optimization step comprises formulating, by one or more processors, an optimization problem using a maximum data rate from a Wi-Fi access point of the plurality of Wi-Fi access point to a client device of the plurality of client devices at a particular time, a data demand of the client device from the Wi-Fi access point, a backhaul limit corresponding to the Wi-Fi access point, a Heaviside step function, an adjustable parameter, a set of equality constraint functions, a set of inequality constraint functions, a transmission channel of the Wi-Fi access point and a subset of available channels on corresponding IEEE 802.11 standard.

### Claim 6 (depends on 5)

The method according to claim 5, wherein the maximum data rate is calculated using SINR corresponding to the client device connected to the Wi-Fi access point and a stepwise function that relates the SINR to the corresponding data rate for a particular IEEE 802.11 standard.

### Claim 7 (depends on 6)

The method according to claim 6, wherein the SINR for the client device is calculated using RSSI from the Wi-Fi access point and a fractional bandwidth overlap between transmission channels of the Wi-Fi access point and a second Wi-Fi access point of the plurality of Wi-Fi access points.

### Claim 8 (depends on 5)

The method according to claim 5, wherein a second optimization step of the two-step optimization process is used for deriving the second optimal setting of the second decision variable for the at least one Wi-Fi access point.

### Claim 9 (depends on 8)

The method according to claim 8, wherein the second optimization step comprises formulating, by one or more processors, an optimization problem to reduce transmit power of a Wi-Fi access point of the plurality of Wi-Fi access points using the SINR of a client device of the plurality of client devices and a maximum data rate from a Wi-Fi access point of the plurality of Wi-Fi access point to the client device.

### Claim 10 (depends on 1)

The method according to claim 1, wherein an optimization problem is formulated as a Mixed Integer Nonlinear Programming (MINLP) problem.

### Claim 11 (depends on 1)

The method according to claim 1 further comprises recalculating, by one or more processors, an optimal setting for the at least one Wi-Fi access point based on one of changes in a data transfer rate, changes in a network management policy, or measurable and immeasurable disturbances.

### Claim 12

A system for managing a plurality of Wi-Fi access points using a cloud based adaptive software defined network (A-SDN), the system comprising:
a memory;
a processor communicatively coupled to the memory, wherein the processor is configured to:
collect measurement data corresponding to a plurality of network variables from at least one Wi-Fi access point of the plurality of Wi-Fi access points and at least one client device of a plurality of client devices using a cloud based network application;
derive a first optimal setting of a first decision variable for the at least one Wi-Fi-access point and a second optimal setting of a second decision variable for the at least one Wi-Fi-access point using a real-time optimization based on at least one network variable of the plurality of network variables, backhaul limits corresponding to the at least one Wi-Fi access point and a download behavior corresponding to the at least one client device, wherein the first decision variable is an operating channel selection corresponding to the at least one Wi-Fi access point and the second decision variable is a transmit power corresponding to the at least one Wi-Fi access point; and
wherein the real-time optimization comprises formulation of optimization problems through heuristic solutions and a machine learning based optimization;
wherein the real-time optimization is a two-step optimization process;
wherein a second optimization step of the two-step optimization process is used for deriving the second optimal setting of the second decision variable for the at least one Wi-Fi access point; and
wherein the processor is configured to formulate an optimization problem to reduce transmit power of a Wi-Fi access point of the plurality of Wi-Fi access points using the SINR of a client device of the plurality of client devices and a maximum data rate from a Wi-Fi access point of the plurality of Wi-Fi access point to the client device.

### Claim 13 (depends on 12)

The system according to claim 12, wherein the processor is configured to adapt a frequency of import of the measurement data corresponding to the plurality of network variables based on a transition profile of the plurality of network variables and environmental disturbance associated with the plurality of Wi-Fi access points.

### Claim 14 (depends on 12)

The system according to claim 12, wherein the processor is configured to:
filter the plurality of network variables to eliminate a channel and an environmental noise, wherein the filtering comprises eliminating outlier measurement data and extracting a representative pattern of the plurality of network variables through smoothing;
decompose the representative pattern of the plurality of network variables into a set of correlated variables, wherein a number of the plurality of network variables is reduced;
select a subset of correlated variables from the set of correlated variables, wherein the subset of correlated variables has a relatively lower condition number of a covariance matrix; and
cluster the subset of correlated variables.

### Claim 15 (depends on 12)

The system according to claim 12, wherein the processor is further configured to:
forecast network dynamics based on density at the least one Wi-Fi access point, wherein an average data demand profile is calculated with less and periodic measurements based on the network dynamics, creating a statistical database for future predictions; and
perform an uncertainty evaluation of the network from the at least one client device to the at least one Wi-Fi access point.

### Claim 16 (depends on 12)

The system according to claim 12, wherein a first optimization step of the two-step optimization process is used for deriving the first optimal setting of the first decision variable for the at least one Wi-Fi access point.

### Claim 17 (depends on 16)

The system according to claim 16, wherein the processor is configured to formulate an optimization problem using a maximum data rate from a Wi-Fi access point of the plurality of Wi-Fi access point to a client device of the plurality of client devices at a particular time, a data demand of the client device from the Wi-Fi access point, a backhaul limit corresponding to the Wi-Fi access point, a Heaviside step function, an adjustable parameter, a set of equality constraint functions, a set of inequality constraint functions, a transmission channel of the Wi-Fi access point and a subset of available channels on corresponding IEEE 802.11 standard.

### Claim 18 (depends on 12)

The system according to claim 12, wherein the processor is further configured to recalculate an optimal setting for the at least one Wi-Fi access point based on one of changes in a data transfer rate, changes in a network management policy, or measurable and immeasurable disturbances.

### Claim 19

A method for managing a plurality of Wi-Fi access points using a cloud based adaptive software defined network (A-SDN), the method comprising:
collecting, by one or more processors, measurement data corresponding to a plurality of network variables from at least one Wi-Fi access point of the plurality of Wi-Fi access points and at least one client device of a plurality of client devices using a cloud based network application;
deriving, by one or more processors, a first optimal setting of a first decision variable for the at least one Wi-Fi-access point and a second optimal setting of a second decision variable for the at least one Wi-Fi-access point using a real-time optimization based on at least one network variable of the plurality of network variables, backhaul limits corresponding to the at least one Wi-Fi access point and a download behavior corresponding to the at least one client device, wherein the first decision variable is an operating channel selection corresponding to the at least one Wi-Fi access point and the second decision variable is a transmit power corresponding to the at least one Wi-Fi access point; and
wherein the real-time optimization comprises formulation of optimization problems through heuristic solutions and a machine learning based optimization;
wherein deriving the first optimal setting and the second optimal setting using the real-time optimization comprises:
forecasting, by one or more processors, network dynamics based on density at the least one Wi-Fi access point, wherein an average data demand profile is calculated with less and periodic measurements based on the network dynamics, creating a statistical database for future predictions; and
performing, by one or more processors, an uncertainty evaluation of the network from the at least one client device to the at least one Wi-Fi access point.

### Claim 20 (depends on 19)

The method according to claim 19, wherein the real-time optimization is a two-step optimization process;
wherein a first optimization step of the two-step optimization process is used for deriving the first optimal setting of the first decision variable for the at least one Wi-Fi access point; and
wherein the first optimization step comprises formulating, by one or more processors, an optimization problem using a maximum data rate from a Wi-Fi access point of the plurality of Wi-Fi access point to a client device of the plurality of client devices at a particular time, a data demand of the client device from the Wi-Fi access point, a backhaul limit corresponding to the Wi-Fi access point, a Heaviside step function, an adjustable parameter, a set of equality constraint functions, a set of inequality constraint functions, a transmission channel of the Wi-Fi access point and a subset of available channels on corresponding IEEE 802.11 standard.

### Claim 21 (depends on 20)

The method according to claim 20, wherein the maximum data rate is calculated using SINR corresponding to the client device connected to the Wi-Fi access point and a stepwise function that relates the SINR to the corresponding data rate for a particular IEEE 802. 11 standard.

### Claim 22 (depends on 21)

The method according to claim 21, wherein the SINR for the client device is calculated using RSSI from the Wi-Fi access point and a fractional bandwidth overlap between transmission channels of the Wi-Fi access point and a second Wi-Fi access point of the plurality of Wi-Fi access points.

### Claim 23 (depends on 20)

The method according to claim 20, wherein a second optimization step of the two-step optimization process is used for deriving the second optimal setting of the second decision variable for the at least one Wi-Fi access point;
wherein the second optimization step comprises formulating, by one or more processors, an optimization problem to reduce transmit power of a Wi-Fi access point of the plurality of Wi-Fi access points using the SINR of a client device of the plurality of client devices and a maximum data rate from a Wi-Fi access point of the plurality of Wi-Fi access point to the client device.

### Claim 24 (depends on 19)

The method according to claim 19, wherein an optimization problem is formulated as a Mixed Integer Nonlinear Programming (MINLP) problem.

### Claim 25 (depends on 19)

The method according to claim 19 further comprises recalculating, by one or more processors, an optimal setting for the at least one Wi-Fi access point based on one of changes in a data transfer rate, changes in a network management policy, or measurable and immeasurable disturbances.