Premises Mesh Network Management via Reinforcement Learning

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Mesh networks face challenges in efficient management due to increased workload on devices acting as both routers and endpoints, leading to potential latency and higher power consumption, which affects network performance and electricity costs.

Innovation Solution

A system utilizing a server to process premises planning data and network data through separate processing pipelines, generating outputs that are then processed using a reinforcement learning-based loop to determine optimal actions for device management, ensuring continuous connectivity and effective coverage without disrupting existing networks.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of manufacture

If mesh devices are configured to work like routers with automatic self-configuration, then network scalability and ease of deployment are improved, but device workload and complexity increase

Engineering Contradiction:
Improveease of deploymentVSAvoiddevice workload
Core Design Contradiction:
Ease of manufactureVSDevice complexity

Solution Approach 1:

The patent introduces a centralized controller as an intermediary that manages mesh network configuration and device coordination. This controller handles the complex tasks of routing decisions, device provisioning, and network management, allowing individual mesh devices to operate with reduced complexity while maintaining overall network scalability and automatic configuration capabilities.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Reliability

If mesh devices act as both endpoints and routes, then network redundancy and reliability are improved, but power consumption increases

Engineering Contradiction:
Improvenetwork redundancyVSAvoidpower consumption
Core Design Contradiction:
ReliabilityVSUse of energy by moving object

Solution Approach 1:

The patent implements dynamic role assignment where mesh devices can switch between endpoint and routing functions based on network conditions, device capabilities, and energy status. This dynamic behavior allows the network to maintain redundancy and reliability when needed while reducing power consumption by allowing devices to operate in lower-power endpoint mode when full routing functionality is not required.

Inventive Principle:
Principle #15Dynamics

3Area of stationary object

If mesh network size is increased for better coverage, then network range is improved, but latency increases due to device management overhead

Engineering Contradiction:
Improvenetwork rangeVSAvoidlatency
Core Design Contradiction:
Area of stationary objectVSLoss of time

Solution Approach 1:

The centralized controller acts as an intermediary that optimizes routing paths and manages data flow across the expanded mesh network. This allows the network to cover larger areas with more devices while the controller efficiently manages communication overhead, reducing latency by making intelligent routing decisions and preventing network congestion.

Inventive Principle:
Principle #24Intermediary (Mediator)

4Ease of operation

If automatic self-configuration is enabled with negligible central coordination, then ease of operation is improved, but network management control decreases

Engineering Contradiction:
Improveautomatic configurationVSAvoidmanagement control
Core Design Contradiction:
Ease of operationVSDevice complexity

Solution Approach 1:

The patent implements a feedback mechanism where the centralized controller receives status information from mesh devices and automatically adjusts configuration parameters, while also providing policies and constraints that guide device behavior. This feedback loop enables automatic self-configuration to maintain ease of operation while the controller preserves necessary management control through policy enforcement and centralized oversight.

Inventive Principle:
Principle #23Feedback

Data Source

PatentEP4312412A1System and method for managing devices in communication network that is implemented at premises
Publication Date: 2024.01.31 ELISA OYJ
  • EP4312412A1 patent drawingFigure 1~2
  • EP4312412A1 patent drawingFigure 3
  • EP4312412A1 patent drawing

AI summary

Disclosed is a system (100) for managing devices (102, 104, 106) in a communication network (108) that is implemented at a premises (110). The system comprises at least one server (112) configured to obtain premises planning data and network data from a service (228) associated with the premises, wherein the premises planning data pertains to at least one floor plan of the premises and the network data pertains to the devices in the communication network and a performance of the communication network; process the premises planning data via a first processing pipeline (202) to generate a first output; process the network data via a second processing pipeline (204) to generate a second output; and process the first output and the second output using a reinforcement learning-based processing loop (224) to determine at least one action that is to be implemented for managing the devices in the communication network.