AI-Based Listen Interval Prioritization for IoT Buffer Management

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

IoT devices with low power consumption face battery life challenges due to limited buffer memory in access points, which are overwhelmed by devices with long listen intervals, impacting both low-power and non-low power clients, necessitating a robust technique for AI-based data delivery prioritization.

Innovation Solution

A device prioritization model using machine learning categorizes stations by type for multicast grouping, prioritizing packets for low power devices, generating a DTIM message to indicate buffered messages, and transmitting data according to assigned listen intervals, employing group keys for each multicast group.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Duration of action of moving object

If access points buffer traffic for devices with long listen intervals to conserve battery life, then battery life of low power devices is extended, but buffer memory in access points is overwhelmed

Engineering Contradiction:
Improvebattery lifeVSAvoidbuffer memory capacity
Core Design Contradiction:
Duration of action of moving objectVSQuantity of substance

Solution Approach 1:

The patent segments stations into different categories (low power stations vs. other stations) and applies different buffer retention policies to each segment. Low power stations receive extended buffer retention to allow longer listen intervals, while other stations use standard retention policies, preventing buffer overflow.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent applies differentiated buffer management quality to different station types. Specifically, buffered frames for low power stations are retained longer than frames for other stations, creating local quality differentiation in buffer retention based on station power characteristics.

Inventive Principle:
Principle #3Local quality

2Use of energy by moving object

If listen interval is increased for low power devices to reduce wake frequency, then power consumption decreases, but data delivery reliability may be compromised

Engineering Contradiction:
Improvepower consumptionVSAvoiddata delivery reliability
Core Design Contradiction:
Use of energy by moving objectVSReliability

Solution Approach 1:

The access point performs preliminary actions by buffering outgoing frames for low power stations before the stations wake up. The buffer retains frames for the duration of the listen interval, ensuring data is ready for immediate delivery when the station wakes, maintaining reliability while allowing extended sleep periods.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The buffered frame storage in the access point acts as an intermediary between the network and low power stations. It holds data during the station's sleep period and delivers it when the station wakes, decoupling the timing of data generation from data reception and enabling extended listen intervals without reliability loss.

Inventive Principle:
Principle #24Intermediary (Mediator)

3Ease of manufacture

If uniform buffer retention policy is applied to all stations, then implementation is simple, but non-low power clients are unduly limited in data exchange opportunities

Engineering Contradiction:
Improveimplementation simplicityVSAvoiddata exchange opportunities
Core Design Contradiction:
Ease of manufactureVSProductivity

Solution Approach 1:

The patent segments the station population into low power stations and other stations, applying different buffer retention policies to each segment. This segmentation enables differentiated service where low power stations receive extended retention while other stations maintain normal data exchange rates, resolving the conflict between simplicity and productivity.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent implements local quality differentiation in buffer management by applying extended buffer retention specifically to low power stations while using standard retention for other stations. This localized quality enhancement improves data exchange opportunities for non-low power clients without complicating the overall system significantly.

Inventive Principle:
Principle #3Local quality

Data Source

PatentUS20230007585A1Machine learning and artificial intelligence model-based data delivery for IoT devices co-existing with high bandwidth devices
Publication Date: 2023.01.05 FORTINET INC
  • US20230007585A1 patent drawing
  • US20230007585A1 patent drawing
  • US20230007585A1 patent drawing

AI summary

Each of the plurality of stations connected to the access point can be profiled to determine device type, and determine a listen interval for each of the plurality of stations based on the device prioritization model based on DTIM periods of the plurality of stations. Delivery of multicast packets is prioritized from the enterprise network destined for a low power device multicast group on the Wi-Fi network and to prioritize delivery of unicast packets for low power device multicast group. The messages are transmitted to the stations over the Wi-Fi network according to the assigned listen interval.