Distributed Edge Cache Replication in Wireless Mesh Networks

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Wireless mesh networks face inefficiencies in digital content distribution due to high latency, network congestion, and variability in network path characteristics, which affect user experience and content accessibility in environments with limited broadband Internet connectivity.

Innovation Solution

Implementing content replication and caching strategies within the wireless mesh network, where popular digital content items are replicated and allocated across network nodes based on popularity and reachability, and multiple resolutions and bit rates are stored to adapt to changing network conditions, allowing for efficient content delivery and reduced latency.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Loss of time

If content is delivered from centralized servers through multiple hops in wireless mesh networks, then content accessibility is achieved, but latency increases and network congestion occurs

Engineering Contradiction:
Improvecontent delivery latencyVSAvoidnetwork efficiency
Core Design Contradiction:
Loss of timeVSProductivity

Solution Approach 1:

The patent implements content replication that proactively copies popular digital content items from centralized servers to multiple network nodes before user requests occur. This preliminary distribution of content to edge locations in the mesh network eliminates the need for long-distance data traversal when users request content, thereby reducing latency and preventing network congestion during peak usage periods.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent segments the centralized content delivery architecture into a distributed network of content-storing nodes. By dividing the content storage function across multiple nodes throughout the mesh network rather than relying on a single centralized server, the system reduces the number of hops required for content delivery and distributes network traffic load, thereby reducing latency and improving overall network efficiency.

Inventive Principle:
Principle #1Segmentation

2Speed

If content is replicated across multiple network nodes, then content delivery speed improves, but network device complexity increases

Engineering Contradiction:
Improvecontent delivery speedVSAvoidnetwork architecture complexity
Core Design Contradiction:
SpeedVSDevice complexity

Solution Approach 1:

The patent implements autonomous content replication mechanisms where network nodes automatically identify popular content items, replicate them to appropriate nodes, and manage their own content caching without requiring complex centralized coordination. Each node independently monitors its content delivery performance and makes local decisions about what content to replicate, thereby simplifying the overall network architecture while still achieving fast content delivery through distributed replication.

Inventive Principle:
Principle #25Self-service

3Reliability

If multiple copies of content are stored across the network, then content accessibility and speed improve, but storage requirements increase

Engineering Contradiction:
Improvecontent accessibilityVSAvoidtotal storage capacity
Core Design Contradiction:
ReliabilityVSQuantity of substance

Solution Approach 1:

The patent implements selective content replication that places different content items at different network nodes based on local demand patterns and node characteristics. Rather than uniformly replicating all content across all nodes, the system optimizes content placement by analyzing which nodes are most likely to serve specific content requests, thereby improving content accessibility while minimizing redundant storage and optimizing the use of total network storage capacity.

Inventive Principle:
Principle #3Local quality

Data Source

PatentUS10530858B1Replication of content using distributed edge cache in wireless mesh networks
Publication Date: 2020.01.07 AMAZON TECH INC
  • US10530858B1 patent drawing
  • US10530858B1 patent drawing
  • US10530858B1 patent drawing

AI summary

A computer system comprises a memory to store instructions and a processor to: generate a list comprising a first entry including an identifier of a first mesh device and a number of neighboring devices that are reachable by the first mesh device over a single-hop path, and a second entry including an identifier of a second mesh device and a number of neighboring devices that are reachable by the second mesh device over a single-hop path; order the initial list in a descending order of the numbers of neighboring devices; determine a replication factor for the digital content item; identify a subset of entries of the ordered list, wherein a number of entries comprised by the subset is equal to the replication factor; and cause a first copy of the digital content item to be stored on the mesh device referenced by the first entry.