Switch Priority Adjustment for IoT Edge Latency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
IoT edge server systems face challenges with high latency, high bandwidth consumption, poor security, poor availability, and poor connectivity due to the reliance on cloud-based data processing, which can be exacerbated by issues with bi-directional communication between networking switches and local compute nodes.
Innovation Solution
A networking switch management system that dynamically adjusts priority based on the ability to establish bi-directional communication with a local compute node, operating as a master or standby switch within a cluster, and utilizing multiple network adapters for flexible communication paths, including wired and wireless links, to ensure high availability and low latency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If cloud-based data processing is used, then data analysis and storage functions are centralized, but latency and bandwidth consumption increase
Solution Approach 1:
The system segments the data processing architecture by introducing edge servers that perform local data analysis and storage, separating critical processing functions from centralized cloud infrastructure. This allows real-time processing at the edge while maintaining cloud connectivity for non-critical operations.
Solution Approach 2:
The patent adds a spatial dimension to data processing by deploying edge servers physically closer to IoT devices. This creates a multi-layered architecture (edge layer, cloud layer) that reduces communication distance and latency for time-sensitive operations.
2Reliability
If networking switch priority is statically configured, then master switch selection is deterministic, but availability decreases when compute node communication fails
Solution Approach 1:
The system implements dynamic priority adjustment where networking switches can change their priority status based on real-time communication conditions. When a switch loses bi-directional communication with its compute node, it dynamically reduces its priority to allow another switch to become master, ensuring continuous system operation.
Solution Approach 2:
The patent implements a feedback mechanism where networking switches continuously monitor their ability to establish bi-directional communication with compute nodes. This feedback loop enables automatic priority adjustment and master switch reselection when communication failures occur.
3Speed
If direct local link communication is required, then communication speed increases, but connectivity reliability decreases when the link fails
Solution Approach 1:
The system prepares for potential link failures by implementing pre-configured backup communication paths through alternative networking switches. When a direct local link fails, the system can rapidly switch to alternative paths without interrupting service, cushioning against the reliability impact.
Solution Approach 2:
The patent introduces alternative networking switches as intermediaries that can relay communication when the direct local link fails. These intermediary switches provide backup communication paths, maintaining connectivity even when primary direct links are unavailable.
Data Source
AI summary
An apparatus includes a local compute node connected to a first networking switch by a direct local link. The first switch reduces its priority in response to determining that that the first networking switch is currently unable to establish the bi-directional communication with the local compute node through the direct local link. If the reduced priority of the first networking switch is greater than a priority of each of at least one additional networking switch in a cluster, then the first networking switch operates as a master networking switch for the cluster. If the reduced priority of the first networking switch is less than the priority of any of the at least one networking switch, then the first networking switch operates as a standby networking switch and one of the atleast one additional networking switch having the highest priority is operated as the master networking switch for the cluster.