Edge Platform Adaptive Peer-to-Peer Communication via Priority Queuing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current edge cloud infrastructures face inefficiencies in managing and processing disparate requests from various devices due to lack of adaptive peer-to-peer data transmission schemes, varying processing priorities, and different data payloads, leading to inefficient communication and processing.
Innovation Solution
Implementing scalable, tiered, and power-efficient architectures with common hardware interfaces that utilize priority and granularity metadata for queuing, scheduling, and pipelining, allowing for dynamic adaptation and optimization of data transmission between edge devices and accelerators, enabling efficient handling of requests with different priorities and granularities.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If current edge cloud infrastructures process requests from various devices, then device connectivity and resource access are enabled, but communication efficiency and processing performance deteriorate due to lack of adaptive peer-to-peer data transmission schemes
Solution Approach 1:
The patent segments requests into different priority levels (high, medium, low) and processes them through separate queues. High-priority requests are handled immediately through direct peer-to-peer transmission, while lower-priority requests are batched and processed efficiently. This segmentation allows the system to adapt to different device types while maintaining high communication efficiency for time-sensitive data.
Solution Approach 2:
The system dynamically adapts its communication strategy based on request characteristics. It uses dynamic priority assignment, adaptive batching thresholds, and real-time queue management to optimize data transmission. The edge platform can adjust processing behavior on-the-fly to handle varying device types, data sizes, and priority levels without sacrificing overall throughput.
2Device complexity
If uniform processing is applied to all incoming requests, then system simplicity is maintained, but latency and throughput performance worsen for requests with different priorities and granularities
Solution Approach 1:
The patent divides the processing architecture into distinct segments: high-priority request handling path and low-priority batch processing path. This segmentation introduces minimal complexity while dramatically reducing latency for time-sensitive requests. The system maintains simple uniform processing for non-critical requests while providing optimized fast paths for high-priority data.
Solution Approach 2:
The patent introduces an intermediary priority classification mechanism that routes requests to appropriate processing queues. This intermediary layer adds minimal complexity but enables differentiated processing - high-priority requests bypass batching and get immediate peer-to-peer transmission, while lower-priority requests are efficiently batched. The intermediary sorter is simple yet effective at reducing overall latency.
3Measurement precision
If small data packets are transmitted individually, then data granularity and precision are maintained, but communication overhead and energy consumption increase
Solution Approach 1:
The patent implements dynamic batching that adapts to request characteristics. For high-priority requests, individual small packets are transmitted immediately without batching, preserving granularity and meeting latency requirements. For lower-priority requests with similar timing characteristics, the system dynamically batches packets to reduce communication overhead and energy consumption. The batching threshold and decision logic adapt in real-time based on traffic patterns.
Solution Approach 2:
The system changes the transmission parameter (batching vs. individual transmission) based on request priority and timing characteristics. High-priority requests maintain original packet granularity with no batching, while lower-priority requests are aggregated into larger batches. This parameter change optimizes energy efficiency without sacrificing data precision, as the same data is transmitted whether batched or individual, just at different times and with different overhead.
4Productivity
If batching is used for all requests, then communication efficiency and throughput are improved, but latency increases for time-sensitive high-priority requests
Solution Approach 1:
The patent segments the transmission system into two parallel channels: a fast path for high-priority requests that transmits individual packets immediately without batching, and a batch processing path for lower-priority requests that achieves high throughput through aggregation. This segmentation ensures that time-sensitive requests maintain low latency while bulk data transmission achieves high efficiency through batching.
Solution Approach 2:
The system dynamically determines whether to batch or transmit individually based on request priority and timing characteristics. High-priority requests with tight deadlines bypass batching entirely and transmit immediately at full speed. Lower-priority requests are candidates for batching to improve throughput. The decision is made dynamically for each request based on current system state and request characteristics, optimizing both speed and throughput.
Data Source
AI summary
In one embodiment, a method includes: receiving, in an edge platform, a plurality of messages from a plurality of edge devices coupled to the edge platform, the plurality of messages comprising metadata including priority information and granularity information; extracting at least the priority information from the plurality of messages; storing the plurality of messages in entries of a pending request queue according to the priority information; selecting a first message stored in the pending request queue for delivery to a destination circuit; and sending a message header for the first message to the destination circuit via at least one interface circuit, the message header including the priority information, and thereafter sending a plurality of packets including payload information of the first message to the destination circuit via the at least one interface circuit. Other embodiments are described and claimed.