Distributed Node Message Routing for Power Efficiency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional data processing systems with distributed processing nodes face high power consumption due to inefficient network communication, particularly in systems using middleware like CORBA, which requires significant memory and processor capacity, and unnecessary polling of messages leads to wasted processor power in nodes like lunar rovers or other battery-powered devices.
Innovation Solution
A data processing system with distributed nodes that uses a routing thread to determine message destination based on application identifiers, reducing unnecessary message processing and polling by directly routing messages to specific applications without central node management, thereby minimizing power consumption.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If common middleware software like CORBA is used for message communication between distributed processing nodes, then communication functionality is provided, but memory consumption and processor capacity utilization increase significantly
Solution Approach 1:
The patent extracts and removes the heavy middleware software layer from the distributed processing system. Instead of using comprehensive middleware like CORBA that provides general-purpose communication functionality, the invention implements a streamlined message routing system with dedicated routing threads and message queues that provide only the essential communication functions needed, thereby eliminating the significant memory and processor overhead associated with full middleware stacks.
Solution Approach 2:
The patent employs lightweight, temporary message queue structures and routing mechanisms that are created and destroyed as needed for specific communication tasks. These temporary communication infrastructure elements replace persistent middleware runtime environments, consuming minimal memory and processor resources during their brief operational lifetime while providing the necessary message passing functionality.
2Ease of operation
If centralized coordination is implemented for message communication between distributed nodes, then message routing is managed, but processor power is consumed through unnecessary polling when data has not changed
Solution Approach 1:
The patent implements a self-service message notification system where the sender node autonomously notifies the receiver node when a message is ready for processing. The routing thread at the sender node places messages in the receiver's message queue and triggers a notification event, eliminating the need for the receiver to continuously poll for new messages. This event-driven approach ensures processor power is consumed only when actual communication occurs.
Solution Approach 2:
The patent establishes a feedback mechanism where message status and availability information flows back from senders to receivers through the routing infrastructure. This feedback loop allows receiving nodes to be informed of message availability without active polling, enabling them to enter low-power states when no messages are pending while maintaining real-time awareness of communication status.
3Productivity
If middleware software is used for distributed communication, then message processing capability is provided, but memory consumption increases by 500k-2M bytes per node
Solution Approach 1:
The patent segments the message processing functionality into separate, dedicated components: routing threads for message reception and distribution, message queues for temporary storage, and application-specific processing threads. This segmentation eliminates the need for a monolithic middleware runtime that consumes large amounts of memory, allowing each component to allocate only the minimal memory required for its specific function while collectively providing comprehensive message processing capability.
4Ease of operation
If conventional network infrastructure with polling is used, then communication coordination is achieved, but data throughput is reduced due to unnecessary message processing
Solution Approach 1:
The patent enables continuous useful action by ensuring that processor cycles are consistently dedicated to actual message processing tasks rather than being periodically wasted on polling operations. The event-driven notification system keeps receiving nodes actively processing messages as they arrive without interruption, while allowing them to enter idle states only when no messages are present, thereby maximizing the continuity and efficiency of useful computational work.
Data Source
AI summary
Methods, systems, and articles of manufacture consistent with the present invention provide for a data processing system having an efficient network communication infrastructure. The data processing system has a plurality of data processing nodes distributed across a network. Each node comprises a processor and a memory having an application and a routing thread operatively configured to receive a message on the network. The message has an application identifier and a message group identifier. Each application has a command thread and a plurality of message group handlers. Each routing thread is operatively configured to determine whether the received message is for the application on the respective node and to notify the application's command thread of the received message in response to determining that the message is for the application. The command thread then directs the message to one of the message group handlers based on the received message's message group identifier.