Switching Device Hash-Tagged Packet Routing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Packet switched communication networks face increased load and decreased Quality of Service (QoS) due to the growing number of data packets, especially for real-time services, as existing switching devices struggle to efficiently route and process these packets without overloading and compromising service quality.
Innovation Solution
A method that uses a hash-function to generate a hash-tag from filtered packet header information, allowing data packets to be routed to sub-engines for further processing without complete knowledge of their content, thereby distributing load and complexity across multiple sub-engines, enhancing scalability and reducing database access frequency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If switching devices process all data packets centrally, then routing accuracy is maintained, but device load increases overproportionally and QoS decreases
Solution Approach 1:
The patent divides the central switching device into multiple sub-engines (first sub-engine, second sub-engine, etc.), each responsible for processing specific data packets. This segmentation distributes the processing load across multiple independent units while maintaining centralized control for routing decisions, thereby reducing the burden on any single processing unit and preventing overloading.
Solution Approach 2:
The patent introduces a hash function as an intermediary mechanism that maps data packet characteristics to specific sub-engines. This hash-based distribution system acts as a mediator between the central controller and sub-engines, automatically routing packets to appropriate processing units based on hash values derived from packet headers, thus balancing load without compromising routing accuracy.
2Productivity
If switching devices increase processing capacity to handle more packets, then throughput increases, but device complexity and cost increase
Solution Approach 1:
The switching device is segmented into multiple identical or similar sub-engines rather than upgrading a single complex processor. Each sub-engine handles a portion of the traffic, achieving increased throughput through parallel processing while keeping individual unit complexity manageable and allowing for standardized, cost-effective implementations.
Solution Approach 2:
The patent employs multiple copies of sub-engines that can be identical or near-identical in structure and function. This copying approach allows the system to scale throughput by adding more copies rather than increasing the complexity of existing components, making the system more cost-effective and easier to maintain while achieving higher processing capacity.
3Measurement precision
If complete packet content is evaluated for routing, then routing precision is improved, but processing time increases
Solution Approach 1:
The patent extracts only the essential routing information from complete packet content, specifically using header fields (source IP, destination IP, source port, destination port, protocol type) to generate hash values. This extraction of critical elements without processing the entire packet content maintains routing precision while dramatically reducing processing time and computational overhead.
Solution Approach 2:
The patent applies partial action by processing only the necessary portion of packet data (header information) rather than the complete packet content. This selective processing of relevant fields provides sufficient routing precision while avoiding the time-consuming evaluation of unnecessary data, achieving an optimal balance between accuracy and speed.
Data Source
Figure 1
Figure 2
AI summary
The invention concerns a method of processing a plurality of data packets in a packet switched communication network comprising at least one switching device (1) as well as a corresponding switching device. The switching device (1) comprises two or more sub-engines (21,..., 29). A control unit (7) of the at least one switching device (1) receives a data packet comprising a header (10) of a communication packet exchanged via the communication network. The control unit (7) applies at least one filter operation (50) to the header (10) and generating thereby filtered information (100). The control unit (7) maps the filtered information (100) on a hash-tag (30) according to a hash-function (300). The control unit (7) provides a tagged-header (11) by means of inserting the hash-tag (30) in the header (10) of the data packet. The control unit (7) sends the data packet to a sub-engine (24) of the two or more sub-engines (21,..., 29) according to the tagged-header (11) for further processing of the data packet.