Dynamic Interpacket Gap Adjustment for Tagged Packet Bandwidth
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional Ethernet networks require overclocking of ports to handle tagged packets, leading to increased power consumption due to higher bandwidth requirements, and existing methods to reduce overclocking are limited by physical coding sublayer restrictions in 10GBASE-R, 100GBASE-R, and 100GBASE-P networks.
Innovation Solution
The method involves reducing interpacket gaps (IPGs) by incrementing a parameter n for each data packet and setting the number of idle bytes in IPGs to NI = Nblock*n - (p mod Nblock), where n is reset after each IPG insertion, allowing for efficient bandwidth management and reduced overclocking.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If tag bytes are added to each packet to help with internal switch and switch fabric operations, then packet processing capability is improved, but bandwidth requirement increases leading to higher power consumption
Solution Approach 1:
The patent extracts bandwidth from the interpacket gap (IPG) period and reallocates it for transmitting tagged packets. By reducing the idle IPG time and utilizing this previously wasted bandwidth for actual data transmission, the system can handle tagged packets without requiring additional overclocking, thus avoiding increased power consumption while maintaining enhanced packet processing capability
Solution Approach 2:
The patent changes the IPG parameter from a fixed value to a dynamic value that is adjusted based on packet size and network conditions. By making the IPG adaptive rather than static, the system optimizes bandwidth utilization to accommodate tagged packets without requiring sustained high-speed operation, thereby reducing power consumption while maintaining processing capability
2Productivity
If ports are overclocked to handle tagged packets, then bandwidth capacity is improved, but power consumption increases
Solution Approach 1:
The patent implements dynamic IPG adjustment where the gap between packets varies based on actual traffic conditions and packet sizes. This dynamic approach allows the system to maintain high bandwidth capacity when needed while avoiding sustained overclocking during lower traffic periods, thereby reducing power consumption while preserving productivity
Solution Approach 2:
The patent reduces idle time in the IPG period and converts this previously wasted time into useful transmission time for tagged packets. By ensuring continuous useful action rather than allowing idle gaps, the system achieves higher effective bandwidth capacity without requiring the ports to operate at continuously elevated speeds, thus reducing power consumption
3Productivity
If interpacket gap is reduced to steal bandwidth, then bandwidth efficiency is improved, but physical coding sublayer restrictions in 10GBASE-R, 100GBASE-R, and 100GBASE-P networks limit the reduction
Solution Approach 1:
The patent makes the IPG dynamic and adaptive rather than fixed, allowing it to be reduced to the maximum extent permitted by physical coding sublayer restrictions for each specific packet. This dynamic adjustment enables the system to achieve high bandwidth efficiency while working within the constraints of 10GBASE-R, 100GBASE-R, and 100GBASE-P networks
Solution Approach 2:
The patent changes the IPG from a fixed parameter to a variable parameter that is optimized for each packet transmission based on packet size and network conditions. This parameter change allows the system to achieve better bandwidth efficiency while adapting to the specific restrictions of different Ethernet physical layer standards
Data Source
AI summary
In a method of transferring a plurality of data packets from a media access control (MAC) layer device to a physical layer (PHY) device, interpacket gaps (IPGs) having a number NI=Nblock*n−(p mod Nblock) of idle bytes are inserted between packets, where p is an integer denoting a length of a data packet in bytes, Nblock is a blocking size in bytes, and n is an integer initialized to one and incremented every time q data bytes of the data packet are transferred. The parameter n is reset to one in connection with each IPG insertion.