Adaptive Packet Size Adjustment for UDT Network Reliability
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
UDP-based data transfer protocols face issues with packet loss due to intermediary devices, leading to reduced bandwidth and failed data transfers, as smaller packets are dropped or lost, causing repeated retransmissions and congestion avoidance mechanisms to reduce bandwidth inefficiently.
Innovation Solution
Implementing adaptive packet size adjustment by subdividing packets into smaller subpackets when a 'black hole' is detected, allowing successful communication and avoiding drastic bandwidth reductions, with the receiving device reassembling the original packet and adding the packet size to a transmission filter to prevent future losses.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If UDT packets of certain sizes are transmitted, then bandwidth utilization is improved, but packet loss occurs due to intermediary device filtering
Solution Approach 1:
The patent segments packets into smaller subpackets when a packet size is identified as being dropped by intermediary devices. Instead of transmitting the original packet as a single unit, it divides the packet into multiple smaller subpackets that can successfully traverse the network path, thereby resolving the contradiction between utilizing larger packet sizes for bandwidth efficiency and avoiding packet loss for reliability.
2Reliability
If MTU is limited to avoid dropped packet sizes, then packet loss is reduced, but usable bandwidth is significantly limited
Solution Approach 1:
The patent implements dynamic packet size adjustment rather than a static MTU limitation. The system dynamically adapts packet sizes based on network conditions and identified black hole ranges, allowing the usable bandwidth to be optimized while avoiding packet loss by adjusting packet sizes in real-time based on feedback from packet transmission results.
Solution Approach 2:
The patent changes the packet size parameter dynamically based on network conditions and identified black hole ranges. Instead of fixing the MTU to a conservative value, the system modifies packet size parameters adaptively, selecting optimal sizes that avoid dropped ranges while maximizing bandwidth utilization for reliable data transfer.
3Reliability
If retransmissions are performed for dropped packets, then data reliability is maintained, but data transfer is delayed due to repeated failures
Solution Approach 1:
The patent performs preliminary action by proactively subdividing packets into subpackets before transmission when packet loss is detected, rather than relying on reactive retransmission of failed packets. This preliminary segmentation prevents the time loss associated with repeated retransmission attempts of packets that are likely to be dropped again.
Solution Approach 2:
The patent creates copies of packet data in the form of multiple subpackets instead of relying on retransmission of the original packet. By distributing the data across multiple smaller subpacket copies, the system avoids the time delay of retransmission while maintaining data integrity through reassembly of the subpacket copies at the destination.
4Reliability
If congestion avoidance features are applied, then packet loss is reduced, but bandwidth utilization is drastically reduced
Solution Approach 1:
The patent applies local quality by targeting congestion avoidance specifically at the problematic packet sizes that are being dropped, rather than applying blanket congestion control to all packets. This localized approach maintains high bandwidth utilization for packets of safe sizes while only applying avoidance measures to the specific size ranges identified as causing losses.
Data Source
AI summary
Described embodiments provide systems and methods for adaptive packet size adjustment for UDT or similar protocols, responsive to actual network conditions. Upon detection of a ‘black hole’ or a size or range of sizes of packets that are repeatedly lost, the system may subdivide packets matching the size or range into smaller subpackets. The receiving device may receive each subpacket and reassemble the original packet, allowing successful communication. Other packets, both smaller and larger than the identified size or range, may be transmitted normally, avoiding drastic reductions to congestion windows or application of other congestion avoidance features, or reducing the MTU value to avoid the ‘black hole’ range while sacrificing overall useable bandwidth. As a result, the system may more efficiently and reliably communicate large data payloads.