Intermittently-Connected Network Congestion Control With Local Flow Feedback
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing congestion control methods for intermittently-connected and lossy networks (ICLNs) are ineffective due to high transmission delays, jitter, and bursty traffic patterns, violating TCP assumptions and leading to network inefficiencies such as increased packet delivery times and resource wastage.
Innovation Solution
A local congestion control method using a buffer-space allocation (BSA) agent and hop-by-hop local-flow-control (LFC) agent to manage network traffic based on a pricing model that considers sojourn time and queue position, employing a Proximal Point Optimization (PPO) deep reinforcement learning algorithm to optimize buffer space allocation and packet flow control.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If TCP-based congestion control is used in ICLNs, then end-to-end congestion control functionality is provided, but network efficiency deteriorates due to high transmission delays, jitter, and bursty traffic patterns violating TCP assumptions
Solution Approach 1:
The patent introduces an intermediary congestion control mechanism at the network node level (between endpoints) to mediate traffic flow. The network node actively monitors queue occupancy and sends congestion control signals back to the source, acting as a mediator that prevents endpoint-based TCP from making decisions based on outdated loss signals. This intermediary control resolves the contradiction by providing reliable congestion control functionality while adapting to ICLN conditions to maintain network efficiency.
Solution Approach 2:
The patent implements a feedback mechanism where the network node monitors its own queue occupancy and sends congestion control signals back to the traffic source based on real-time buffer status. This local feedback loop allows the system to respond to congestion conditions immediately without waiting for packet loss to occur, resolving the contradiction by maintaining both reliability of congestion control and efficiency under high-delay conditions where traditional TCP feedback is too slow.
2Reliability
If buffer space is allocated to accommodate bursty traffic in ICLNs, then packet loss is reduced, but sojourn time increases due to packets waiting in queues
Solution Approach 1:
The patent implements dynamic buffer space allocation where the network node adjusts the amount of buffer space allocated to different flows based on real-time traffic conditions and congestion signals. Rather than static allocation, the buffer capacity is dynamically modified in response to congestion feedback, allowing the system to reduce sojourn time during low-congestion periods while maintaining packet delivery reliability during high-congestion periods. This resolves the contradiction by making buffer allocation adaptive rather than fixed.
Solution Approach 2:
The patent changes the parameter of buffer space allocation dynamically based on congestion conditions. When congestion is detected, the system adjusts buffer allocation parameters to prioritize certain flows or reduce overall buffer usage, thereby controlling sojourn time while maintaining essential packet delivery reliability. This parameter change approach resolves the contradiction between reliability and time loss by allowing flexible adjustment based on network state.
3Productivity
If congestion control signals are sent back to sources in ICLNs, then traffic flow is regulated, but signal transmission is affected by high delays and packet loss
Solution Approach 1:
The patent implements preliminary congestion control by having the network node proactively send congestion signals back to the source before the buffer becomes completely full or packet loss occurs. This preliminary action allows the source to reduce its transmission rate in advance, preventing congestion before it severely impacts signal transmission reliability. By acting preliminarily rather than reactively, the system maintains both traffic flow regulation and signal transmission reliability despite high delays and loss conditions.
4Reliability
If packet retransmissions are used to handle losses in ICLNs, then data integrity is maintained, but network resources are wasted due to retransmission overhead
Solution Approach 1:
The patent applies preliminary anti-action by having the network node send congestion control signals to prevent packet loss in the first place, rather than relying on retransmissions to correct losses afterward. By proactively regulating traffic flow based on buffer occupancy, the system prevents congestion-related losses, thereby maintaining data integrity while avoiding the energy waste of retransmissions. This preliminary prevention approach resolves the contradiction between reliability and resource consumption.
Data Source
AI summary
A method for controlling congestion in intermittently-connected and lossy computer networks comprising: determining, at a local network node, a payoff score for each of a plurality of active flows of network traffic, wherein each active flow consists of a stream of in-transit packets at the local network node that come from a common source and share a common destination, wherein each active flow's payoff score is based on a pricing model that considers both a sojourn time and a position in a queue of each of an active flow's constituent packets; allocating unused buffer space across all active flows in the local network node based on relative traffic loads with a buffer-space allocation (BSA) agent; and controlling a rate at which packets from all active flows are received at the local network node with a hop-by-hop local-flow-control (LFC) agent according to each flow's payoff score.