Hybrid Access Network Packet Processing for Bandwidth Congestion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In hybrid access networks, the token bucket-based load sharing mechanism fails to adjust load splitting proportion dynamically with changing bandwidth, leading to congestion in the DSL tunnel and underutilization of the LTE tunnel, resulting in transmission errors and reduced resource utilization.
Innovation Solution
A packet processing method that dynamically adjusts the bandwidth of links based on feedback from the second network device, guiding overflow traffic to another link to relieve primary link congestion and improve network bandwidth, using a virtual binding tunnel with dynamically managed sending and receiving windows.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If token bucket-based load sharing is used with fixed bandwidth parameters, then load sharing can be implemented, but the system cannot dynamically adjust load splitting proportion when bandwidth changes, causing congestion in one tunnel and underutilization of another
Solution Approach 1:
The patent implements dynamic bandwidth adjustment by allowing the HAAP to modify the bandwidth parameters of DSL and LTE tunnels in real-time based on actual network conditions. The HAAP receives feedback about packet loss and congestion from the HG, then dynamically adjusts the bandwidth allocation between tunnels to optimize load sharing and prevent congestion, transforming the static token bucket system into a dynamic adaptive system.
Solution Approach 2:
The patent establishes a feedback mechanism where the HG monitors packet loss and congestion status in each tunnel and reports this information back to the HAAP. The HAAP uses this feedback to adjust bandwidth parameters and redirect traffic, creating a closed-loop control system that continuously optimizes load sharing based on actual network performance and prevents transmission errors.
2Productivity
If fixed bandwidth is assigned to DSL and LTE tunnels, then initial load sharing can be established, but system resource utilization is reduced when one tunnel becomes congested and the other remains idle
Solution Approach 1:
The system dynamically adjusts bandwidth allocation between DSL and LTE tunnels based on real-time congestion status. When one tunnel becomes congested, the HAAP redirects traffic to the other tunnel with available capacity, maximizing network resource utilization and preventing wasted network capacity while maintaining high productivity.
Solution Approach 2:
The patent changes the bandwidth parameter dynamically based on network conditions. The HAAP modifies the bandwidth allocation parameters of the tunnels in response to congestion feedback, allowing the system to adapt to changing network topology and traffic patterns, thereby optimizing resource utilization and preventing capacity waste.
3Reliability
If congestion occurs in the DSL tunnel with fixed bandwidth parameters, then traffic continues to be injected into the congested tunnel, but this causes packet loss and prevents proper utilization of the idle LTE tunnel
Solution Approach 1:
The HG monitors packet loss and congestion status in the DSL tunnel and provides feedback to the HAAP. Upon detecting congestion, the HAAP uses this feedback to redirect traffic to the LTE tunnel, effectively reducing packet loss rate while maintaining flexible traffic distribution across available tunnels.
Solution Approach 2:
The system dynamically redirects traffic from congested tunnels to idle tunnels in real-time. When DSL tunnel congestion is detected, the HAAP dynamically adjusts traffic distribution to utilize the LTE tunnel, preventing packet loss and maintaining adaptability in traffic routing decisions.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
This application provides a packet processing method and a network device in a hybrid access network, to implement traffic balancing and real-time monitoring on network bandwidth, thereby relieving primary link congestion caused by a network emergency, and improving network bandwidth. The method is: sending, by a first network device, a first data packet in a first sending window to a second network device by using a first tunnel; in response to receiving a first acknowledgement response sent by the second network device, increasing, by the first network device, a size of the first sending window based on a first proportion; in response to not receiving, within a first predetermined time, the first acknowledgement response, decreasing the size of the first sending window based on a second proportion; and in response to determining that the size of the first sending window is greater than or equal to a first threshold, sending a second data packet to a second receiving window of the second network device by using a second sending window, where a first acknowledgement number carried in the first acknowledgement response is obtained based on a first sub-link number in the first data packet.