Packet Scheduling via Expected Arrival Time Comparison
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing packet scheduling methods struggle to efficiently allocate bandwidth among multiple flows in a network, particularly in large-scale networks, due to complexity and the need for frame size adjustments and synchronization, which affects system performance and fairness in bandwidth usage.
Innovation Solution
A packet scheduling method that calculates the expected time of arrival for each packet based on allocated bandwidth and packet length, comparing it to actual arrival times, and adjusts scheduling accordingly, including discarding packets if they arrive too late and recalculating expected times if they arrive too early, to ensure fair bandwidth distribution among flows.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If weighted fair queuing or fair queuing algorithms are used to allocate bandwidth fairly among flows, then bandwidth fairness is improved, but system complexity increases to O(logN)
Solution Approach 1:
The patent changes the scheduling parameter from virtual time calculation to expected arrival time based on packet length and allocated bandwidth. By using the formula ETA = previous_ETA + packet_length/allocated_bandwidth, the system maintains bandwidth fairness while avoiding complex O(logN) calculations, reducing computational complexity to O(1) per packet.
2Device complexity
If stop-end-go or hierarchical-round-Robin scheduling is used with frame-based timing, then bandwidth allocation is simplified, but system performance is affected by frame size and inter-frame synchronization requirements
Solution Approach 1:
The patent replaces static frame-based scheduling with dynamic per-packet expected arrival time calculation. Instead of being constrained by fixed frame sizes and synchronization intervals, the system dynamically adjusts scheduling decisions based on actual packet lengths and allocated bandwidth, eliminating frame size constraints and synchronization overhead while improving system performance.
3Reliability
If bandwidth is allocated fairly to each flow without packet arrival time control, then bandwidth fairness is improved, but packet congestion occurs and bandwidth efficiency decreases
Solution Approach 1:
The patent implements feedback control by comparing actual packet arrival times with expected arrival times. When packets arrive earlier than expected, the system adjusts subsequent expected times to prevent congestion. This feedback mechanism ensures that flows use only their fair share of bandwidth while maintaining high bandwidth efficiency through congestion prevention.
Data Source
AI summary
A packet scheduling method and apparatus which allows multiple flows that require data transmission to the same output port of a network device such as a router to fairly share bandwidth. The packet scheduling method includes calculating an expected time of arrival of a (k+1)-th packet subsequent to a currently input k-th packet of individual flows by use of bandwidth allocated fairly to each of the flows and a length of the k-th packet; in response to the arrival of the (k+1)-th packet, comparing the expected time of arrival of the (k+1)-th packet to an actual time of arrival of the (k+1)-th packet; and scheduling the (k+1)-th packet of each flow according to the comparison result.