Dynamic Bandwidth Allocation Circuit for PON Delay and Efficiency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In PON networks, existing dynamic bandwidth allocation methods face challenges in simultaneously restricting maximum delay time for low delay service classes while maintaining high bandwidth efficiency for normal delay service classes, accurately reflecting assured bandwidth, and reducing unused bandwidth, especially when packet data sizes are variable.
Innovation Solution
A dynamic bandwidth allocation circuit that classifies service paths into low and normal delay classes, allocates bandwidth based on requested values and excess allocated bandwidth, and adjusts allocations to ensure the sum value of allocated bandwidths falls within predetermined ranges, allowing for efficient distribution and utilization of bandwidth across cycles.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If bandwidth is allocated to low delay service class first in each cycle, then maximum delay time is restricted, but bandwidth efficiency for normal delay service class deteriorates
Solution Approach 1:
The patent segments the bandwidth allocation process into two distinct phases: a first phase dedicated to low delay service class allocations and a second phase for normal delay service class allocations. This segmentation allows each service class to receive appropriate bandwidth treatment without interfering with the other, resolving the contradiction between delay restriction and bandwidth efficiency.
Solution Approach 2:
The patent implements periodic action by allocating bandwidth to normal delay service class in alternating cycles rather than every cycle. In cycles where low delay service receives bandwidth, normal delay service is deferred to the next cycle. This periodic allocation pattern ensures that low delay requirements are met while normal delay service still receives adequate bandwidth opportunities, maintaining overall bandwidth efficiency.
2Reliability
If maximum bandwidth is allocated to service path terminating sections, then bandwidth requests are satisfied, but unused bandwidth increases when packet sizes are variable
Solution Approach 1:
The patent applies partial action by allocating bandwidth based on actual packet data sizes rather than always providing maximum requested bandwidth. The bandwidth allocation calculation section determines the exact bandwidth needed for each service path terminating section's packet data, avoiding excessive allocation that would result in unused bandwidth while still fully satisfying the actual bandwidth requirements.
3Adaptability or versatility
If bandwidth allocation calculation is performed for all service path terminating sections in each cycle, then bandwidth distribution is comprehensive, but transmission cycle length increases
Solution Approach 1:
The patent segments the bandwidth allocation calculation into two distinct phases: a first calculation phase for low delay service class service path terminating sections and a second calculation phase for normal delay service class sections. This segmentation reduces the computational burden in each phase, allowing comprehensive bandwidth distribution to be achieved while keeping individual calculation cycles shorter and more manageable.
Data Source
AI summary
The aim of the present invention is to achieve a low delay and high bandwidth efficiency in an upstream bandwidth and to accurately reflect assured bandwidth in an actual transmitted bandwidth. To achieve this, in the present invention the maximum value and the transmission cycle of a requested bandwidth are set for each service class. The present invention allocates bandwidth to a service path terminating section based on excess allocated bandwidth and requested bandwidth of each service path terminating section. For each service path terminating section the present invention also selects one bandwidth out of a plurality of bandwidths requested by a service path terminating section in such a manner that the sum value of the allocated bandwidths of the plurality of service path terminating sections is within a predetermined range.