Call Capacity Allocation via Linear Programming Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing telecommunications systems face challenges in predicting traffic demands, leading to blocked calls and inefficient capacity allocation across geographic regions and call destinations, due to manual intervention and sub-optimal capacity partitioning.
Innovation Solution
A capacity allocation system that uses historical data to predict traffic demands and rank call-termination devices based on their capability to satisfy regional demands, employing linear programming to maximize capacity allocation and prioritize service levels, thereby optimizing call routing and reducing waste.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If manual intervention is used for capacity partitioning, then system complexity is reduced, but capacity allocation efficiency deteriorates
Solution Approach 1:
The system automatically performs capacity partitioning and allocation without manual intervention. The capacity allocation module autonomously determines optimal capacity distribution across call-termination devices based on forecasted traffic demands and service priorities, eliminating the need for manual configuration while maximizing allocation efficiency.
2Device complexity
If sub-optimal capacity partitioning is used, then device complexity is reduced, but call blocking increases
Solution Approach 1:
The system performs preliminary capacity partitioning based on forecasted traffic demands before actual call traffic arrives. By pre-calculating optimal capacity allocations for different geographic regions and service priorities, the system ensures that sufficient capacity is reserved to meet anticipated demand, thereby reducing call blocking while maintaining manageable system complexity.
3Productivity
If automated capacity allocation is implemented, then capacity allocation efficiency is improved, but device complexity increases
Solution Approach 1:
The automated capacity allocation system is segmented into distinct functional modules: a capacity allocation module that determines optimal distribution, a routing module that implements allocations, and a forecast module that provides traffic predictions. This modular segmentation manages system complexity by dividing the automated allocation function into manageable, independent components that can operate autonomously.
4Measurement precision
If historical data analysis is used for traffic prediction, then prediction accuracy is improved, but processing time increases
Solution Approach 1:
The system performs traffic demand forecasting in advance using historical data analysis, completing predictions before actual call traffic needs to be routed. By pre-processing historical data and generating traffic forecasts ahead of time, the system achieves high prediction accuracy without causing time loss during actual call handling, as the predictions are already available when needed.
Data Source
Figure 1A
Figure 1B
Figure 2
AI summary
A system and method are disclosed that allocate call capacity based on the need to divide the call capacities of at least some call-termination devices across geographic regions. Accordingly, the allocation system uses various input parameters as constraints in a linear programming optimization, which has the objective of maximizing the capacity allocation of a device to fulfill the traffic demands of each region being processed. The input parameters that are used include i) the traffic demand forecast of each geographic region being evaluated, ii) the available call capacity of each call-termination device, iii) the call destinations associated with each region, and iv) service levels associated with each given call destination. Call-capacities are separately allocated for i) the predicted traffic that is expected in the next time period and ii) an additional margin of excess traffic above and beyond the expected traffic.