Optimizing 5G Base Station Placement Using Multiple-Criteria Decision Analysis (MCDA)
- Customer
- Tele2/Altel
- Project manager on the customer side
- Year of project completion
- 2023
- Project timeline
- July, 2022 - September, 2023
- Project scope
- 3600 man-hours
- Goals
-
-
High Capital Expenditure: Deploying 5G technology comes at a significant cost, including replacing or upgrading existing 4G infrastructure.
-
Operational Complexity: The decision-making process for station placement needs to consider multiple variables like traffic volume, existing user behavior, technological adaptability, etc.
-
Sustainability: The energy demands of 5G are significantly higher, making it essential to optimize for energy efficiency.
-
Quality of Service: Balancing the demands of existing 4G subscribers with the expectations of newer 5G subscribers is a challenging task.
-
- Project Results
-
Tangible Measurements
-
8% reduction in CAPEX through optimized placement.
-
5% increase in efficiency by reducing energy consumption.
- 5% increase in ROI through a more efficient deployment strategy.
-
The uniqueness of the project
Dynamic re-evaluation: As data changes over time, the algorithm recalibrates recommendations.
Predictive Analysis: Forecasting future 5G uptake based on current trends.
- Used software
- Big data from various company's systems.
- Difficulty of implementation
- Computational costs
- Project Description
-
Harness the power of Multiple-criteria decision analysis (MCDA) to evaluate potential 5G mobile station placements. By focusing on parameters like traffic volume, subscriber count, 5G device prevalence, user activity, and 5G-generated traffic, we can determine the most effective transition points.
Design Principles:
• Data-driven decision-making: Prioritize decisions based on real-world data.
• Scalability: Ensuring that the solution can adapt to future technological advancements.
• Customer-centric: Prioritize areas with high 5G device usage to cater to demand. - Project geography
- Whole country (Kazakhstan).