LTE Network Congestion Reduction via ML Prediction and Traffic Redistribution
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
4G LTE cellular networks face congestion issues due to dynamic capacity demands, leading to suboptimal manual fine-tuning of network parameters, which can result in degraded user experiences and call drops, as the fixed number of Physical Resource Blocks (PRBs) per cell limits throughput, especially under high demand scenarios.
Innovation Solution
A method using a Multi-Layer Perceptron Deep Learning (MLPDL) structure to predict PRB utilization congestion thresholds in 4G LTE cellular towers, combined with Block Coordinated Descent Simulated Annealing (BCDSA) and Genetic Algorithm (GA) optimization techniques to dynamically adjust cell power and handover thresholds, redistributing traffic from congested to non-congested cells, thereby reducing network congestion.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If manual fine-tuning of cellular network parameters is performed to alleviate congestion, then congestion is partially reduced, but the optimization results are suboptimal and require significant operator effort
Solution Approach 1:
The system enables self-service through automated algorithms that independently analyze network congestion patterns, predict PRB utilization thresholds using machine learning models, and adjust cell power and handover parameters without human intervention. The congestion management system serves itself by continuously monitoring and optimizing network parameters based on real-time conditions.
Solution Approach 2:
Manual mechanical tuning of network parameters by operators is replaced with automated computational algorithms. The system uses machine learning models and optimization algorithms to automatically adjust cell power, handover thresholds, and other network parameters, substituting human-operated mechanical processes with intelligent automated systems.
2Productivity
If the number of Physical Resource Blocks (PRBs) per cell is increased to handle high demand, then throughput capacity is improved, but infrastructure cost and complexity increase significantly
Solution Approach 1:
The system implements dynamic allocation of network resources by continuously adjusting cell power and handover parameters based on real-time traffic conditions. Instead of statically increasing PRB counts, the system dynamically optimizes the utilization of existing resources through automated parameter adjustment, allowing the network to adapt to varying demand patterns without permanent infrastructure changes.
Solution Approach 2:
The system changes operational parameters (cell power levels, handover thresholds, frequency allocation) to optimize network capacity and throughput. By adjusting these parameters dynamically, the system maximizes the utilization of existing PRBs and achieves higher effective capacity without physically adding more resource blocks or infrastructure elements.
3Area of stationary object
If cell power is increased to expand coverage area, then coverage is improved, but traffic congestion in the cell increases
Solution Approach 1:
The system implements feedback mechanisms where the automated algorithms continuously monitor cell load, PRB utilization, and traffic patterns. Based on this feedback, the system dynamically adjusts cell power levels and handover parameters to maintain optimal balance between coverage area and traffic capacity, preventing congestion while preserving necessary coverage.
Solution Approach 2:
The system applies different power levels and handover parameters to different cells based on their local conditions and traffic patterns. Instead of uniform power adjustment across the network, each cell receives customized parameter optimization tailored to its specific coverage area, traffic density, and congestion patterns, achieving local optimality that balances coverage and capacity.
Data Source
AI summary
Optimal reduction of 4G LTE cellular network congestion utilizes two components of learning and optimization. First, an MLPDL learning approach is used to model cellular network congestion measured in terms of PRB utilization and predict 80% utilization as breakpoint thresholds of cellular towers as a function of average connected user equipments. Then, an optimization problem is formulated to minimize LTE network congestion subject to constraints of user quality and load preservation. Two alternative solutions, namely Block Coordinated Descent Simulated Annealing (BCDSA) and Genetic Algorithms (GA) are presented to solve the problem. Performance measurements demonstrate that GA offers higher success rates in finding the optimal solution while BCDSA has much improved runtimes with reasonable success rates. Accordingly, integrated iterative methods, programs, and systems are described aiming at minimizing the congestion of 4G LTE cellular networks by redistributing traffic from congested cellular towers to non-congested cellular towers.