LTE eNB Load Balancing via Dynamic Session Assignment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Long Term Evolution (LTE) eNBs face challenges in evenly distributing and balancing incoming traffic load due to varying user equipment (UE) activity, leading to performance issues, especially with increased throughput demands from advancements like Carrier Aggregation and advanced MIMO techniques.
Innovation Solution
A controller with multiple processing engines and a distributor/load balancer that assigns new UE sessions to the least loaded processing engine, dynamically adjusting based on processing load differentials, entering a second mode to redirect sessions when the load differential exceeds a threshold, ensuring efficient load balancing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If new UE sessions are assigned to the PE with the fewest sessions (first mode), then session distribution is improved, but processing load balance deteriorates when traffic patterns vary significantly
Solution Approach 1:
The system dynamically switches between two load balancing modes based on real-time traffic conditions. In the first mode, new UE sessions are assigned to PEs with the fewest sessions to ensure even distribution. In the second mode, sessions are assigned to the least loaded PE based on current processing load metrics. This dynamic adaptation resolves the contradiction by adjusting the assignment strategy according to actual system state, maintaining both session distribution and processing load balance under varying traffic conditions.
2Productivity
If load balancing mode switching is implemented, then processing efficiency is improved, but system complexity increases
Solution Approach 1:
The system pre-establishes two distinct load balancing modes with clearly defined switching criteria. The controller is configured in advance to monitor processing load differentials and automatically transition between modes when thresholds are exceeded. This preliminary configuration simplifies the control logic by providing a predetermined decision framework, avoiding the need for complex real-time optimization algorithms while still achieving improved processing efficiency through adaptive load balancing.
Data Source
AI summary
A controller that processes signals in a baseband frequency band is described. The controller includes a plurality of processing engines (PEs). The controller also includes a distributor/load balancer that includes at least one processor or processing core. The at least one processor or processing core is configured to (1) enter a first mode in which each new UE session is assigned to a PE that is currently processing a fewest number of UE sessions; (2) determine whether a processing load differential between a least loaded PE and a highest loaded PE exceed a first threshold for at least a time threshold; and (3) in response to the processing load differential between the least loaded PE and the highest loaded PE exceeding the first threshold for at least the time threshold, enter a second mode in which each new UE session is assigned to the least loaded PE.