DRL-Based Pre-Connect Handover for Low-Loss Mobility
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing handover procedures in wireless communication systems, such as those defined by 3GPP for 4G and 5G, suffer from handover failures due to poorly established parameter configurations and suboptimal target cell selection, leading to increased handover interruption time and packet loss.
Innovation Solution
Implementing a Pre-connect Handover (PHO) technique that uses deep reinforcement learning (DRL) for target cell selection based on RSRQ conditions and adjusts buffer capacity dynamically, combined with early data forwarding and buffering to ensure seamless handovers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional handover procedures are used, then handover operations are simple, but handover failures occur due to suboptimal target cell selection and poor parameter configurations
Solution Approach 1:
The patent applies preliminary action by performing target cell selection and resource allocation before the actual handover occurs. The network node identifies candidate target cells in advance, pre-allocates resources, and establishes pre-connections during stable radio conditions, so that when handover is triggered, the process can proceed smoothly without failure.
Solution Approach 2:
The patent implements self-service through machine learning models that automatically learn optimal handover parameters and target cell selection strategies from historical data. The system continuously improves its own performance by processing measurement reports and adjusting parameters without manual intervention, reducing handover failures while managing complexity automatically.
2Loss of time
If handover is performed quickly, then handover interruption time is reduced, but packet loss increases due to insufficient preparation
Solution Approach 1:
The patent resolves this contradiction by performing preliminary resource allocation and pre-connection establishment before handover is triggered. Data is forwarded to candidate target cells in advance, and resources are pre-configured, enabling the handover to execute quickly without causing packet loss since the target cell is already prepared to receive data immediately.
Solution Approach 2:
The patent applies beforehand cushioning by maintaining pre-connections with candidate target cells and pre-forwarding data during stable radio conditions. This creates a buffer of prepared resources and data at the target cell, cushioning against the potential packet loss that would otherwise occur during the handover transition.
3Reliability
If pre-connection is established during stable radio conditions, then handover success rate improves, but network resource consumption increases
Solution Approach 1:
The patent applies local quality by establishing pre-connections and allocating resources selectively only for specific candidate target cells that are most likely to be selected, rather than uniformly for all possible targets. The machine learning model identifies which cells deserve pre-preparation based on measurement reports and predicted handover likelihood, concentrating resources where they provide maximum benefit.
Solution Approach 2:
The patent uses parameter changes by dynamically adjusting pre-connection establishment parameters based on radio conditions and handover predictions. The system modifies which cells receive pre-connections, the duration of pre-connections, and resource allocation levels based on changing conditions, optimizing the balance between handover success rate and resource consumption.
Data Source
AI summary
A method, system and apparatus are disclosed. In at least one embodiment, a serving node is configured to communicate with a target node and a wireless device. The serving node is configured to cause transmission of wireless device data to the target node for storage in a buffer of the target node. The serving node is configured to cause, after transmission of the data, handover of the wireless device to the target node.