Explore Our Knowledge Base Covering A Range Of Topics

Minimizing Handover Failures in 5G

Written by Philip Nugent | Oct 10, 2022 11:00:00 AM

Whilst working with various 5G service providers, it has been great to see them incorporate enhanced features to minimize handover failures in their 5G Standalone deployments. One such feature relates to the term “conditional handover”. 

There are numerous reasons why a handover may fail, for example poor coverage, fast mobility or poor configuration. Generally, when performing a handover, the decision is made either too late, perhaps too early and sometimes even to the wrong cell. Whilst the 5G specifications include mechanisms to re-establish the connection, this does not mitigate the potential impact to the user’s service.  

Figure 1 illustrates a scenario whereby the device is handed over to a wrong cell. In this scenario the handover from source gNB to target gNB is unsuccessful and the connection quickly drops. This leaves the device having to re-establish the connection to a new gNB (the actual target). 

Figure 1 Example of Handover Failure (Wrong Cell)

Conditional Handover

The latest enhancements to 5G include a feature known as “conditional handover”. This is designed to mitigate the potential issue illustrated in Figure 1. In summary, rather than the device having only one option/target, conditional handover enables the source cell to configure multiple targets. These are then linked to new measurements the device is performing, enabling the device to make the final decision based on the radio conditions and additional triggers.

Figure 2 illustrates the intial Measurement Report message from the device. This may indicate a potential target cell or cells. In this scenario, the source gNB knows that typically the handover may fail and the device usually ends up on one of 2 candidates. It therefore sends a XnAP Handover Request messages to both target cells (one on each gNB). Next, admission control is perfomed and both cells respond with the Handover Request Acknowledge message. Each contains a RRC Reconfiguration message, i.e. the handover message. The source gNB then sends the RRC Reconfiguration message which contains information/configuration for both target cells. At this point the device sends the RRC Reconfiguration Complete message back to the source gNB and then begins monitoring (no handover has been performed yet).

Figure 2 Conditional Handover (Part 1)

As part of the handover execution phase, the device will monitor the potential targets to see which cell triggers the “execution condition”. This execution condition is in the form of specific measurement event triggers.

Figure 3 Conditional Handover (Part 2)

The device then accesses the chosen target, which triggers the XnAP Handover Success message to the source. In so doing, the source gNB can send the SN (Sequence Number) Status Transfer to the new target gNB, as well as cancelling the handover towards the gNB which was not chosen. Finally, the new target gNB will perform a N2 Path Switch procedure.

Overall, conditional handovers enable a faster decision with pre-configured target options. In so doing, minimizing some of the handover failure scenarios.