How to maintain high performance radio connections during mobility

Mobility is at the heart of cellular networks, we expect services to work also when a mobile device is moving. 5G enables new high-performance services that require very low latency and large bandwidth. This is possible thanks to the introduction of new frequency bands in the mid- and high spectrum range, where the higher frequencies unlock entirely new bandwidth capabilities and much faster speeds.

These higher frequencies however provide limited coverage and performance degrades faster when the mobile device is moving, which leads to higher risk for disruption of services.

A question then is how to maintain the radio connections on the higher frequency bands when the mobile device is moving, to ensure service continuity.

Use of multiple carriers

A way to take advantage of both the higher data rates provided by higher frequency bands and the robustness of the lower frequency bands is to have simultaneous radio connections between the mobile device and the network both on higher and lower frequency bands. This can in 5G be achieved either by Carrier Aggregation or by Dual Connectivity.

In Carrier Aggregation, the mobile device is configured with multiple carriers (frequencies) towards one node in the network – the base station. A base station terminates user plane and control plane protocols over the radio connection towards the mobile device. All the carriers for the mobile device are then controlled by this base station, including radio connection maintenance during mobility.

Dual connectivity

Dual connectivity allows a UE to simultaneously transmit and receive data on multiple carriers from two serving nodes (a master node, MN, and a secondary node, SN).

Multiple cells can be available for both nodes, where each cell corresponds to a radio connection on a separate frequency between the mobile device and the network. To the MN, the device is connected through a group of cells called a Master Cell Group (MCG), and to the SN through a Secondary Cell Group (SCG). Both groups consist of a primary cell (PCell and PSCell, respectively) and possibly one or more additional, or secondary, cells (SCells) for additional capacity.

Figure 1 shows an example of a mobile device in Dual Connectivity.

Figure 1: Example of a mobile device in Dual Connectivity.

In this example, the mobile device has simultaneous radio connections to two separate network nodes, through a PCell controlled by the MN and a PSCell controlled by the SN.

The cell(s) in the MCG typically use lower frequency bands that provide a more robust radio connection. The control signaling between the mobile device and the network, for example, is mainly handled by the MN and transmitted over the MCG. The cell(s) in the SCG instead typically use higher frequency bands and are suitable for services requiring very low latency and large bandwidth.

With two different base stations handling simultaneous radio connections to the same mobile device, there is a need for coordination between them. Such coordination includes for example what type of radio connections each one can use and how to change radio connection when the mobile device is moving.

Change of cell at mobility

As a mobile device moves in the network, sometimes a change of radio connection is needed to keep the device on the best possible network connection. The change is either to another cell in the same base station or to a cell in a different base station, and it is valid both for the MCG and for the SCG.

The MCG/PCell cell changes are referred to as handovers. For a mobile device in Dual Connectivity, such a handover often means that also the SCG configuration, meaning the radio connections on the higher frequencies, needs to be updated. 

The different cell changes are controlled by the network but they are normally based on information from the mobile device. If the radio connection to the current cell is getting degraded and/or if another neighboring cell is becoming better than the current cell, the mobile device sends a measurement report to the network indicating this. The network then prepares a radio connection to the neighboring (target) cell and instructs the mobile device to move there. That way high performance and interruption free radio conditions can be maintained for the mobile device. 

The below figure shows a handover procedure from the source cell to a target cell.

Figure 2 Signaling at a handover procedure

The signaling that takes place after the need for a cell change is detected until the actual execution is started takes time. This delays the actual cell change. If the radio connection to the cell where the signaling is performed (the PCell) is degrading, there is also a risk that the signaling fails. In that case, the radio connections for both the MCG and the SCG will typically be lost, leading to an interruption of the entire data connection for the mobile device.

Conditional handover

A conditional mobility procedure, the “Conditional Handover”, was introduced in 3GPP Rel-16 to increase mobility robustness. Read more in this blog post: This is the key to mobility robustness in 5G networks. 

In this procedure, the handover instruction for a target PCell is sent to the mobile device in advance, together with associated execution conditions. The device then performs the handover as soon as those execution conditions are fulfilled, which avoids the need for signaling towards the network.

In the 3GPP Rel-16 specification, however, no SCG was supported with conditional handover, meaning that the SCG would be released at execution of conditional handover. That then would cause interruption to those high frequency radio connections.

Conditional handover with SCG in 3GPP Rel-17 and Rel-18

One part of achieving service continuity for the high performing services is to maintain the radio connections on high frequencies during the robust mobility procedures for the low frequencies. This decreases the risk for interruptions for those services when there is mobility for the MCG/PCell. Therefore, support for an SCG configuration (a target PSCell) in the pre-configured conditional handover was introduced in Rel-17. This is illustrated in Figure 3. 

Figure 3 Conditional handover with target PSCell

When the associated execution conditions for the candidate PCell are fulfilled, and the mobile device performs the handover procedure, a radio connection to a PSCell after the handover is then included. This way, it’s possible to maintain the previous PSCell/SCG configuration, or setup a new one, thus minimizing interruption for the high frequency connections at the handover.

The conditional handover configurations in the Rel-17 solution only include execution conditions associated to the target PCell. This means that the mobile device only checks radio conditions of the target PCell to determine when to execute the handover procedure, even if the configuration also includes a target PSCell. Without any evaluation of the target PSCell, there is a risk that the radio conditions for that are not good. There can be some time from the configuration of the conditional handover until it is executed, which increases the risk that the included target PSCell is in bad coverage at the time of execution. In summary, there is a risk that the mobile device will have a bad radio connection to the target PSCell, or even suffer from an interruption for the SCG, after the conditional handover execution.

To overcome this, a solution was introduced in 3GPP Rel-18 where the mobile device receives and evaluates conditions for both the target PCell and the target PSCell. Conditional handover is then only executed when the conditions for both cells are fulfilled. The network can provide the device with multiple conditional handovers for the same candidate PCell, but with different candidate PSCells. This is illustrated in Figure 4, where each combination of candidate PCell and candidate PSCell is included in a separate conditional handover configuration. 

Figure 4 Conditional handover with different candidate PSCells for the same candidate PCell

The mobile device will only execute the conditional handover, including the SCG configuration, when both the PCell and the PSCell are in good radio conditions. A robust mobility on the lower frequency (MCG), while maintaining the radio connection on higher frequencies (SCG) with minimized risk for degradation or interruptions, is thereby possible.

For the case that the mobile device does not have good radio conditions for any of the candidate PSCells when the execution conditions for the PCell are fulfilled, the network can also provide a separate conditional handover configuration for the same candidate PCell without any PSCell execution condition. This prevents the mobile device from staying too long in the source PCell, which could lead to a loss of the entire data connection to the network.

Conditional PSCell change in 3GPP Rel-16 and Rel-17

The shorter coverage on higher frequencies means that cell changes are much more frequent for the SCG (PSCell) than for the MCG (PCell). Another important part of achieving service continuity for the high performing services is therefore to avoid interruptions and degradation when there is a need to change the radio connection for the SCG.

With the higher level of fluctuations on high frequency bands, there is a risk that the performance degrades quicker as the mobile device is moving. Since there is a delay from the point in time where the mobile device detects the need to change cell until it has sent a measurement report to the network and got back the instruction to change cell, there is a risk that the radio connection gets disrupted before the PSCell change is performed.

In 3GPP Rel-16 and Rel-17 conditional PSCell change procedures have therefore been introduced to overcome the delay caused by this signaling. Similar to the conditional handover, the mobile device receives an instruction to perform a PSCell change in advance and stores it together with associated execution conditions. When the execution conditions are fulfilled, the mobile device here performs a PSCell change. The risk for performance degradation, or even loss of connection, in the serving PSCell can thereby be avoided. This enables service continuity on the higher frequencies also when there is a need for cell changes for the SCG. Multiple conditional configurations (up to 8) can be configured, where an example of two candidate PSCells can be seen in Figure 5. 

Figure 5 Mobile device configured with candidate PSCells for a conditional PSCell change

3GPP Rel-16 introduced support for conditional PSCell changes between different cells in the same Secondary Node (SN). In 3GPP Rel-17 support for conditional PSCell changes between cells in different SNs was added.

PSCell change enhancements in 3GPP Rel-18

The conditional PSCell change procedures avoids the need for signaling of measurement reports and cell change configuration when the need for cell change is detected. There is however a need to signal the conditional configurations to the mobile device in advance. In 3GPP Rel-17, the mobile device deletes the stored conditional configurations after each cell change and receives new ones in the new network cell/node. 

This can cause a lot of signaling to the mobile device considering that the cell changes can be very frequent for the SCG on high frequency bands. It also means that, after execution of a PSCell change, there is a delay until it is possible to trigger the next following PSCell change for the mobile device.

In 3GPP Rel-18 the conditional PSCell change has therefore been further enhanced to avoid the need to send a new configuration to the mobile device after a PSCell change. The mobile device then keeps the stored conditional PSCell change configurations, and continues to use them, when it changes cell. This decreases the signaling load due to frequent cell changes as well as the delay until the next PSCell change can be triggered. To allow the network to still decide which candidate PSCells that are relevant, and the associated execution conditions, depending on the current PSCell for the mobile device, the mobile device is also pre-configured with this information. This is illustrated in Figure 6, where the mobile device has different candidate PSCells and associated execution conditions for those in the different PSCells.

Figure 6 Mobile device in Dual Connectivity with Rel-18 candidate PSCell configurations.

Each PSCell configuration includes associated execution conditions for other relevant candidate PSCell configurations. The Rel-18 enhancements are useful for scenarios that involve very frequent PSCell transitions within a limited set of cells. This can include scenarios where the mobile device moves back and forth between a limited set of cells in an area or along a path where the future cell changes can be predicted.

Improving mobility

Dual connectivity is one way to benefit from both the high performance of higher frequency bands, and the robustness of lower frequencies. Dual connectivity comes with specific challenges when devices move in the network, since there are two base station environments to consider. 

From Rel-16 to Rel-18, 3GPP has added support in the specifications to make cell changes smooth and interruption free for devices in dual connectivity.
3GPP Rel-18 also introduces a new way of doing a cell change for a mobile device, called L1/L2 Triggered Mobility (LTM), which enables a shorter connection interruption time during the cell change. The shorter interruption time is achieved by the mobile device synchronizing to the target cell before the actual cell change and that lower-layer signaling is used between the network and the mobile device for controlling the cell change. LTM is supported for the higher frequency bands, including for SCG cell changes. Please check this blog post to find more information about LTM: 5G Advanced handover: L1/L2 Triggered mobility - Ericsson.