What is Handover Interruption Time?

Handover Interruption Time (HIT) is the shortest period of time supported by a cellular mobile network during the handover process from one node to the next. When a handover is carried out from a source cell (existing) to a target (next) cell, a finite amount of time exists during which the mobile user can neither transmit nor receive the data. This time is called Handover Interruption Time.

In a typical 4G LTE cellular deployment scenario, the handover interruption time is between 30 to 60 ms and can possibly attain the upper limit depending on the handover scenario, density of the network, and other radio conditions. Future cellular networks face a big challenge to further reduce the handover interruption time.

As far as the progress towards advancing latency-free networks is concerned, 3GPP releases such as the 16 and 17 have introduced new features and functionalities to enhance the network operations and lower latency and reduce HIT in applications such as drones in smart manufacturing, autonomous vehicles that are connected, electrical power distribution and much more. These and other potential applications require a low latency networking and ultra high reliability in the range of a few milli-seconds.

In today’s mobile networks, the mobile user usually releases a connection to the source cell before the link is established with the target cell. In other words, both uplink and downlink exchange of data is confirmed in the source cell before mobile user initiates a communication with the target cell. As the mobile user moves towards the edge of the cell and the mobile is not formally exchanging information with the target cell, this inevitably leads to an interruption during the handover in the range of a few tens or even hundreds of milliseconds. 

There are a number of ways to reduce HIT. One of the techniques to reduce the HIT time to milliseconds is detailed below.

Dual Active Protocol Stack (DAPS)

In this method, the mobile user will allow the connection to the source cell to continue to remain active until it is able to effectively transmit and receive data in the target cell.

New Features of the DAPS Method:

• Continued exchange of data in the source cell upon the reception of handover request.
• Able to simultaneously receive information from both source and target cell, and
• Upon completion of the random access procedure, the uplink transmission of data is switched to the target cell.

The mobile user continues to exchange information with the source cell upon receiving the handover request from the source cell. In parallel, a new connection to the target cell is established and hence the user will perform synchronization and random access procedure in the target cell. Due to the new connection, a new user plane protocol stack will be executed for the target cell. This stack contains PHY (physical layer), MAC (Medium Access Control layer), and RLC (Radio Link Control layers). While this process is carried out, the existing user plane protocol stack for source cell remains active for allowing continuous transmission and reception of data with the same.

However, in order to be able to simultaneously exchange data with both the source and target cell, the Packet Data Convergence Protocol (PDCP) should be reconfigured such that a common PDCP entity exists where the user planes for both the source as well as target cell is present. To receive packets in sequence for efficient reordering, a PDCP sequence number continuation is maintained throughout the handover procedure which is provided in the common PDCP entity for both the source and target cell. Other functions such as encryption/decryption and header compression and decompression are handled separately in the PDCP entity.

During the random access procedure, user data from the 5G core is relayed from the source to the target cell so that the target cell is enabled for downlink transmissions while the user is enabled for uplink transmissions. This implies that the same information is transferred to the mobile user too.

Upon completion of this random access procedure, the uplink transmission will be switched from the source cell to the target cell and the mobile user informs the target node of the last received user data from the source cell. This information is useful for target cell so that it will not send duplicate or repeating packets in the downlink. Once the target node acknowledges the request from the mobile user, it will inform the source node of the successful establishment of link and hence will trigger the source node to stop further uplink transmissions to the mobile user. This reduces the handover interruption time compared to the single active protocol technique in that mobile user will continue to remain active in the source cell until the random access procedure is completed with the target cell. Although, an inevitable latency exists, this latency will be relatively less in the order of a few milliseconds.