5G EN-DC: Flow Control Between 4G and 5G

One thing I’ve been wondering for some time now is how in 5G EN-DC, user data is split between the 4G and 5G base station in a way that takes changing RF conditions into account. It’s no good to just assign a certain data rate to the 4G and 5G side when a split bearer is initially established, as achievable speeds over the air interface can change very quickly. In other words, there must be some sort of throttling mechanism.

So I went a bit through the specs and finally found what I was looking for. In 3GPP TS 36.300 (E-UTRAN overall description), chapter 20.1.1 says the following:

Flow Control Functions: The Flow Control function is applied when an E-RAB is configured for the split bearer option and only for DL i.e. the flow control information is provided only by the SeNB to the MeNB for the MeNB to control the downlink user data flow to the SeNB. The flow control function is further described in TS 36.425.

Very good, so let’s have a look at TS 36.425 which is the specification for the X2 interface (between radio nodes), User Plane Protocol. Here it says in chapter 5.4.2:

Downlink Data Delivery Status: The purpose of the Downlink Data Delivery Status procedure is to provide feedback from the SeNB to the MeNB to allow the MeNB to control the downlink user data flow via the SeNB for the respective E-RAB.

In other words, The master cell forwards user data packets to the secondary cell and adapts the data rate depending on the delivery status reports that are returned. There we go, another mystery solved!

P.S.: You might have noticed that the quotes in this post are from 3GPP Release 14 of the spec which do not include NR. I did this on purpose because Dual Connectivity existed before Release 15 NR and the concept must have already existed before. Older specs are smaller and easier to read…