For example, when downlink data is transmitted from DgNB(a) to RN(b) while RN(b) transmits downlink data to RN(c) in the example of 
Accordingly, directions between links connected to nodes need to be aligned. According to this rule, only RN(c) and UE2 can transmit signals to RN(b) when RN(b) receives a signal from DgNB(a). Here, RN(c) can perform only transmission to UE3.
On the other hand, when RN(b) transmits an uplink signal to DgNB(a), RN(b) can transmit a downlink signal to RN(c) and UE2 and RN(c) can perform only reception from UE3 and RN(d).
When an RN simultaneously processes multiple links, symbol boundaries need to be carefully processed. If not, performance may be considerably deteriorated.
Assume that access links of other nodes have the same Tx/Rx timing.
In this case, a DgNB can set the same backhaul link Tx/Rx timing as the access link timing. However, RN1 can receive a DL signal from the DgNB at a delayed timing and transmit a UL signal to the DgNB as a configured TA value. In this case, it is difficult for RN1 to process symbol boundaries, which causes interference in both the access link and backhaul link.
Since problems may occur even when the access link can be flexibly scheduled, the network needs to perform appropriate scheduling/coordination between backhaul/access links.
<CLI (D/U Alignment Between Access Links) for UE>