Congestion control based inter-GNB carrier aggregation

1 . A method performed by a first network node comprising a primary cell gNodeB (PCell-gNB) having one or more carriers in a network, the method comprising:

defining a number of media access control (MAC) flows from the first network node that each forms a source end-point to a destination end-point, wherein the first network node comprises the source end-point and each destination end point comprises a second network node comprising a secondary cell gNB (SCell-gNB) having one or more destination carriers;

defining a MAC flow packet size that is scaled based on carrier bandwidth where each MAC flow packet has an approximately equivalent spectrum usage and an equivalent quality of service (QoS) on a destination carrier on the second network node; and

transmitting a MAC flow packet to the second network node, wherein the MAC flow packet comprises a bundle of one or more MAC sub-protocol data units (sub-PDUs).

2 . The method of claim 1 wherein defining the MAC flow packet size comprises defining the MAC flow packet size as a percent of a channel bandwidth of the MAC flow.

3 . The method of claim 1 , further comprising:

decoupling PCell and SCell scheduling decisions by separating hybrid automatic repeat request (HARQ) feedback opportunity separately into virtual physical uplink control channel (PUCCH) groups (VPGs) per network node and per User Equipment (UE).

4 . The method of claim 3 wherein separating HARQ feedback opportunity separately into VPGs comprises:

for each UE, defining one or more VPGs wherein each VPG is for all serving cells of a network node including one VPG associated with the first network node; and

mapping each VPG per UE to a unique uplink (UL) slot.

5 . The method of claim 4 further comprising:

for each newly-active connection with traffic towards a VPG-i, reserving one or more physical uplink control channel (PUCCH) resources;

indicating the one or more PUCCH resources reserved to a network node scheduler associated with the VPG-i; and

for each of the one or more PUCCH resources reserved, providing a minimum payload and a maximum payload of hybrid automatic repeat request acknowledgement (HARQ-ACK) uplink control information (UCI) to the network node scheduler associated with the VPG-i.

6 . The method of claim 1 , further comprising:

for a slot k to which virtual physical uplink control channel (PUCCH) group VPG-i is mapped, receiving an indication of a number of allocated hybrid automatic repeat request acknowledgement HARQ-ACK bits on the slot k from a VPG-i node and an indication of a PUCCH resource selected at an end of slot k.

7 . The method of claim 6 , further comprising:

responsive to receiving the number of allocated HARQ-ACK bits, setting up a PUCCH receiver or a physical uplink shared channel (PUSCH) receiver with a number of expected uplink control information (UCI) bits for decoding allocated resources on the PUCCH or the PUSCH.

8 . The method of claim 6 wherein a delay between the VPG-i node sending the indication of the number of allocated HARQ-ACK bits and receiving the indication is a one-way inter-node delay, the method further comprising setting up a PUCCH receiver for PUCCH transmission including ensuring that mink 1 for VPG-i is greater than the one-way inter-node delay where mink 1 is a minimum value of allowed k 1 for VPG-i.

9 . The method of claim 1 further comprising defining the MAC flow packet size for user equipment (UE) towards an SCell, wherein the MAC flow packet scales with UE-specific channel quality and a target number of physical resource blocks (PRBs) required based on a current estimation of channel state information (CSI).

10 . The method of claim 9 wherein defining the MAC flow packet size comprises defining the MAC flow packet size in accordance with

MAC

⁢

flow

⁢

packet

⁢

size

=

ICC

[

i

]

×

N

PRB

scell

×

n

RE

B

where ICC[i] is information carrying-capacity in units of bits per resource element (RE),

N

PRB

scell

is a number of PRBs in the SCell, n RE is an approximate number of Res per PRB available for physical downlink shared channel (PDSCH); and B is a parameter that controls the target-fraction of the SCell bandwidth that is needed to fit one MAC flow packet.

11 . The method of claim 1 , further comprising:

responsive to receiving data in at least one MAC flow, using quality of service (QoS) differentiation defined for each MAC flow to provide prioritization of traffic among each of the at least one MAC flow and local user traffic.

12 . The method of claim 1 , wherein defining a number of MAC flows comprises defining a MAC flow for each quality of service (QoS) level such that data from all user equipments (UEs) associated with the first network node of a defined QoS level are transported to a second network node having a destination SCell using a MAC flow corresponding to the defined QoS level.

13 . The method of claim 1 , further comprising performing congestion control on each MAC flow independent of other MAC flows.

14 . The method of claim 13 wherein performing congestions control for a MAC flow comprises:

tracking a number of MAC flow packets transmitted to a SCell in the MAC flow;

receiving flow control feedback from the SCell;

responsive to the flow control feedback indicating no congestion has been detected, increasing the number of MAC flow packets transmitted to the SCell in the MAC flow by one of a predetermined step size or an adaptively adjusted step size;

continually increasing the number of MAC flow packets transmitted to the SCell until flow control feedback is received indicating the MAC flow is congested or is starting to be congested; and

responsive to the flow control feedback being received that indicates the MAC flow is congested or is starting to be congested, decreasing the number of MAC flow packets by a predetermined number or by an adaptively adjusted number.

15 . A first network node comprising a primary cell gNodeB (PCell-gNB), the first network node comprising:

processing circuitry; and

memory coupled with the processing circuitry, wherein the memory includes instructions that when executed by the processing circuitry causes the processing circuitry to perform operations comprising:

defining a number of media access control flows from the first network node that each forms a source end-point to a destination end-point, wherein the first network node comprises the source end-point and each destination end point comprises a second network node comprising a secondary cell gNB (SCell-gNB) having one or more destination carriers;

defining a MAC flow packet size that is scaled based on carrier bandwidth where each MAC flow packet has an approximately equivalent spectrum usage and an equivalent quality of service (QoS) on a destination carrier on the second network node; and

transmitting a MAC flow packet to the second network node, wherein the MAC flow packet comprises a bundle of one or more MAC sub-protocol data units (sub-PDUs).