US 10,764,959 B2 · Communication system of quality of experience oriented cross-layer admission control and beam allocation for functional-split wireless fronthaul communications

Abstract

Two algorithms for different functional-split network models are provided, which are the CU-based (central unit based) beam allocation and admission control algorithm (CU-BAACA) and the DU-based (distributed unit based) beam allocation and admission control algorithm (DU-BAACA). Difference between the CU-BAACA and DU-BAACA includes whether the algorithm is implemented at the CU site or at the DU site. Proposed algorithms aim to optimize DUs' quality of experience (QoE) by admission control to determine the amount of data from application layer entering into traffic queue and allocating beam in physical layer at the fronthaul network between CU and DUs. On the other hand, queueing delay and queue stability are taken into consideration to maintain the system steadiness. Simulation results compare performance of two functional split models to find the appropriate scenario for each function split option, which provide technical requirement and applicability of the proposed algorithms for practical system.

A communication system for quality of experience oriented cross-layer admission control and beam allocation for functional-split wireless fronthaul communications for a cloud radio access network, the system comprises:

a central unit, having a portion of a plurality of functional-split protocol layers of the cloud radio access network;
and
a plurality of distributed units, wirelessly communicating with the central unit and having another portion of the plurality of functional-split protocol layers of the cloud radio access network;
wherein the central unit comprising:
an admission controller, configured to receive data of a plurality of data sources;
a virtual machine, connected to the admission controller and configuring the data of the plurality of data sources as a plurality of data queues;
a beam allocator, connected to the virtual machine and managing a plurality of beam resources to transmit data of the plurality of data queues to the plurality of distributed units;
and
a processor, connected to the admission controller, the virtual machine and the beam allocator and receiving a channel state information, a data arrival and a queue state information;
wherein the channel state information is from each of the plurality of distributed units, the data arrival is from the plurality of data sources, and the queue state information is from the plurality of data queues of the virtual machine;
wherein the processor further operates a central unit based beam allocation and admission control algorithm in order to compute a data transmission rate of the plurality of beam resources and a data bulk tolerable for capacity of the communication system in accordance with the channel state information, the data arrival and the queue state information so as to control the beam allocator to allocate the plurality of beam resources and control the admission controller to control data entering the plurality of data queues;
wherein in the central unit based beam allocation and admission control algorithm, a dynamic equation of the queue state information is expressed as:
Qk(t+1)=[Qk(t)−Rk(t)T]++Ck(t)T
wherein Qk(t) is a length of a data queue transmitted to a k-th distributed unit in a t-th time slot, Rk(t) is a data transmission rate allocated to the k-th distributed unit in the t-th time slot, Ck(t) is a data bulk in the t-th time slot, T is a length of time of a single time slot, and [x]+ is expressed as max(x,0);
wherein a time-averaged queue delay dk of a k-th queue of the plurality of data queues is expressed as:
d
k
_
=
lim
M
→
∞
⁢
1
M
⁢
∑
t
=
1
M
⁢
𝔼
⁢
{
Q
k
⁡
(
t
)
}
lim
M
→
∞
⁢
1
M
⁢
∑
t
=
1
M
⁢
𝔼
⁢
{
C
k
⁡
(
t
)
}
wherein M represents a total number of time slots;
wherein entire time-averaged quality of experience (QoE) of the distributed units is expressed as:
QoEk=log2(Ck(t)T)
wherein the central unit based beam allocation and admission control algorithm is used to solve an optimization question as:
max
ϕ
n
,
k
⁡
(
t
)
,
C
⁡
(
t
)
⁢
lim
M
→
∞
⁢
1
M
⁢
∑
t
=
1
M
⁢
∑
k
=
1
K
⁢
𝔼
⁢
{
log
2
⁡
(
C
k
⁡
(
t
)
⁢
T
)
}
with conditions listed here:
C
⁢
⁢
1
⁢
:
⁢
⁢
∑
k
=
1
K
⁢
ϕ
n
,
k
⁡
(
t
)
≤
1
C
⁢
⁢
2
⁢
:
⁢
⁢
ϕ
n
,
k
⁡
(
t
)
∈
{
0
,
1
}
C
⁢
⁢
3
⁢
:
⁢
⁢
d
k
_
≤
d
k
max
C
⁢
⁢
4
⁢
:
⁢
⁢
0
≤
C
k
⁡
(
t
)
≤
A
k
⁡
(
t
)
C
⁢
⁢
5
⁢
:
⁢
⁢
Queues
⁢
⁢
Q
k
⁡
(
t
)
⁢
⁢
are
⁢
⁢
mean
⁢
⁢
rate
⁢
⁢
stable
wherein C(t) represents a set of Ck(t), condition C1 represents that each of the beam resources is allocated to only a single one of the distributed units, condition C2 represents that ϕn,k(t) indicates whether or not one of the beam resources is allocated to the one of the distributed units, condition C3 represents a restriction of the time-averaged queue delay, condition C4 represents a restriction of the admission control, and condition C5 represents that all of the data queues of the communication system must be mean rate stable to maintain a stability of the communication system.

236 more elements

Communication system of quality of experience oriented cross-layer admission control and beam allocation for functional-split wireless fronthaul communications

Life of this patent

Communication system of quality of experience oriented cross-layer admission control and beam allocation for functional-split wireless fronthaul communications

Life of this patent

Abstract

Claims11