Dynamic self-optimization of application quality-of-service requests to mobile networks
Systems and methods for network bandwidth optimization, including transmitting sensor data from one or more sensors over a wireless network into a generated network slice, submitting a Quality-of-Service (QoS) request for one or more applications by specifying desired network slice characteristics, and predicting network bandwidth needed for granting the QoS request for the one or more applications using a cost function based on magnitude, direction, and frequency of error. Time-varying network bandwidth usage is continuously monitored, and new QoS requests for the one or more applications are periodically requested based on the monitoring. An updated prediction for updated bandwidth needed for the new QoS request is generated using the cost function, and network bandwidth reservations are iteratively adjusted based on the updated prediction for the new QoS request to provide an amount of network resources to the one or more applications to support the new QoS request.
1. A method for network bandwidth optimization comprising:
transmitting sensor data from one or more sensors over a wireless network into a generated network slice;
submitting a Quality-of-Service (QoS) request for one or more applications by specifying desired network slice characteristics;
predicting network bandwidth needed to grant the QoS request for the one or more applications using a cost function based on magnitude, direction, and frequency of error;
continuously monitoring time-varying network bandwidth usage and periodically issuing new QoS requests for the one or more applications based on the monitoring;
generating an updated prediction for an updated bandwidth needed for the new QoS request using the cost function; and
iteratively adjusting network bandwidth reservations based on the updated prediction for the new QoS request to provide an amount of network resources to the one or more applications to support the new QoS request.
2. The method as recited in claim 1 , wherein the wireless network is a 5G network.
3. The method as recited in claim 1 , wherein the predicting the network bandwidth needed for granting the QoS request further comprises identifying any oversubscription or undersubscription of network resources to optimally balance application accuracy and network operational cost.
4. The method as recited in claim 1 , wherein the cost function is represented as follows:
C
total
=
∑
t
∈
T
F
u
×
(
GBR
t
-
A
t
)
⊗
P
u
+
∑
t
∈
T
F
o
×
(
A
t
-
GBR
t
)
⊗
P
o
,
where C total represents a total cost, t represents a time unit in a total duration T, F u represents a flag for undersubscription, F o represents a flag for oversubscription, A t represents actual value, GBR t represents reserved guaranteed bandwidth, P u represents a penalty cost for undersubscription, and P o represents a penalty cost for oversubscription.
5. The method as recited in claim 3 , wherein the undersubscription is identified by an actual value A t being less than a reserved guaranteed bandwidth GBR t , and the oversubscription is identified by the actual value A t being greater than the reserved guaranteed bandwidth GBR t .
6. The method as recited in claim 1 , wherein the generated network slice is a single network slice and is isolated from other network slices upon generation of the single network slice.
7. The method as recited in claim 3 , wherein the updated prediction comprises selecting a highest magnitude of the network bandwidth detected during the monitoring as a baseline, adding a mean of a magnitude of the oversubscription to the baseline if an upward trend is detected during the monitoring, and subtracting a mean of a magnitude of the undersubscription to the baseline if a downward trend is detected during the monitoring.
8. A system for network bandwidth optimization comprising:
one or more processors operatively coupled to one or more computer-readable storage mediums, the processors being configured for:
transmitting sensor data from one or more sensors over a wireless network into a generated network slice;
submitting a Quality-of-Service (QoS) request for one or more applications by specifying desired network slice characteristics;
predicting network bandwidth needed for granting the QoS request for the one or more applications using a cost function based on magnitude, direction, and frequency of error;
continuously monitoring time-varying network bandwidth usage and periodically issuing new QoS requests for the one or more applications based on the monitoring;
generating an updated prediction for an updated bandwidth needed for the new QoS request using the cost function; and
iteratively adjusting network bandwidth reservations based on the updated prediction for the new QoS request to provide an amount of network resources to the one or more applications to support the new QoS request.
9. The system as recited in claim 8 , wherein the wireless network is a 5G network.
10. The system as recited in claim 8 , wherein the predicting the network bandwidth needed for granting the QoS request further comprises identifying any oversubscription or undersubscription of network resources to optimally balance application accuracy and network operational cost.
11. The system as recited in claim 8 , wherein the cost function is represented as follows:
C
total
=
∑
t
∈
T
F
u
×
(
GBR
t
-
A
t
)
⊗
P
u
+
∑
t
∈
T
F
o
×
(
A
t
-
GBR
t
)
⊗
P
o
,
where C total represents a total cost, t represents a time unit in a total duration T, F u represents a flag for undersubscription, F o represents a flag for oversubscription, A t represents actual value, GBR t represents reserved guaranteed bandwidth, P u represents a penalty cost for undersubscription, and P o represents a penalty cost for oversubscription.
12. The system as recited in claim 10 , wherein the undersubscription is identified by an actual value A t being less than a reserved guaranteed bandwidth GBR t , and the oversubscription is identified by the actual value A t being greater than the reserved guaranteed bandwidth GBR t .
13. The system as recited in claim 8 , wherein the generated network slice is a single network slice and is isolated from other network slices upon generation of the single network slice.
14. The system as recited in claim 10 , wherein the updated prediction comprises selecting a highest magnitude of the network bandwidth detected during the monitoring as a baseline, adding a mean of a magnitude of the oversubscription to the baseline if an upward trend is detected during the monitoring, and subtracting a mean of a magnitude of the undersubscription to the baseline if a downward trend is detected during the monitoring.
15. A non-transitory computer readable storage medium comprising a computer readable program operatively coupled to a processor device for network bandwidth optimization, wherein the computer readable program when executed on a computer causes the computer to perform the steps of:
transmitting sensor data from one or more sensors over a wireless network into a generated network slice;
submitting a Quality-of-Service (QoS) request for one or more applications by specifying desired network slice characteristics;
predicting network bandwidth needed for granting the QoS request for the one or more applications using a cost function based on magnitude, direction, and frequency of error;
continuously monitoring time-varying network bandwidth usage and periodically issuing new QoS requests for the one or more applications based on the monitoring;
generating an updated prediction for an updated bandwidth needed for the new QoS request using the cost function; and
iteratively adjusting network bandwidth reservations based on the updated prediction for the new QoS request to provide an amount of network resources to the one or more applications to support the new QoS request.
16. The non-transitory computer readable storage medium as recited in claim 15 , wherein the wireless network is a 5G network.
17. The non-transitory computer readable storage medium as recited in claim 15 , wherein the predicting the network bandwidth needed for granting the QoS request further comprises identifying any oversubscription or undersubscription of network resources to optimally balance application accuracy and network operational cost.
18. The non-transitory computer readable storage medium as recited in claim 15 , wherein the cost function is represented as follows:
C
total
=
∑
t
∈
T
F
u
×
(
GBR
t
-
A
t
)
⊗
P
u
+
∑
t
∈
T
F
o
×
(
A
t
-
GBR
t
)
⊗
P
o
,
where C total represents a total cost, t represents a time unit in a total duration T, F u represents a flag for undersubscription, F o represents a flag for oversubscription, A t represents actual value, GBR t represents reserved guaranteed bandwidth, P u represents a penalty cost for undersubscription, and P o represents a penalty cost for oversubscription.
19. The non-transitory computer readable storage medium as recited in claim 17 , wherein the undersubscription is identified by an actual value A t being less than a reserved guaranteed bandwidth GBR t , and the oversubscription is identified by the actual value A t being greater than the reserved guaranteed bandwidth GBR t .
20. The non-transitory computer readable storage medium as recited in claim 17 , wherein the updated prediction comprises selecting a highest magnitude of the network bandwidth detected during the monitoring as a baseline, adding a mean of a magnitude of the oversubscription to the baseline if an upward trend is detected during the monitoring, and subtracting a mean of a magnitude of the undersubscription to the baseline if a downward trend is detected during the monitoring.