Network capacity planning considering the control plane bottleneck

A non-transitory machine-readable medium comprising executable instructions that, when executed by a processing system including a processor, facilitate performance of operations, the operations comprising:

• receiving first data associated with a first frequency band of a wireless network, wherein the first data characterizes utilization of a data plane of the wireless network at the first frequency band;
• receiving second data associated with a second frequency band of the wireless network, wherein the second data characterizes utilization of the data plane of the wireless network at the second frequency band, and wherein the second frequency band is different from the first frequency band;
• receiving third data associated with the first frequency band of the wireless network, wherein the third data includes first control plane statistics which characterize utilization of a control plane of the wireless network at the first frequency band;
• receiving fourth data associated with the second frequency band of the wireless network, wherein the fourth data includes second control plane statistics which characterize utilization of the control plane of the wireless network at the second frequency band;
• applying the first data and the third data to a first machine learning algorithm;
• determining, based on the first control plane statistics, a presence of a first control plane congestion condition associated with the first frequency band. wherein the first control plane statistics indicate a physical downlink control channel (PDCCH) utilization statistics and the presence of the first control plane congestion condition is determined based on the PDCCH utilization statistics exceeding a first overload threshold, and wherein the determining further comprises determining, using the first machine learning algorithm, that the first overload threshold is exceeded;
• generating, via the first machine learning algorithm, one or more first outputs, wherein the one or more first outputs indicates whether one or more first network resources associated with the first frequency band should be adjusted to improve data plane utilization at the first frequency band;
• applying the second data and the fourth data to a second machine learning algorithm, wherein the second machine learning algorithm is different from the first machine learning algorithm;
• determining, based on the second control plane statistics, a presence of a second control plane congestion condition associated with the second frequency band, wherein the second control plane statistics include physical downlink control channel (PDCCH) utilization statistics and the presence of the second control plane congestion condition is determined based on the PDCCH utilization statistics exceeding a second overload threshold;
• and wherein the determining further comprises determining, using the second machine learning algorithm, that the second overload threshold is exceeded;
• and
• generating, via the second machine learning algorithm, one or more second outputs, wherein the one or more second outputs indicates whether one or more second network resources associated with the second frequency band should be adjusted to improve data plane utilization at the second frequency band.

A method comprising:

• obtaining, by a processing system including a processor, first training data associated with a first frequency band of a wireless network, wherein the first training data characterizes utilization of a data plane of the wireless network at the first frequency band;obtaining, by the processing system, second training data associated with a second frequency band of the wireless network, wherein the second training data characterizes utilization of the data plane of the wireless network at the second frequency band, and wherein the second frequency band is distinct from the first frequency band;obtaining, by the processing system, third training data associated with the first frequency band of the wireless network, wherein the third training data includes first control plane statistics which characterizes utilization of a control plane of the wireless network at the first frequency band;obtaining, by the processing system, fourth training data associated with the second frequency band of the wireless network, wherein the fourth training data includes second control plane statistics which characterizes utilization of the control plane of the wireless network at the second frequency band;utilizing, by the processing system, the first training data and the third training data to train a first machine learning algorithm that is associated with the first frequency band;determining, based on the first control plane statistics, a presence of a first control plane congestion condition associated with the first frequency band, wherein the first control plane statistics include physical downlink control channel (PDCCH) utilization statistics and the presence of the first control plane congestion condition is determined based on the PDCCH utilization statistics exceeding a first overload threshold;
• and wherein the determining further comprises determining, using the first trained machine learning algorithm, that the first overload threshold is exceeded;utilizing, by the processing system, the second training data and the fourth training data to train a second machine learning algorithm that is associated with the second frequency band, wherein the second machine learning algorithm is distinct from the first machine learning algorithm;
• anddetermining, based on the second control plane statistics, a presence of a second control plane congestion condition associated with the second frequency band, wherein the second control plane statistics include physical downlink control channel (PDCCH) utilization statistics and the presence of the second control plane congestion condition is determined based on the PDCCH utilization statistics exceeding a second overload threshold;
• and wherein the determining further comprises determining, using the second trained machine learning algorithm, that the second overload threshold is exceeded.