US 12,309,640 B2
Predictive user plane function (UPF) load balancing based on network data analytics
Kazi Bashir, Lewisville, TX (US); Mehdi Alasti, Reston, VA (US); and Siddhartha Chenumolu, Broadlands, VA (US)
Assigned to DISH Wireless L.L.C., Englewood, CO (US)
Filed by DISH Wireless L.L.C., Englewood, CO (US)
Filed on Feb. 16, 2024, as Appl. No. 18/443,872.
Application 18/443,872 is a continuation of application No. 17/529,128, filed on Nov. 17, 2021, granted, now 11,950,138.
Prior Publication US 2024/0196273 A1, Jun. 13, 2024
This patent is subject to a terminal disclaimer.
Int. Cl. H04W 28/02 (2009.01); H04W 28/08 (2023.01); H04W 28/082 (2023.01)
CPC H04W 28/0942 (2020.05) [H04W 28/082 (2023.05)] 16 Claims
OG exemplary drawing
 
1. A method comprising, comprising:
receiving, by a session management function, a request to anchor a protocol data unit session of a new user equipment;
predicting, by the session management function over a series of times for each respective user plane function in a plurality of user plane functions that serve as anchor points between user equipment in a cellular telecommunication network and a data network, a respective load of the respective user plane function at each respective time in the series of times to generate a series of predicted loads;
selecting, by the session management function, a user plane function from the plurality of user plane functions based on an analysis of factors that includes the series of predicted loads for each respective user plane function in the plurality of user plane functions; and
anchoring, by the session management function, the protocol data unit session of the new user equipment on the selected user plane function;
wherein:
the method further comprises generating an average, for each respective user plane function in the plurality of user plane functions that serve as anchor points between user equipment in the cellular telecommunication network and the data network, of the series of predicted loads;
the factors include the average of the series of predicted loads for each respective user plane function in the plurality of user plane functions; and
generating the average of the series of predicted loads comprises generating a weighted average.