US 12,143,861 B2
Efficiency of routing traffic to an edge compute server at the far edge of a cellular network
Alastair Wolman, Seattle, WA (US); Paramvir Bahl, Bellevue, WA (US); and Landon Prentice Cox, Seattle, WA (US)
Assigned to Microsoft Technology Licensing, LLC, Redmond, WA (US)
Filed by Microsoft Technology Licensing, LLC, Redmond, WA (US)
Filed on Oct. 13, 2021, as Appl. No. 17/500,441.
Prior Publication US 2023/0110752 A1, Apr. 13, 2023
Int. Cl. H04W 28/10 (2009.01); H04L 47/765 (2022.01)
CPC H04W 28/10 (2013.01) [H04L 47/767 (2013.01)] 20 Claims
OG exemplary drawing
 
1. In a fifth generation (5G) cellular network including a core network, a radio access network (RAN), and an edge portion having a near edge and a far edge, a method for improving efficiency of routing edge compute traffic from a user equipment (UE) to an edge compute server at the far edge, the method comprising:
provisioning a near edge control unit (CU) and a near edge user plane function (UPF) at the near edge of the edge portion of the 5G cellular network in response to detecting under-utilized computing resources at the far edge, components of the near edge being implemented on a first set of servers being located closer to the radio units of the RAN than a datacenter that hosts one or more components of the core network;
provisioning a far edge CU, a far edge UPF, and an edge compute workload at the far edge of the edge portion of the 5G cellular network, components of the far edge being implemented on a second set of servers being located closer to radio units of the RAN than the first set of servers of the near edge;
receiving UE traffic at one or more distributed units on the second set of servers of the far edge, the UE traffic comprising edge compute traffic and non-edge compute traffic;
identifying the edge compute traffic among the UE traffic;
routing the edge compute traffic of the UE traffic to the edge compute workload on the second set of servers at the far edge;
routing the non-edge compute traffic of the UE traffic to the near edge UPF on the first set of servers at the near edge;
consuming telemetry of a current load at the far edge based on physical layer processing operations performed by one or more distributed units on the far edge;
detecting, based on the telemetry, that the current load at the second set of servers at the far edge exceeds a first threshold level; and
in response to determining that the current load at the second set of servers at the far edge exceeds the first threshold level:
causing the edge compute workload at the second set of servers at the far edge to be migrated to the first set of servers at the near edge; and
deprovisioning the far edge CU, the far edge UPF, and the edge compute workload at the far edge.