US 12,335,786 B2
Techniques for adaptively allocating resources in a cloud-computing environment
Bozidar Radunovic, Cambridge (GB); Sanjeev Mehrotra, Kirkland, WA (US); Yongguang Zhang, Redmond, WA (US); Paramvir Bahl, Bellevue, WA (US); and Xenofon Foukas, Cambridge (GB)
Assigned to MICROSOFT TECHNOLOGY LICENSING, LLC, Redmond, WA (US)
Filed by Microsoft Technology Licensing, LLC, Redmond, WA (US)
Filed on Mar. 5, 2024, as Appl. No. 18/596,028.
Application 18/596,028 is a division of application No. 17/323,680, filed on May 18, 2021, granted, now 12,004,013.
Prior Publication US 2024/0214871 A1, Jun. 27, 2024
Int. Cl. H04W 28/16 (2009.01); G06F 11/32 (2006.01); G06F 11/34 (2006.01); G06N 20/00 (2019.01); H04W 24/10 (2009.01); H04W 28/02 (2009.01); H04W 72/21 (2023.01); H04W 72/23 (2023.01)
CPC H04W 28/16 (2013.01) [G06F 11/324 (2013.01); G06F 11/3428 (2013.01); G06N 20/00 (2019.01); H04W 24/10 (2013.01); H04W 28/0268 (2013.01); H04W 72/21 (2023.01); H04W 72/23 (2023.01)] 20 Claims
OG exemplary drawing
 
1. A computer-implemented method for reallocating resources among nodes in a cloud-computing environment, comprising:
monitoring a traffic load of a virtual radio access network (vRAN) workload, executing on a node of the cloud-computing environment, that provides wireless network access to one or more user equipment (UEs); and
based on comparing the traffic load to a threshold, reallocating resources of a hardware accelerator, including multiple encoding queues for different quality-of-service levels of vRAN traffic, to the vRAN workload or one or more other vRAN workloads, wherein the hardware accelerator is operable to perform signal processing for vRAN workloads, wherein the multiple encoding queues are distributed across multiple virtual functions, of a physical function that performs the signal processing, and wherein reallocating resources of the hardware accelerator includes adjusting a number of queues used for the vRAN workload and the one or more other vRAN workloads.