Scheduled and random access uplink resource unit allocation
Techniques for resource unit (RU) allocation for wireless communication are disclosed. These techniques include identifying a first allocation of scheduled access (SA) and random access (RA) RUs for uplink from one or more wireless stations (STAs) to a wireless access point (AP) in a wireless local area network (WLAN). The techniques further include measuring one or more performance characteristics for the one or more STAs, determining a second allocation of uplink SA and RA RUs in the WLAN based on the measured one or more performance characteristics, and advertising the second allocation of uplink SA and RARUs to the one or more STAs.
1 . A method, comprising:
identifying a first allocation of scheduled access (SA) and random access (RA) resource units (RUs) for uplink from one or more wireless stations (STAs) to a wireless access point (AP) in a wireless local area network (WLAN);
measuring a total aggregate throughput of the one or more STAs while using the first allocation of SA and RA RUs;
determining, using a machine learning (ML) model, a second allocation of uplink SA and RA RUs in the WLAN based on the total aggregate throughput; and
advertising the second allocation of uplink SA and RA RUs to the one or more STAs.
2 . The method of claim 1 , wherein the second allocation of uplink SA and RA Rus is based on the first allocation.
3 . The method of claim 1 , wherein the ML model is a reinforcement learning ML model.
4 . The method of claim 3 , further comprising:
modifying the ML model by providing at least one of a penalty or a reward to the ML model based on the first allocation and the total aggregate throughput.
5 . The method of claim 1 , wherein the one or more STAs use the second allocation to select between SA and RA for uplink transmission to the AP.
6 . The method of claim 1 , further comprising measuring per-traffic identifier (TID) wait times for the one or more STAs.
7 . The method of claim 6 , wherein determining the second allocation of uplink SA and RA RUs in the WLAN comprises:
determining to increase an allocation of RA RUs and decrease an allocation of SA RUs based on the per-TID wait times for the one or more STAs.
8 . A system, comprising:
a processor; and
a memory having instructions stored thereon which, when executed on the processor, performs operations comprising:
identifying a first allocation of scheduled access (SA) and random access (RA) resource units (RUs) for uplink from one or more wireless stations (STAs) to a wireless access point (AP) in a wireless local area network (WLAN);
measuring a total aggregate throughput of the one or more STAs while using the first allocation of SA and RA RUs;
determining, using a machine learning (ML) model, a second allocation of uplink SA and RA RUs in the WLAN based on the total aggregate throughput; and
advertising the second allocation of uplink SA and RA RUs to the one or more STAs.
9 . The system of claim 8 , wherein the second allocation of uplink SA and RA Rus is based on the first allocation.
10 . The system of claim 8 , wherein the ML model is a reinforcement learning ML model, the operations further comprising:
modifying the ML model by providing at least one of a penalty or a reward to the ML model based on the first allocation and the total aggregate throughput.
11 . The system of claim 8 , wherein the one or more STAs use the second allocation to select between SA and RA for uplink transmission to the AP.
12 . The system of claim 8 , further comprising measuring per-traffic identifier (TID) wait times for the one or more STAs.
13 . A non-transitory computer-readable medium having instructions stored thereon which, when executed by a processor, performs operations comprising:
identifying a first allocation of scheduled access (SA) and random access (RA) resource units (RUs) for uplink from one or more wireless stations (STAs) to a wireless access point (AP) in a wireless local area network (WLAN);
measuring a total aggregate throughput of the one or more STAs while using the first allocation of SA and RA RUs;
determining, using a machine learning (ML) model, a second allocation of uplink SA and RA RUs in the WLAN based on the total aggregate throughput; and
advertising the second allocation of uplink SA and RA RUs to the one or more STAs.
14 . The non-transitory computer-readable medium of claim 13 , wherein the second allocation of uplink SA and RA Rus is based on the first allocation.
15 . The non-transitory computer-readable medium of claim 13 , wherein the ML model is a reinforcement learning ML model, the operations further comprising:
modifying the ML model by providing at least one of a penalty or a reward to the ML model based on the first allocation and the total aggregate throughput.
16 . The non-transitory computer-readable medium of claim 13 , wherein the one or more STAs use the second allocation to select between SA and RA for uplink transmission to the AP.
17 . The non-transitory computer-readable medium of claim 13 , further comprising measuring per-traffic identifier (TID) wait times for the one or more STAs.