# US 2026/0292595 A1 · Scaling network capability using baseband unit pooling in fifth generation networks and beyond

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2026/0292595 A1
- Document ID: US20260292595A1
- Status: Published application
- Kind: A1
- Application number: 19/693,698
- Filed: 2026-06-01
- Published: 2026-09-24
- Earliest claimed: 2020-11-16
- Assignees: AT&T Intellectual Property I, L.P.
- Applicants: AT&T Intellectual Property I, L.P.
- Inventors: Minsung Jang; Aleksandr Zelezniak; Kaustubh Joshi
- Web page: https://patentgrep.com/patent/US20260292595A1

## CPC

- H04W 28/0861: {between base stations}
- H04W 28/0263: {involving mapping traffic to individual bearers or channels, e.g. traffic flow template [TFT]}
- H04W 28/12: using signalling between network elements
- H04W 80/02: Data link layer protocols
- H04W: WIRELESS COMMUNICATION NETWORKS (broadcast communication H04H; communication systems using wireless links for non-selective communication, e.g. wireless extensions H04M1/72)

## Abstract

The disclosed technology is directed towards load balancing baseband units in a communications network. A baseband physical layer 1 unit’s functions are disaggregated into Layer 1 (L1) distributed units and radio units, instead of deploying full-fledged baseband units at a service’ provider’s service areas (cells). A load balancer scales up the number of active distributed units based on increased actual demand, and scales down the active distributed units based on decreased demand. The L1 distributed units and radio units can be software-defined network functions, and need not be collocated, whereby the distributed units can be in the cloud or hub remotely located relative to the radio units deployed at the service areas. Examples of load balancing can be load balancing of transmitted data per carrier, per subcarrier, per user equipment, per transmission time interval (TTI / slot), per bearer, or per channel.

## Claims

### Claim 1

A system, comprising:
a processing system including a processor; and
a memory that stores executable instructions that, when executed by the processing system, facilitate performance of operations, the operations comprising:
determining a utilization condition of one or more active distributed units of a pool of distributed units in a mobility network based on one or more load measurements;
in response to determining that the utilization condition satisfies a scale-up criterion, activating an additional distributed unit from the pool; and
in response to determining that the utilization condition satisfies a scale-down criterion, deactivating a distributed unit of the one or more active distributed units.

### Claim 2 (depends on 1)

The system of claim 1, wherein the scale-up criterion comprises the one or more load measurements exceeding a first predetermined threshold.

### Claim 3 (depends on 2)

The system of claim 2, wherein the first predetermined threshold comprises an average load measurement across the one or more active distributed units exceeding a first threshold value.

### Claim 4 (depends on 1)

The system of claim 1, wherein the scale-down criterion comprises the one or more load measurements falling below a second predetermined threshold.

### Claim 5 (depends on 4)

The system of claim 4, wherein the second predetermined threshold comprises an average load measurement across the one or more active distributed units falling below a second threshold value.

### Claim 6 (depends on 1)

The system of claim 1, wherein the one or more load measurements comprise at least one of: processor utilization, memory utilization, input/output utilization, network interface utilization, queue depth, buffer occupancy, packet processing rate, or physical layer processing latency.

### Claim 7 (depends on 1)

The system of claim 1, wherein the determining the utilization condition comprises:
determining the utilization condition based on a weighted combination of the one or more load measurements.

### Claim 8 (depends on 1)

The system of claim 1, wherein the activating the additional distributed unit comprises:
instantiating a virtualized network function corresponding to the additional distributed unit in a cloud computing environment.

### Claim 9 (depends on 8)

The system of claim 8, wherein the activating the additional distributed unit comprises:
allocating compute resources for the additional distributed unit based on at least one of a requested number of CPU cores, a requested memory size, or a requested acceleration resource.

### Claim 10 (depends on 1)

The system of claim 1, wherein the operations further comprise:
load balancing downlink traffic among the one or more active distributed units by modifying at least one of a destination media access control address or a destination internet protocol address associated with control messages or corresponding user data.

### Claim 11 (depends on 10)

The system of claim 10, wherein the load balancing the downlink traffic comprises:
load balancing the downlink traffic at a selected granularity comprising at least one of per carrier, per subcarrier, per user equipment, per transmission time interval, per bearer, per channel, per frame, per subframe, or per slot.

### Claim 12

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:
determining a downlink data load associated with a radio unit deployed at a service area, wherein the radio unit is remotely located relative to a pool of distributed units comprising physical layer distributed units;
selecting, based on a load balancing criterion, a first distributed unit of the pool for routing a downlink control message;
routing the downlink control message via the first distributed unit to the radio unit;
routing downlink user data corresponding to the downlink control message via the first distributed unit to the radio unit;
determining that the downlink data load satisfies a load overutilization criterion; and
in response to determining that the downlink data load satisfies the load overutilization criterion, activating a second distributed unit from the pool and routing at least a portion of subsequent downlink control messages and corresponding downlink user data via the second distributed unit to load balance downlink traffic between the first distributed unit and the second distributed unit.

### Claim 13 (depends on 12)

The non-transitory machine-readable medium of claim 12, wherein the routing the downlink control message and the routing the downlink user data comprise:
inserting, by the first distributed unit, a distributed unit identifier into an enhanced common public radio interface (eCPRI) header field.

### Claim 14 (depends on 13)

The non-transitory machine-readable medium of claim 13, wherein the operations further comprise:
performing, at a cloud switch, address translation based on the distributed unit identifier to forward uplink data traffic to the first distributed unit that transmitted an uplink control message.

### Claim 15 (depends on 12)

The non-transitory machine-readable medium of claim 12, wherein the operations further comprise:
maintaining, by a load balancer, a mapping between distributed unit identifiers and at least one of media access control addresses or internet protocol addresses of the distributed units.

### Claim 16

A method, comprising:
receiving, by a processing system including a processor, downlink scheduling control information for a radio unit deployed at a service area, wherein the radio unit is remotely located relative to a pool of physical layer distributed units executing at one or more compute nodes;
selecting, by the processing system, a first physical layer distributed unit from the pool based on a load balancing policy, and modifying at least one of a destination media access control address, a destination internet protocol address, or a tunnel endpoint identifier to steer (i) the downlink scheduling control information and (ii) corresponding downlink user data through the first physical layer distributed unit;
inserting, by the processing system, a distributed unit identifier into an enhanced Common Public Radio Interface (eCPRI) header of a downlink transmission toward the radio unit;
after the selecting, determining, by the processing system, that a utilization metric of the first physical layer distributed unit satisfies an overutilization criterion;
responsive to the determining, instantiating, by the processing system, a second physical layer distributed unit from the pool at a different compute node, and configuring the second physical layer distributed unit with static radio parameters for the service area; and
routing, by the processing system, at least a portion of subsequent downlink scheduling control information and corresponding downlink user data through the second physical layer distributed unit to load balance downlink processing for the service area, while maintaining, for each downlink transmission time interval, consistency that the downlink scheduling control information and the corresponding downlink user data are processed by a same selected physical layer distributed unit.

### Claim 17 (depends on 16)

The method of claim 16, wherein the selecting the first physical layer distributed unit based on a load balancing policy comprises:
selecting, by the processing system, the first physical layer distributed unit based on at least one of round robin selection, least-utilized selection according to the utilization metric, or weighted selection among active physical layer distributed units.

### Claim 18 (depends on 16)

The method of claim 16, comprising:
configuring, by the processing system, prior to the receiving the downlink scheduling control information, each physical layer distributed unit of the pool with the static radio parameters via a functional application platform interface (FAPI) or network functional application platform interface (nFAPI) message exchange.

### Claim 19 (depends on 16)

The method of claim 16, wherein the inserting the distributed unit identifier into the eCPRI header comprises:
inserting, by the processing system, into the eCPRI header a value usable by a cloud switch to map uplink traffic to a physical layer distributed unit that transmitted a corresponding uplink control message.

### Claim 20 (depends on 16)

The method of claim 16, wherein the maintaining the consistency for each downlink transmission time interval comprises:
ensuring, by the processing system, that a downlink grant and downlink data corresponding to the downlink grant are both routed through a same selected physical layer distributed unit for that downlink transmission time interval.
