Methods and apparatuses for radio unit sharing in wireless communications networks

Abstract

A Radio Unit (RU) (28) concurrently transmits two or more radio carriers (54), each radio carrier (54) generated from a respective traffic stream (52) incoming to the RU (28) from a respective scheduler (50). The schedulers (50) reside in one or more remote Digital Units (DUs) (26), and the traffic streams (52) share one or more operational resources (140, 142, 144) at the RU (28). The RU (28) determines future times at which over-utilization of the shared operational resource(s) (140, 142, 144) is expected to occur and reduces resource allocation(s) for one or more selected ones of the traffic streams (52) at such future times, to reduce or eliminate the over-utilization. In at least one embodiment, the RU (28) sends feedback (100) to the scheduler(s) (50) associated with the selected traffic stream(s) (52), and the scheduler(s) (50) adapt scheduling with respect to the future time(s), to avoid or lessen the effects of the reduced allocation(s).

Description

[0001] Attorney Docket No. 1009-6680 / P108970W001

[0002] METHODS AND APPARATUSES FOR RADIO UNIT SHARING IN WIRELESS COMMUNICATIONS NETWORKS

[0003] TECHNICAL FIELD

[0004] Disclosed methods and apparatuses relate to wireless communications and, particularly, relate to the sharing of a Radio Unit for concurrent transmissions.

[0005] BACKGROUND

[0006] Wireless communications equipment typically supports multiple radio frequency (RF) bands corresponding to a multiplicity of radio access technologies (RATs). Such support requires digital front end (DFE) and analog front end (AFE) equipment to operate over multiple bands and multiple modes. An essential module for all such wireless equipment is the power amplifier (PA), and PAs stand as a potentially significant cost driver for communications equipment.

[0007] Generally, PA costs increase with the maximum power level supported. Thus, limiting the maximum PA gain is a straightforward means of controlling costs. But the issue is not so simple in certain scenarios. For example, consider a transmitter of a wireless communications network that transmits multiple carriers, where each carrier must have sufficient power spectral density at the borders corresponding to a desired cell coverage. Broadly, when transmitting multiple sector carriers for the same or different frequency bands and trying to achieve similar coverage for each carrier, using a shared PA for such transmissions introduces meaningful complexities.

[0008] For example, satisfying power spectral density (PSD) targets associated with desired cell coverage for all carriers transmitted via the PA may exceed the power limits of the PA, thus triggering scaling of the total power, for protection of the PA. But blind power scaling compromises the underlying radio transmissions, potentially degrading throughput.

[0009] SUMMARY

[0010] A Radio Unit (RU) concurrently transmits two or more radio carriers, each radio carrier generated from a respective traffic stream incoming to the RU from a respective scheduler. The schedulers reside in one or more remote Digital Units (DUs), and the traffic streams share one or more operational resources at the RU. The RU determines future times at which over-utilization of the shared operational resource(s) is expected to occur and reduces resource allocation(s) for one or more selected ones of the traffic streams at such future times, to reduce or eliminate the Attorney Docket No. 1009-6680 / P108970W001 over-utilization. In at least one embodiment, the RU sends feedback to the scheduler(s) associated with the selected traffic stream(s), and the scheduler(s) adapt scheduling with respect to the future time(s), to avoid or lessen the effects of the reduced allocation(s).

[0011] One embodiment comprises a method of operation by a RU, where the method includes the RU receiving two or more traffic streams for concurrent radio transmission. Each traffic stream is received from a respective scheduler and contributes to an aggregate utilization of a shared operational resource of the RU. The method further includes the RU forecasting the aggregate utilization on an ongoing basis with respect to future time intervals and determining that the forecasted aggregate utilization will exceed a limit for the shared operational resource at a respective one of the future time intervals. Still further, the method includes, in response to the determination, the RU performing a mitigation operation that includes: selecting one or more of the two or more traffic streams according to stream priorities; and, with respect to occurrence of the respective future time interval, reducing a per-stream utilization of the shared operational resource by each selected traffic stream.

[0012] A related embodiment comprises a RU that includes transmission circuitry, interface circuitry, and processing circuitry. The interface circuitry is configured to receive two or more traffic streams for concurrent radio transmission via the transmission circuitry. Each traffic stream is received from a respective scheduler, and each traffic stream individually contributes to an aggregate utilization of a shared operational resource of the RU. Correspondingly, the processing circuitry is configured to: forecast the aggregate utilization on an ongoing basis with respect to future time intervals and determine that the forecasted aggregate utilization will exceed a limit for the shared operational resource at a respective one of the future time intervals. In response to the determination, the processing circuitry performs a mitigation operation that includes: selecting one or more of the two or more traffic streams according to stream priorities; and with respect to occurrence of the respective future time interval, reducing a per-stream utilization of the shared operational resource by each selected traffic stream.

[0013] Another embodiment comprises a method of operation by a scheduler of a DU, which is remote from a RU that is shared by the scheduler with one or more other schedulers. The method includes the scheduler forming a traffic stream based on making ongoing scheduling decisions with respect to per-user traffic associated with a plurality of users, and outputting the traffic stream for radio transmission by the RU. Further, the method includes the scheduler receiving feedback generated by the RU, indicating that the RU will impose a reduction in utilization by the traffic stream of a shared operational resource at the RU for a future time interval, responsive to the RU determining that an aggregate utilization associated with sharing of the RU will exceed a limit for the shared operational resource. Correspondingly, the method further includes the Attorney Docket No. 1009-6680 / P108970W001 scheduler making the scheduling decisions with respect to the future time interval in dependence on the imposed reduction in utilization.

[0014] A related embodiment comprises a scheduler implemented in a DU. The scheduler comprises processing circuitry configured to form a traffic stream based on making ongoing scheduling decisions with respect to per-user traffic associated with a plurality of users, and to output the traffic stream via interface circuitry of the DU, for radio transmission by a remote RU that is shared with one or more other schedulers in the same or other DUs. The processing circuitry is further configured to receive, via the interface circuitry, feedback from the RU, indicating that the RU will impose a reduction in utilization by the traffic stream of a shared operational resource at the RU for a future time interval. The RU sends the feedback responsive to the RU determining that an aggregate utilization associated with sharing of the RU will exceed a limit for the shared operational resource. Correspondingly, the processing circuitry comprising the scheduler is configured to make the scheduling decisions with respect to the future time interval in dependence on the imposed reduction in utilization.

[0015] Of course, the present invention is not limited to the above features and advantages. Indeed, those skilled in the art will recognize additional features and advantages upon reading the following detailed description, and upon viewing the accompanying drawings.

[0016] BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 is a block diagram of a communications network, according to one embodiment.

[0018] Figure 2 is a block diagram of example details for a sharing controller operating as part of a scheduler in a Digital Unit (DU), and for a utilization controller operating in a Radio Unit (RU), according to one embodiment.

[0019] Figure 3 is a block diagram of further example details for two Digital Units (DUs) and a shared RU, according to one embodiment.

[0020] Figure 4 is a diagram of a technique for mitigating an overutilization of transmission power at a RU, according to one embodiment.

[0021] Figure 5 is a logic flow diagram of a method of operation at a RU, according to one embodiment.

[0022] Figure 6 is a block diagram of a RU, according to one embodiment.

[0023] Figure 7 is a logic flow diagram of a method of operation at a Digital Unit, according to one embodiment.

[0024] Figure 8 is a block diagram of a Digital Unit, according to one embodiment. Attorney Docket No. 1009-6680 / P108970W001 DETAILED DESCRIPTION

[0025] Figure 1 illustrates an example communications network 10 — network 10 — that is configured according to Fifth Generation (5G) specifications promulgated by the Third Generation Partnership Project (3GPP). The network 10 provides communication services to a potentially large number of user equipments (UEs) 12. Although the diagram suggests one set or cluster of UEs 12, it shall be understood that the network 10 may provide communications service coverage — network coverage — over one or more areas, which may be divided into network “cells,” with multiple UEs 12 operating in each cell. “Cell” in this context refers broadly to a particular coverage area and the corresponding allocation of specific communication frequencies and / or other communication resource for providing coverage in that area. Cells may be static or dynamic and may, for example, be defined according to transmit and / or receive beamforming.

[0026] The network 10 includes a core network (CN) 14 that interfaces to one or more external networks, such as the Internet, which is not shown explicitly in the diagram but is suggested by the depiction of user traffic incoming to CN 14, for transmission to targeted ones among the UEs 12 served by the network 10. The CN 14 comprises a number of computer servers or other nodes that are configured to provide various network services or functions. These functions include an Access and Mobility Management Function (AMF) 18, which is a Control Plane (CP) function that performs various operations in association with managing the access and mobility of UEs 12 in the network 10. The CN 14 also includes a User Plane Function (UPF) 20, which serves as an interconnect point between the mobile infrastructure and external NW(s) and performs routing and forwarding of packets carrying user traffic to and from respective UEs 12. Other functions included in the CN 14, such as Session Management Functions (SMFs), are not shown. Refer to 3GPP Technical Specification (TS) 38.300 V18.1.0 for example architecture and function details.

[0027] The network 10 further includes a Radio Access Network (RAN) 22 which provides the wireless connectivity for transmitting downlink user traffic to respective UEs 12 and receiving uplink user traffic from respective UEs 12, along with exchanging access and control signaling with the UEs 12. As illustrated, the RAN 22 adopts an “Open RAN” architecture in which various functions are split. In particular, the RAN 22 is split into one or more Central Units (CUs) 24, one or more Digital Units (DUs) 26, and one or more Radio Units (RUs) 28. While the diagram illustrates only one CU 24, one DU 26, and one RU 28, it shall be understood that the RAN 22 may include potentially large numbers of CUs 24, DUs 26, and RUs 26, for providing network coverage over one or more geographic regions. Further, it shall be understood that the one-to-one associations suggested in the diagram are not limiting. For example, one RU 28 may be shared by multiple — two or more DUs 26. Attorney Docket No. 1009-6680 / P108970W001

[0028] Among other things, the functional splitting between the DU and RU, such as defined by Open RAN, provides increased flexibility with respect to implementation, such as the use of cloud data centers for implementation of higher level RAN functions. Functional splitting also simplifies network scaling. In one or more embodiments, the DU / RU are based on a Lower Layer Split (LLS).

[0029] In at least one embodiment, CUs 24 provide non-real time, higher Layer 2 (L2) and Layer 3 (L3) processing. See, e.g., the Radio Resource Control (RRC) and Packet Data Convergence Protocol (PDCP) entities 30 and 32 of the CU 24 for supporting CP operations and see the RRC and PDCP entities 34 and 36 of the CU 24 for supporting UP operations. With respect to L2, L3, and so on, see 3GPP TS 23.501 vl 8.5.0 for example details, and note that these protocol processing entities comprise logical functions implemented via underlying computer processing circuitry, where the implementation may be direct or virtualized.

[0030] DUs 26 in one or more embodiments provide real time Layer 1 (LI) and Layer 2 (L2) scheduling functions. A DU 26 receives CP signaling from a supporting CU 24 via a Fl-C interface and receives user traffic from the supporting CU 24 via a Fl-U interface, and it includes multiple protocol processing layer entities, including a Radio Link Control (RLC) entity 40, a Medium Access Control (MAC) entity 42, and a Physical (PHY1) entity 44. “PHY1” denotes a split PHY layer, with certain PHY functions performed in the DU 26 and further PHY functions performed in the supporting RU 28 via the “PHY2” entity 60. Further elements of RU 28 include transmission circuitry comprised in a radio frequency (RF) transceiver circuitry that is configured for transmitting downlink radio signals to UEs 12 and receiving uplink radio signals from UEs 12.

[0031] Downlink operations are of particular interest here, with downlink user traffic for given UEs 12 supported by the DU 26 incoming from the CU 24 via the Fl-U interface. A scheduler 50 of the DU 26 schedules the transmission of the downlink user traffic according to a variety of factors, including the type of traffic and its associated Quality-of-Service (QoS) requirements, the radio conditions associated with the UEs 12 that are candidates for scheduling, etc. The PHY1 entity 44 of the DU 26 outputs one or more downlink traffic streams (DTSs) 52. Each downlink (DL) traffic stream 52 conveys scheduled user traffic as decided by the scheduler 50, for transmission by the RU 28 on a respective DL RF carrier 54. In operational scenarios of interest herein, a RU 28 receives two or more DL traffic streams 52, from one or more DUs 26, for transmission on respective DL RF carriers 54.

[0032] Correspondingly, the RU 28 can be understood as transmitting a DL radio signal 56 that is a composite or aggregation of two or more DL RF carriers 54 — i.e., the DL radio signal 56 is a multi-carrier radio signal. Such an aggregate radio signal may be referred to as a multi-carrier Attorney Docket No. 1009-6680 / P108970W001 radio signal, to reflect the fact that it comprises an aggregation of two or more DL RF carriers 54. In one or more embodiments, or under one or more operational circumstances, two or more of the DL RF carriers 54 are in different frequency bands, and in such instances, the DL radio signal 56 is a multi-band radio signal.

[0033] Each DL traffic stream 52 output by a DU 26 for transfer to a remote RU 28 comprises a digital baseband signal that may be encapsulated into packets transmitted from the DU 26 to the RU 28 over a “fronthaul” connection. The fronthaul connection follows Common Public Radio Interface (CPRI) or enhanced (eCPRI) protocols, for example, because of the efficiency of these protocols in handling packet-based data at high bandwidths. The underlying baseband signal comprises In-phase and Quadrature (IQ) data samples, representing the modulated data symbols that need to be transmitted over the air by the RU 28 on the corresponding DL RF carrier 54.

[0034] The RU 28 includes the additional or final PHY layer processing for these IQ data samples — see the PHY2 entity 60 — and includes RF transceiver circuitry 62 for transmission of the respective DL RF carriers 54 from the antenna system of the RU 28 — not shown in the diagram. In addition to the IQ data samples, the DU 26 sends synchronization and control information to the RU 28, thereby enabling the RU 28 to properly modulate and transmit the signals over the air.

[0035] The DU 26 and the RU 28 incorporate complementary features or behaviors that improve the sharing of operational resources at the RU 28 for concurrent reception of two or more DL traffic streams 52 and corresponding transmission of two or more DL RF carriers 54. Improvement with respect to overload events stands as a particular advantage. “Overload events,” also referred to as “overbooking events” or “overload conditions,” refers to instances of overbooking of one or more operational resources at the RU 28 with respect to concurrent transmission by the RU 28 of two or more DL RF carriers 54.

[0036] For each DL RF carrier 54 being transmitted, the RU 28 experiences one or more forms of loading. In one example, the transmission of each DL RF carrier 54 requires the allocation of a time-varying amount of transmit power from a transmit power budget defined by the power and / or temperature limits associated with the RU PA that is shared at the RU 28 for the transmission of multiple DL RF carriers 54. Other examples include “consumption” of the Digital Front End (DFE) resources of the RU 28 for each DL traffic stream 52 incoming to the RU 28 and / or corresponding “consumption” of the Analog Front End (AFE) resources available in the RU 28 for managing the conversion from the digital domain to the analog domain.

[0037] The DFE of the RU 28 handles digital signal processing tasks such as digital upconversion / downconversion, filtering, sample rate conversion, beamforming, digital predistortion, and FFT / IFFT operations. In complementary fashion, the AFE handles the Attorney Docket No. 1009-6680 / P108970W001 conversion between digital and analog signals, including DAC, ADC, mixing, amplification, filtering, and signal conditioning. Together, the DFE and AFE ensure that the DL traffic streams 52 incoming to the RU 28 are accurately and efficiently converted to corresponding DL RF carriers 54, for transmission over the air interface.

[0038] The advantageous RU sharing operations are provided via the cooperative operation of a sharing controller 70 implemented in the scheduler 50 of the DU 26 and a utilization controller 72 implemented in the RU 28. To the extent that two or more schedulers 50 share one RU 28, each scheduler 50 has a like-configured sharing controller 70, and the utilization controller 72 is configured to cooperate with any one or more such sharing controllers 70 for improved sharing of one or more operational resources of the RU 28. It shall be understood that the sharing controller 70 and the utilization controller 72 are respectively implemented as logical functions realized via fixed circuitry, programmatically configured circuitry, or a mix of fixed and programmatically configured circuitry. Further, it shall be understood that references to the RU 28 receiving a traffic stream from a scheduler 50 implies receipt of the traffic stream over a DU- RU interface. In at least one embodiment, each one of two or more schedulers 50 sharing the same RU 28 resides in a respective DU 26, such that each traffic stream received at the RU 28 comes from a respective DU 26.

[0039] In one or more embodiments, the sharing controller 70 in any given DU 26 comprises one or more processors configured to carry out the operations described herein, based on the execution of computer program instructions stored in a memory that is included in or accessible to the one or more processors. In one or more embodiments, the utilization controller 72 in any given RU 28 comprises one or more processors configured to carry out the operations described herein, based on the execution of computer program instructions stored in a memory that is included in or accessible to the one or more processors.

[0040] The utilization controller 72 of a RU 28 uses historical data and / or data provided by the associated DU(s) 26 to predict overbooking events and takes action to reduce a per-stream utilization of the affected operational resources at the RU 28 by one or more of the DL traffic streams 52 with respect to the predicted time(s). Saying that a DL traffic stream 52 utilizes operational resources at the RU 28 refers to the overall “signal path” in the RU 28, including any one or more of the reception processing, D / A conversion, RF signal generation, and transmission.

[0041] The involved resources comprise, for example, any one or more of signal buffer space, processing memory, compute cycles, operating temperature margin, and transmit power, and it shall be understood that each RF carrier 54 transmitted by the RU 28 involves the corresponding Attorney Docket No. 1009-6680 / P108970W001 reception, processing, conversion, amplification, etc., of a respective DL traffic stream 52, all at the expense of a certain portion or fraction of the operational resources available in the RU 28.

[0042] Figure 2 illustrates example details for a sharing controller 70 of a DU 26 and a corresponding utilization controller 72 of a RU 28 that is used for transmission of DL traffic streams 52-1 through 52-N, as output from the DU 26 to the RU 28 via a communication interface 80. Non-limiting example implementations of the communication interface 80 include CPRI or eCPRI.

[0043] In the forward or DL direction, the interface communication 80 carries the DL traffic streams 52-1 through 52-N from the DU 26 to the RU 28, where each such stream comprises packets of I / Q waveform values, and where N is an integer greater than or equal to two. The DU 26 also outputs mitigation control information (MCI) 84, for use by the RU 28 in mitigating overutilization of one or more operational resources at the RU 28 in support of concurrent generation and transmission of DL RF carriers 54-1 through 54-A corresponding to the incoming DL traffic streams 52-1 through 52-N.

[0044] In one or more embodiments, or under one or more operational scenarios, the MCI 84 comprises only stream priority information (SPI) 86, which comprises priority information for the DL traffic streams 52-1 through 52-N. A stream priority indication module 88 that is logically realized as part of the sharing controller 70 outputs the SPI 86. The SPI 86 comprises, for example, an indication of a prioritized one among the DL traffic streams 52-1 through 52-N, relative to priorities of all streams, or a priority of each stream relative to some predefined scheme, e.g., low, medium, and high. An over-utilization mitigator module 90 that is logically realized as part of the utilization controller 72 in the RU 28 uses the indicated stream priorities to decide which DL traffic stream(s) 52, from among the DL traffic streams 52-1 through 52-N, is / are selected for mitigation control with respect to the utilization controller 72 forecasting an overutilization event.

[0045] In one or more embodiments, or under one or more operational scenarios, the MCI 84 further includes load forecasting information (LFI) 92 for one or more of the DL traffic streams 52-1 through 52-N. For example, the sharing controller 70 receives PDCP data from the associated CU 24. The PDCP data comprises, for example, PDCP statistics such as any one or more of packet arrival times and sizes, scheduling request information, buffer size information, etc. In such embodiments, the sharing controller 70 of the DU 26 includes a utilization forecaster module 94 that is configured to forecast the loading associated with each individual DL traffic streams 52-1 through 52-N. Rather than expressing the forecasted loading directly in units corresponding to the operational resource(s) at the RU 28 that are consumed by the processing of the DL traffic streams 52 and the transmission of the DL RF carriers 54 derived therefrom, the Attorney Docket No. 1009-6680 / P108970W001 loading may be expressed as “demand” information that can be mapped to resource consumption.

[0046] For example, forecasted loading may be expressed in terms of Physical Resource Block (PRB) usage, at least in embodiments where the DL RF carriers 54 are Orthogonal Frequency Division Multiple Access (OFDMA) carriers comprised of time / frequency resources that are aggregated or grouped into PRBs for purposes of scheduling the transmission of DL traffic to given UEs 12. The greater the number of PRBs allocated for conveying user traffic, generally the higher the “loading” is at the RU 28, e.g., in terms of required transmit power.

[0047] Notably, in embodiments where the sharing controller 70 outputs LFI 92, the LFI 92 in at least one such embodiment is variable in dependence on the bandwidth available on the communication interface 80. Or, more particularly, the amount of LFI 92 may be varied as a function of bandwidth headroom available on the communication interface 80, after accounting for the bandwidth requirements associated with transporting the DL traffic streams 52 and related control / timing.

[0048] For example, the resolution or richness of the LFI 92 may be varied as a function of available reserve bandwidth on the communication interface 80. Resolution refers to granularity of indicated loading, for example, where loading could be quantized as low, medium, or high in limited reserve bandwidth scenarios, versus indicating predicted PRB requirements in scenarios where ample reserved bandwidth exists. However, in one or more embodiments, the utilization controller 72 at the RU 28 includes a utilization forecaster module 96, which lessens or obviates the need for load forecasting at the sharing controller 70 of the DU 26. In such embodiments, the utilization forecaster module 96 at the utilization controller 72 of the RU 28 tracks historical data for the incoming DL traffic streams 52- 1 through 52-N, such as PRB utilization, and forecasts loading at the RU 28.

[0049] This module and other “modules” described herein are realized in one or more embodiments via the execution of computer program instructions by corresponding processors. That is, a “module” can be understood as a logical function realized via programmatic configuration of processing circuitry.

[0050] Continuing with the example details of Figure 2, the utilization controller 72 provides mitigation control feedback (MCF) 100 to the DU 26. With respect to the utilization controller 72 predicting an over-utilization event at the RU 28, the MCF 100 at a minimum indicates the timing of the predicted over-utilization event and indicates the DL traffic stream(s) 52 that are selected by the over-utilization controller 72 for mitigation control. Such information is referred to as stream selection information (SSI) 102 and mitigation timing information (MTI) 104. Attorney Docket No. 1009-6680 / P108970W001

[0051] For example, the utilization controller 72 uses forecasted loading to determine that the transmit power requirements for the individual DL traffic streams 52 will exceed an aggregate transmit power limit of the RU 28 at some future time. The utilization controller 72 then uses the SPI 86 to identify one or more non-priority or lower priority ones among the DL traffic streams 52- 1 through 52-N, and it selects at least one such non-priority or low-priority DL traffic stream 52 for mitigation control. In this transmit-power context, mitigation control means, with respect to the predicted time of over-utilization, allocating less transmit power than is nominally required for transmission. The SSI 102 will indicate the affected DL traffic stream(s) 52, and the MTI 104 will indicate the future time(s) at which the mitigation(s) will be applied to the affected DL traffic stream(s) 52.

[0052] In one or more embodiments, the MCF 100 further includes, at least when sufficient reserve bandwidth is available on the communication interface 80, mitigation control details (MCD) 106 and / or scheduling adjustment recommendations (SAR) 108. For example, the MCD 106 indicates the nature or extent of mitigation control. For example, in a context where the mitigation is a reduction in allocated transmit power below the nominal transmit power required for satisfaction of QoS or other requirements, the MCD 106 may indicate the extent of underallocation. The SAR 108 may indicate one or more other DL traffic streams 52 that are candidates for scheduling adjustment, such as future time intervals where the operational resource(s) at issue are under-utilized at the RU 28. Such information allows, for example, a scheduling adaptor module 110 of the sharing controller 70 to shift traffic that was scheduled for transmission at the time(s) associated with the forecasted over-utilization to one or more subsequent times associated with under-utilization. These operations can be understood as an intelligent approach to spreading the load.

[0053] Figure 3 illustrates further details for the DU 26 and the RU 28 and illustrates another dimension to RU sharing. Rather than the RU 28 being shared with respect to two or more DL traffic streams 52 originating from the same DU 26, Figure 3 illustrates an example of two DUs 26-1 and 26-2 sharing a same RU 28. In one or more embodiments, or in one or more operational scenarios, the two DUs 26-1 and 26-2 belong to the same wireless communication networks 10. However, in at least one embodiment, or under one or more other operational scenarios, the DU 26-1 belongs to a first wireless communication network 10-1 and the DU 26-2 belongs to a second wireless communication network 10-2. In such embodiments or scenarios, mitigation control feedback provided by the RU 28 to the respective DUs 26 effects a kind of indirect resource utilization control between the two networks 10-1 and 10-2, without requiring explicit coordination signaling or interdependencies. Attorney Docket No. 1009-6680 / P108970W001

[0054] Each DU 26 outputs one or more DL traffic streams 52, meaning that the RU 28 is tasked with transmitting a DL radio signal 56 that contains at least two DL RF carriers 54. The illustration depicts the DU 26-1 outputting a DL traffic stream 52-1 and the DU 26-2 outputting a DL traffic stream 52-2, with the DL radio signal 56 correspondingly containing the DL RF carrier 54-1 corresponding to the DL traffic stream 52-1, and the DL RF carrier 54-2 corresponding to the DL traffic stream 52-2. The DL RF carrier 54-1 serves DL traffic to targeted UEs 12 in a first group 120 of UEs 12, and the DL RF carrier 54-2 serves DL traffic to targeted UEs 12 in a second group 122 of UEs 12. However, it shall be understood that the DL radio signal 56 will contain as many DL RF carriers 54 as the number of DL traffic streams 52 it receives in the sharing arrangement, and each DU 26 itself may output more than one DL traffic stream 52.

[0055] According to the illustrated details, the RU 28 includes interface circuitry 130 that is configured to support transmission and reception of signaling via the communication interface 80. The RU 28 further includes one or more transmit signal paths, also referred to as transmission circuitry, comprising one or more digital front ends (DFEs) 132, one or more analog front ends (AFEs) 134, one or more power amplifiers (PAs) 136, and one or more antennas 138, e.g., one or more phased array antennas used for transmission of the DL radio signal 56.

[0056] The DFE(s) 132 process the incoming DL traffic streams 52 and comprise corresponding operational resources 140, such as one or more digital signal processors (DSPs), working memory, and signal sample buffers. Each incoming DL traffic stream 52 “consumes” a portion of the finite operational resources 140 represented by the DFE(s) 132.

[0057] The AFE(s) 134 provide for conversion of the digital signals output from the DFE(s) 132 into the analog domain, along with corresponding modulation and up-conversion to form respective RF communications signals, which are amplified by the PA(s) 136 for transmission as the DL RF carriers 54. The AFE(s) 134 comprise corresponding operational resources 142, such as digital-to-analog converters (DACs), modulators, etc., and each incoming DL traffic stream 52 can be understood as consuming a portion of the operational resources 142. Likewise, each incoming DL traffic stream 52 can be understood as consuming a portion of the operational resources 144 represented by the PA(s) 136. The operational resources 144 comprise, for example, an overall available transmit power, with the transmission of each DL RF carrier 54 consuming a portion of the overall available transmit power.

[0058] Other example operational resources include operating temperature margin, which indicates the temperature headroom between a current operating temperature of one or more components or circuits of the RU 28 and a maximum operating temperature. In this regard, the Attorney Docket No. 1009-6680 / P108970W001 activity at the RU 28 with respect to each incoming DL traffic stream 52 may be understood as contributing to the dissipation of electrical power at the RU 28 and, hence, contributing to the temperature rise-over- ambient at the RU 28.

[0059] The RU 28 further includes processing circuitry 150 that is configured to control the DFE(s) 132, the AFE(s) 134, and the PA(s) 136 for processing of the DL traffic streams 52 and corresponding transmission of the DL RF carriers 54. Such circuitry 150 shall be understood as implementing the aforementioned utilization controller 72 and, hence, the processing circuitry 150 outputs mitigation control signaling 152 for imposition of its mitigation decisions, e.g., power reductions, etc.

[0060] The processing circuitry 150 comprises fixed circuitry or programmatically configured circuitry or a mix of both. In at least one embodiment the processing circuitry 150 comprises one or more microprocessors or other digital processors that is / are configured to operate as the described utilization controller 72 based on the execution of computer program instructions (CPI) 154 stored in associated storage 156. The storage 156 in one or more embodiments further stores data 158, which may be provisioned data or dynamically determined data. In at least one embodiment, the data 158 includes stream priority information, load forecasting information, etc.

[0061] As described above, the RU 28 outputs MCF 100. Also, as noted, the MCF 100 pertains to the DL traffic stream(s) 52 / DL radio carrier(s) 54 selected for mitigation. Thus, the diagram depicts MCF 100-1 fed back to the DU 26-1 and MCF 100-2 fed back to DU 26-2. More particularly, if the RU 28 predicts a given over-utilization event and selects for corresponding mitigation control one of the DL traffic streams 52 originating from the DU 26-1, it sends MCF 100-1 to DU 26-1. The DU 26-2 may receive no feedback in this example. Conversely, if the RU 28 selects for mitigation one of the DL traffic streams 52 originating from the DU 26-2, the RU 28 sends MCF 100-2 to the DU 26-2, and the DU 26-1 may receive no feedback.

[0062] The type and amount of feedback provided as MCF 100 may be adjusted or varied as a function of available bandwidth on the communication interface 80. In at least one embodiment, the utilization controller 72 of the RU 28 acts as a type of arbitrator for the DUs 26 that are sharing the RU 28. For example, the utilization controller 72 predicts the loading of one or more of its operational resources with respect to the two or more DL traffic streams 52 it is receiving from the DUs 26, or the utilization controller 72 receives load forecasting information from the DUs 26. Either way, the utilization controller 72 in one or more embodiments predicts upcoming times when it is underutilized or has more than a defined threshold of operating margin with respect to one or more operational resources, such as transmit power. In response to predicting underutilization, the utilization controller 72 sends MCF 100 to one or more of the DUs 26, Attorney Docket No. 1009-6680 / P108970W001 indicating that there is an upcoming opportunity to increase loading of the RU 28 and those one or more DUs 26 respond by adjusting traffic stream scheduling to account for such opportunity.

[0063] Consider a more detailed example that assumes each DL RF carrier 54 is an OFDMA carrier serving a corresponding cell of the network 10. Each OFDMA carrier comprises a plurality of narrowband subcarriers and the intersection of a subcarrier with an OFDM symbol time represents a resource element or “RE,” which is the smallest allocable resource on the OFDMA carrier. A PRB, in turn, is a group of REs and represents the smallest scheduled allocation of resources on the OFDMA carrier for conveying traffic to a UE 12. Thus, at any given OFDM symbol time, the overall amount of user traffic to be sent determines the number of allocated PRBs and, in general, more traffic per OFDM symbol time in a DL traffic stream 52 means more transmit power will be needed at the RU 28 for transmission of the corresponding DL RF carrier 24.

[0064] Thus, either the involved DU(s) 26 and / or the RU 28 in one or more embodiments are configured to monitor “utilization” in terms of PRB allocations and use historical PRB allocations for each of the DL RF carriers 54 being transmitted to predict upcoming transmit power overbooking at the RU 28. Such forecasting is performed either at the DU(s) 26 via DU- based utilization forecasting — see module 94 in Figure 2 — or is performed at the RU 28 via RU based utilization forecasting — see module 96 in Figure 2. Also, as noted, forecasting may be extended to include forecasting of underutilization. For example, it may be forecasted that a particular one of the DL RF carriers 54 will require relatively low transmit power at a future time, and the utilization controller 72 of the RU 28 may indicate this to the DU(s) 26 that are responsible for one or more of the other DL traffic streams 52. Such indication serves as a trigger to one or more such DUs 26 to increasing loading at the RU 28 with respect to the future time, e.g., by adapting user scheduling such that more traffic is sent at the future time, thereby making better use of the available transmit power of the RU 28.

[0065] Each DU 26, e.g., DUs 26-1 and 26-2, include interface circuitry 170 supporting the communication interface 80 with the RU 28. Each DU 26 further includes processing circuitry 172 and storage 174. In at least one embodiment, the processing circuitry 172 comprises one or more processors that are programmatically configured to carry out the DU operations described herein, based at least in part on executing computer program instructions (CPI) 176 stored in the storage 174. As a particular example, the processing circuitry 172 in one or more embodiments is configured to implement a scheduler 50 as described herein — see schedulers 50-1 and 50-2. The storage 174 in one or more embodiments further stores one or more types of data 178, e.g., live data associated with scheduling traffic streams 52 and responding to mitigation feedback from the RU 28. See the traffic streams 52-1 and 52-2. Attorney Docket No. 1009-6680 / P108970W001

[0066] Turning to the operations suggested in Figure 4, it is assumed that a DU 26 outputs two DL traffic streams 52 to a RU 28, for concurrent transmission as respective DL RF carriers 54. Particularly, the DU 26 outputs DL traffic stream 52-1 to the RU 28 for transmission as DL RF carrier 54-1, and outputs DL traffic stream 52-2 to the RU 28 for transmission as DL RF carrier 54-2. The DU 26 also outputs SPI 86 to the RU 28, indicating stream priority.

[0067] Each column represents a symbol time, and at each symbol time, transmission of the DL RF carrier 54-1 requires an amount of transmit power that depends on multiple factors, including the amount of traffic and the radio condition(s) of the UE(s) 12 to be served at that symbol time. The required — nominal — transmit power required at each symbol time for the DL RF carrier 54- 1 is shown as a white (unfilled) bar, and the transmit power required at each symbol time for the DL RF carrier 54-2 is shown as a hatched (filled) bar. The bars are shown in a stacked arrangement because the transmit power allocations come from the same power budget, e.g., the two DL RF carriers 54-1 and 54-2 are being transmitted via the same PA.

[0068] One sees that for certain symbol times — also referred to as time slots — the combined required or nominal power transmit for the two DL RF carriers 54-1 and 54-2 exceeds an aggregate power limit. Such times represent instances of over-utilization of the available transmit power at the RU 28 and the utilization controller 72 performs or initiates the performance of power scaling at such time instances, to reduce the transmit power allocation for the lower priority one of DL RF carriers 54-1 and 54-2. Of course, transmitting a RF DL carrier with less power than is nominally required may cause problems, e.g., signal-to-noise ratios (SNRs) not being met for the traffic affected by the power scaling. Therefore, the utilization controller 72 returns MCF 100 to the sharing controller 70 in advance of the time(s) at which power scaling is applied, giving the sharing controller 70 time to make or adjust its scheduling decisions with respect to the upcoming time(s) during which the RU 28 applies the power scaling.

[0069] One adaptation is choosing to reschedule some of the traffic that would have been transmitted at the involved time(s), for transmission in other symbol times / slots that are not predicted as being overbooked. As shown in Figure 4, such spreading of traffic is a form of timedomain “waterfilling,” where some of the traffic originally scheduled, or to be scheduled, for transmission at a given future time that is predicted as being overbooked is moved to one or more other times. Additionally, or alternatively, the sharing controller 70 performs any one or more of the following adaptations: modulation and coding scheme (MCS) adaptation, power spectral density (PSD) adaptation, PRB adaptation, and priority adaptation. Adaptations by the sharing controller 70 — e.g., by the scheduling adaptor module 110 shown in Figure 2 — reduce or eliminate the deleterious effects of the corresponding reduction(s) in resource allocations performed by the utilization controller 72 to mitigate overloading at the RU 28. Attorney Docket No. 1009-6680 / P108970W001

[0070] Thus, one technique disclosed herein is that a RU 28 that estimates future transmit power requirements associated with its concurrent transmission of two or more DL RF carriers 54, determines when a future transmit power requirement exceeds a power limit, and takes action at the future time(s) to mitigate the power overload, such as by reducing the power allocation for one or more of the streams below the allocation that would have been given absent the overload condition. Further, in advance of taking such action, the RU 28 indicates the upcoming allocation reduction to the affected DU(s) 26, so that the affected DU(s) 26 can update scheduling with respect to the future time(s), to reduce the impact of the allocation reduction.

[0071] Such action may be shifting some of the traffic scheduled for those future time(s) to one or more other times — traffic spreading. Of course, in addition to performing waterfilling to eliminate or reduce overutilization of resources, the utilization controller 72 of the RU 28 may indicate to one or more DU(s) 26 upcoming underutilizations of transmit power and / or one or more other operational resources at the RU 28, such that the sharing controller(s) 70 of those DU(s) 26 adjust scheduling to spread traffic to the one or more upcoming times of underutilization.

[0072] Consider a further example where the utilization controller 72 of a RU 28 determines that the aggregate transmit power needed at a future time for its concurrent transmission of two or more DL RF carriers 54 will exceed a safe operating limit for the involved PA at the RU 28 — i.e., the utilization controller 72 identifies an upcoming power overbooking event. For example, the utilization controller 72 receives load forecasting information from the DU(s) 26 associated with the two or more DL traffic streams 52 corresponding to the two or more DL RF carriers 54 and determines from the load forecasting information that a power overbooking will occur at the future time. Alternatively, the utilization controller 72 maintains historic data regarding the transmit power requirements of each DL RF carrier 54 and uses this historic data to predict future instances of power overbooking.

[0073] Regardless, in response to detecting the upcoming power overbooking event, the utilization controller 72 sends MCF 100 to one or more DUs 26, indicating that the RU 28 will scale the power allocation for the transmission of one or more DL RF carriers 54, where “scaling” the power allocation means reducing the allocated transmit power below the nominal power allocation that otherwise would have been used absent the power overbooking. In response, to receiving such indication, the DU(s) 26 adjust scheduling with respect to the future time indicated in the MCF 100, such as by using a lower-order modulation or a higher coding rate, to make the transmission more robust in view of the lower power.

[0074] As a further example, consider a case where multiple DUs 26 share one RU 28, with each DU 26 outputting at least one DL traffic stream 52 to the RU 28, for transmission as a Attorney Docket No. 1009-6680 / P108970W001 corresponding DL RF carrier 54. Further assume that at least one of the DUs 26 outputs two or more such DL traffic streams 52. In such cases, the DUs 26 operate independently from one another, with each DU 26 determining the priority / priorities of only the DL traffic streams 52 it originates. As such, the utilization controller 72 in the RU 28 selects which DL traffic stream(s) 52 are “under-allocated” with respect to predicted over-utilization events at the RU 28, with the RU 28 sending corresponding signaling to the DU(s) 26 associated with the selected DL traffic stream(s) 52, such that they can adjust scheduling with respect to the predicted over-utilization event(s) and thereby avoid or ameliorate the negative effects that might otherwise arise from the under-allocations. Again, if the operational resource at issue is transmit power, under-allocating transmit power may be ameliorated by reducing the required transmit power by shifting traffic, changing one or more transmission parameters such as MCS, etc.

[0075] A key advantage in such embodiments is that there is at least a loosely closed control loop involving the RU 28 and its associated DU(s) 26. Namely, in such embodiments, upon predicting that an overloading event will occur at the RU 28 at a future time, the RU 28 selects one or more of the incoming DL traffic streams 52 for overloading mitigation, which means reducing one or more resource allocations at the RU 28 below that required for nominal processing and transmission of the selected DL traffic stream(s) 52 as corresponding DL RF carriers 54. By notifying the affected DU(s) 26 of the forthcoming mitigation operations at the RU 28, the affected DU(s) 26 adapt scheduling such that the selected DL traffic stream(s) 52 are adapted to account for the reduced resource allocation(s) at the RU 28. As noted, with sufficient bandwidth on the communication interface(s) 80 between the RU 28 and the DU(s) 26, the mitigation feedback going from the RU 28 to the affected DU(s) 26 may be “rich,” such as by indicating the extent of the under-allocation, thereby allowing the affected DU(s) 26 to optimize the scheduling adaptations. When the traffic streams originate from schedulers residing in DUs from different wireless communications networks, the feedback effects indirect coordination between the networks.

[0076] Consider the utilization forecaster 96 depicted in Figure 2 and assume that the shared operational resource at issue at the RU 28 is transmit power — that is, K DL RF carriers 54 transmitted by the RU 28 share the same transmit power budget or allowance. Here, “ ”’ is an integer greater than or equal to two.

[0077] The utilization forecaster 96 receives the DL traffic streams 52 corresponding to the two or more DL RF carriers 54, along with receiving corresponding SPI 86 and LFI 92 and uses that information to forecast transmission powers p i (0 through p^(t) for the K DL RF carriers 54. The corresponding forecasted aggregate transmission power is therefore determined using the Attorney Docket No. 1009-6680 / P108970W001 forecasted per-carrier transmission powers and power overbooking (POB) events are identified for mitigation.

[0078] Thus, utilization forecaster 96 receives load and priority information on a per stream basis and combines that information with the actual per carrier physical resource block (PRB) utilization or, more particularly, power measurements per DL RF carrier 54, for prediction of the total power utilization in the transmission upcoming slots. In Long Term Evolution (LTE), a PRB comprises 12 consecutive OFDM subcarriers for a duration of one slot (0.5 ms). For 5G New Radio (NR), a PRB comprises 12 subcarriers, with the spacing between the subcarriers based on the numerology used.

[0079] Depending on the implementation of utilization forecaster 96 in terms of computing complexity, various multi-variate regression or smoothing methods as well as recurrent neural network based learning methods (such as Long Short Term Memory (LSTM)) can be employed for the forecasting function. The output of the module can be reconfigured based on the estimated load fluctuations per DL RF carrier 54. For example, one may specify inter arrival times of overbooking events and duration of bursts defined as the time between first overbooking event and the instance after a prespecified number of slots, denoted by T, following the last overbooking event such that in between the two times, inter arrival time never exceeds T. When such events occur, the utilization forecaster 96 predicts the duration of such bursts in the coming prediction window. The length of prediction window can be limited by the granularity and prediction depth associated with generation of the LFI 92 at the DU(s) 26.

[0080] Further, the outputs of the utilization forecaster 96 can be varied depending on the prediction window and the bandwidth and latency of the communication interface(s) 80. For example, the output from the RU 28 back to the DU(s) 26 may contain the predicted traffic load (low, medium, heavy) per DL RF carrier 54 and per slot within a prespecified time window in case of a low-latency RU-DU reporting channel. For a high-latency feedback, a summary of average traffic load over the upcoming time-window can be generated.

[0081] In one or more embodiments, the MCF 100 sent from the utilization controller 72 of a RU 28 to the sharing controller 70 of an associated DU 26 comprises a power overbooking (POB) status report. An example POB status report is a multi-functional report. For example, the POB status report may indicate a recommended action for the DU 26 to take with respect to a DL traffic stream 52 selected by the RU 28 for an upcoming mitigation. Alternatively, the POB status report may indicate a power load summary according to the available granularity from the utilization forecaster 96 of the utilization controller 72 at the RU 28. In this latter case, the sharing controller 70, not the RU 28, decides what action(s) to take with respect to the power scaling that will be applied by the RU 28 to mitigate the forecasted POB event. Attorney Docket No. 1009-6680 / P108970W001

[0082] POB status reports may be sent on a recurring basis, and each one may contain as much compressed information as possible, such as a single bit per DL RF carrier 54 to indicate “no overbooking event” or “heavy overbooking event” to trigger / de-trigger POB adaptor actions at the respective DUs 26. Another alternative is that a report targeting any given DL RF carrier 54 may contain a total power load from all other DL RF carriers 54, instead of individual power load information. Here, the targeted DL RF carrier 54 has been selected for an upcoming mitigation control at the RU 28 and the report received at the involved DU 26 summarizes the potential load situation from all other DL RF carriers 54 that are sharing the same power budget at the RU 28.

[0083] Depending on the priority of the DL traffic stream 52 corresponding to the targeted DL RF carrier 54 and further depending upon associated traffic buffer status, the sharing controller 70 of the responsible DU 26 may perform a defensive action, such as by reducing PSD or reducing the number of PRBs used, with respect to the future time at which mitigation at the RU 28 will be applied. Other example adaptation actions that may be applied by the sharing controller 70 is causing the involved scheduler 50 to schedule UEs 12 that are tolerant to power scaling and the resulting increases in error vector magnitude (EVM). This approach allows the DU 26 to maintain the planned PSD and / or bandwidth of the targeted DL RF carrier 54 during the time at which the RU 28 performs the overbooking mitigation(s). Additionally, or alternatively, the POB status report may contain power allowances per DL RF carrier 54 on a per slot or per sliding window basis.

[0084] The content of the POB status report can be varied based on the RU-to-DU signal path latency and bandwidth. If resources for RU-to-DU communications are scarce, a short summary of overbooking events is provided. If there is a fast report status channel, the RU 28 can provide more detailed info such as per slot power allowance forecast.

[0085] One point to note is that scheduling updates may be needed only when the likelihood of a resource overutilization event at the RU 28 is high. For example, when the predicted total power at the RU 28 is already below a permissible threshold, no action is needed. But if the POB status report indicates a potential overbooking event or bursts of overbooking events within a prespecified window length (the window length can be indicated by the POB status report or configured by a network operator or system administrator), the sharing controller 70 at a DU 26 receiving such reports may trigger adaptions to the ongoing scheduling behavior of the associated scheduler 50. The adaptations may span multiple slots depending on the interarrival times of power overshooting events. Without violating the quality-of-service (QoS) requirements, such as packet delivery latency, target UE data rate, the data rates of the UE traffic can be adapted to the upcoming predicted overbooking conditions. Attorney Docket No. 1009-6680 / P108970W001

[0086] If POB status report includes per-slot information, the scheduler 50 has more degrees of freedom to disperse user traffic for different UEs 12 being serviced along the time windows. The POB status report may contain priority info for the DL RF carriers 54 in the form of maximum back-off thresholds and / or tokens for power allowances per slot. With such priority info, multi- step power allocation can be performed independently by the scheduler 50 at each DU 26, with respect to the DL traffic streams 52 originating from the scheduler 50.

[0087] For example, assume that a RU 28 transmits K DL RF carriers 54 using a shared transmit power budget, and assume that the utilization controller 72 of the RU 28 has predicted an upcoming power overbooking event and has selected the k-th one of the K DL RF carriers 54 for application of mitigation control — i.e., the k-th DL RF carrier 54 is selected for power scaling according to which the nominal power allocation for the k-th DL RF carrier 54 at the future time will be reduced via power scaling.

[0088] The k-th DL RF carrier 54 is generated at the RU 28 from the k-th DL traffic stream 52 and the DU 26 originating that k-th DL traffic stream 52 includes a scheduler 50 responsible for determining traffic scheduling in the k-th DL traffic stream 52. The responsible scheduler 50 includes a sharing controller 70, which is responsible for adapting scheduling by the responsible scheduler 50, to account for the mitigation to be applied at the RU 28. In one approach, the sharing controller 70 causes the scheduler 50 to redistribute power allocation over multiple transmission slots, to remove power over shooting events (e.g., power pooling) or reduce overshooting levels to tolerable levels (e.g., power-back-off per sector carrier / user within a threshold for each slot).

[0089] If the POB status report received at the DU 26 provides a single power load summary for some specified window length, then action by the responsible sharing controller 70 may be set as a conservative behavior by limiting the total power (or bandwidth) scheduled over a prespecified window length indicated by the POB status report based on the interarrival times / burstiness of power overshooting events as forecasted at the RU 28. Notably, the proposed POB status report mechanism can be utilized for managing priority between competing schedulers 50 across multiple DUs 26 sharing a RU 28.

[0090] Figure 5 illustrates an example method 500 of operation by a RU 28 according to one or more embodiments. One or more of the steps of the method 500 may be performed in an order other than that suggested by the diagram, or performed in parallel, or in a looping fashion, e.g., such as an ongoing process that repeats.

[0091] The method 500 includes the RU 28 receiving (Block 502) two or more DL traffic streams 52 for concurrent radio transmission as respective DL RF carriers 54. The two or more DL traffic streams 52 may originate from the same DU 26 or from two or more DUs 26. Attorney Docket No. 1009-6680 / P108970W001 However, in one or more embodiments, each of the DL traffic streams 52 is scheduled independently — here, each traffic stream 52 is received from a respective scheduler 50. The reception step may be an ongoing process, e.g., with traffic stream data received for successive timeslots forming subframes, with defined groupings or sets of subframes forming frames that repeat regularly.

[0092] The method 500 further includes the RU 28 forecasting (Block 504) an aggregate utilization on an ongoing basis with respect to future time intervals. This forecasting may be ongoing and operate some defined temporal distance in the future, on a recurring or ongoing basis. “Aggregate utilization” refers to the consumption of one or more shared operational resources at the RU 28, for reception and processing of the DL traffic streams 52 and generation and transmission of the corresponding DL RF carriers 54. In at least one embodiment, “aggregate utilization” refers to the aggregate transmit power requirements for transmission of the DL traffic streams 52.

[0093] Responsive to determining that the forecasted aggregate utilization will exceed a limit for the shared operational resource at a future time interval (YES from Block 506), the method 500 includes the RU 28 performing (Block 508) a mitigation operation, with processing also continuing with ongoing reception of the DL traffic streams 52. If there are no forecasted overutilization events (NO from Block 506), processing simply continues with reception of the DL traffic streams 52.

[0094] Performing the mitigation operation comprises the RU 28 selecting (Block 510) one or more of the two or more DL traffic streams 52 according to stream priorities, and, with respect to occurrence of the respective future time interval, reducing (Block 512) a per-stream utilization of the shared operational resource by each selected DL traffic stream 52. For each DL traffic stream 52, there is a corresponding nominal utilization of the shared operational resource for the respective future time interval, and reducing the per-stream utilization for each selected DL traffic stream 52 comprises modifying transmission or associated transmit processing of the selected DL traffic stream 52 such that actual utilization of the shared operational resource for the selected DL traffic stream 52 is below the corresponding nominal utilization. As was explained earlier, the corresponding nominal utilization may be understood as the utilization — resource allocation — that would be made absent the overload condition.

[0095] In an example embodiment or operational scenario where the shared operational resource is transmit power, each DL traffic stream 52 has a corresponding nominal or required transmit power, and the mitigation operation comprises selecting one or more of the DL traffic streams 52 according to stream priorities, and, at the future time, reducing the transmit power allocation for Attorney Docket No. 1009-6680 / P108970W001 the DL RF carrier(s) 54 corresponding to the selected DL traffic stream(s) 52 below the nominal allocation(s) that would have been used absent the forecasted overutilization.

[0096] Further, in at least one embodiment, performing the mitigation operation (Block 508) includes the RU 28 notifying the affected DU(s) 26 of the upcoming mitigation, with the notification(s) sent in response to determining that the forecasted aggregate utilization will exceed a limit for the shared operational resource at the future time interval. The notification takes the form of MCF 100 in one or more embodiments, where each DU 26 associated with a DL traffic stream 52 that is selected for reduced allocation receives SSI 102 indicating that selection. In at least one such embodiment, the MCD 106 includes details regarding the reduced allocation, allowing the involved scheduler 50 to make more informed decisions about how to change the scheduling with respect to the future time(s) at which the reduce allocation(s) is / are applicable.

[0097] The method 500 in one or more embodiments includes the RU 28 receiving the stream priorities as incoming control information from one or more schedulers 50 — e.g., MCI 84 in Figure 2, including SPI 86. If a given DU 26 transmits more than one DL traffic stream 52 to the RU 28, each such stream can be considered at least logically as originating from a respective scheduler 50. That is, it is assumed that each DL traffic stream 52 is scheduled independently of the others.

[0098] Forecasting the aggregate utilization comprises, for example, the RU 28 predicting the aggregate utilization from per-stream demand forecasts. Here, the per-stream demand forecasts are received from one or more of the schedulers 50 in one or more embodiments — e.g., see LFI 92 as a component of MCI 84 in Figure 2. In one or more other embodiments, the RU 28 determines the demand forecasts, with such operation being transparent to the DUs 26. For example, the RU 28 in one or more embodiments determines the per-stream forecasts based on current or previous characteristics of each traffic stream.

[0099] In one example, the shared operational resource is a computing resource of the RU 28 that is used for processing the two or more traffic streams to form a multi-carrier radio signal that conveys the two or more traffic streams. As another example, the shared operational resource is an operating temperature margin of the RU 28, where processing or transmission of each DL traffic stream 52 reduces the operating temperature margin by contributing to an overall operating temperature of the RU 28.

[0100] As yet another example, the shared operational resource is an overall transmit power of the RU 28 or of a particular PA within the RU 28. Reducing the per-stream utilization of the overall transmit power of the RU 28 by each selected DL traffic stream 52 comprises reducing a transmit power used for transmission of the selected DL traffic stream 52. That is, reduced Attorney Docket No. 1009-6680 / P108970W001 transmit power is allocated to the corresponding DL RF carrier 54. Reducing the transmit power used for transmission of the selected traffic stream comprises reducing an average transmit power and / or reducing an instantaneous transmit power.

[0101] The method 500 in one or more embodiments includes the RU 28 providing feedback to one or more of the two or more schedulers 50-see the MCF 100 in Figure 2. The feedback provides advance indication that the forecasted aggregate utilization will exceed the limit for the shared operational resource at the respective future time interval. As such, the feedback enables the scheduler(s) 50 that is / are responsible for the one or more selected traffic streams to adjust scheduling operations with respect to the respective future time interval to compensate for the reduced per-stream utilization that will be imposed by the RU 28.

[0102] Figure 6 illustrates an example RU 28 according to another example embodiment, where the RU 28 comprises a number of processing units or modules. For example, the modules comprise functional logic circuits realized via the execution of computer program instructions by general-purpose or special-purpose computing hardware.

[0103] In the illustration, the modules include a receiving module 600, a forecasting module 602, a determining module 604, a mitigating module 606, and a transmitting module 608. The receiving module 600 is configured for receiving and processing the multiple DL traffic streams 52, the forecasting module 602 is configured to forecast the aggregate resource utilization represented by the multiple DL traffic streams 52 with respect to any one or more shared operational resources of the RU 28. The determining module 604 is configured to determine whether the forecasted aggregate utilization exceeds utilization limits at any future times, and the mitigating module 606 is configured to perform one or more mitigations with respect to any predicted overutilization. The transmitting module 608 is configured for transmission of the DL RF carriers 54, e.g., via control of RF transmit circuitry.

[0104] In one or more embodiments of the RU 28, the collection of modules includes a feedback module 610. Here, the feedback module 610 is configured to provide feedback — e.g., a simple indication or richer information — to each scheduler 50 that is associated with a DL traffic stream 52 that has been selected for mitigation operations at a forthcoming time for which the RU 28 has forecasted an overutilization of one or more operational resources at the RU 28.

[0105] Figure 7 illustrates an example method 700 of operation by a DU 26 according to one or more embodiments. Here, the DU 26 in question is one among two or more DUs 26 that are sharing a RU 28. One or more of the steps of the method 700 may be performed in an order other than that suggested by the diagram, or performed in parallel, or in a looping fashion, e.g., such as an ongoing process that repeats. Attorney Docket No. 1009-6680 / P108970W001 In particular, the method 700 can be understood as a method of operation by a scheduler 50 of a DU 26, with the method 700 including the scheduler 50 forming (Block 702) a DL traffic stream 52, based on making ongoing scheduling decisions with respect to per-user traffic associated with a plurality of users — a plurality of UEs 12. The method 700 further includes the scheduler 50 outputting (Block 704) the DL traffic stream 52 for radio transmission by the RU 28 as a corresponding DL RF carrier 54. The scheduler 50 is one among two or more schedulers 50 sharing the RU 28, with each such scheduler 50 outputting a corresponding DL traffic stream 52, and with one or more operational resources at the RU 28 shared with respect to processing the DL traffic streams 52 and transmitting the corresponding DL RF carriers 54.

[0106] The method 700 further includes the scheduler 50 receiving (Block 706) feedback generated by the RU 28, indicating that the RU 28 will impose a reduction in utilization by the traffic stream of a shared operational resource at the RU 28 for a future time interval. This reduction decision is made by the RU 28 responsive to the RU 28 determining that an aggregate utilization associated with sharing of the RU 28 will exceed a limit for the shared operational resource. Correspondingly, the method 700 includes the scheduler 50 making (Block 708) the scheduling decisions with respect to the future time interval in dependence on the imposed reduction in utilization — e.g., the scheduler 50 adapts its scheduling to reduce or eliminate deleterious effects that would otherwise be caused by the reduced resource allocation at the RU 28.

[0107] Note that in a variation of the method 700, the RU 28 provides no feedback to the DU 26, such that steps 706 and 708 are not performed. With or without the feedback and adaptation steps, the method 700 in one or more embodiments includes the scheduler 50 / DU 26 providing demand information in or with the traffic stream it outputs in Block 704, for use by the RU 28 forecasting resource demands of the traffic stream at the RU 28. For example, the demand information may be sent as or within the LFI 92 shown in Figure 2. Thus, in at least one embodiment, the method 700 includes the scheduler 50 outputting demand information for the RU 28, for use by the RU 28 in forecasting aggregate utilization of a shared operational resource at the RU 28.

[0108] For example, the demand information indicates a demand placed directly or indirectly on the shared operational resource in association with transmission of the involved traffic stream by the RU 28. The demand information comprises, for example, forecasted demand information and the method 700 further comprises the scheduler 50 generating the forecasted demand information based on monitoring one or more characteristics of the traffic stream.

[0109] In embodiments in which the RU 28 provides feedback for scheduling adaptation, the feedback in at least one such embodiment comprises a flag indicating a forthcoming imposed Attorney Docket No. 1009-6680 / P108970W001 reduction in utilization. In at least one such embodiment, the flag indicates an estimated amount of reduction to be imposed at the RU 28 on the traffic stream of the DU 26. As for imposition of the reduction, the future time interval is one or more ones among a continuing succession of timeslots, for example, where the scheduler 50 makes the scheduling decisions on a per-timeslot basis, and where the feedback indicates the one or more ones of the future timeslots corresponding to the imposed reduction.

[0110] Figure 8 illustrates a DU 26 according to an example embodiment, where the DU 26 comprises a number of processing units or modules. For example, the modules comprise functional logic circuits realized via the execution of computer program instructions by general- purpose or special-purpose computing hardware.

[0111] In the illustration, the modules include a traffic receiving module 800, a scheduling module 802, an outputting module 804, and, in one or more embodiments, a feedback (FB) receiving module 806.

[0112] The traffic receiving module 800 receives user traffic for transmission to respective UEs, and the scheduling module 802 processes the received user traffic according to its scheduling decisions to form a traffic stream for transmission by the RU 28. The outputting module 804 outputs the traffic stream via the interface between the DU 26 and the RU 28. In one or more embodiments, the outputting module 804 also outputs demand information, or at least stream priority information, for use by the RU 28 in determining when and how to apply mitigation to avoid or reduce overutilization events.

[0113] In one or more embodiments, the collection of modules includes the FB receiving module 806, which receives feedback from the RU 28 regarding forthcoming mitigations. The FB receiving module 806 provides the feedback to the scheduling module 802, which, in such embodiments, is configured to adapt scheduling with respect to the future time interval(s) indicated in the feedback. The feedback also may indicate the nature or extent of mitigation to be applied to the involved traffic stream, and the scheduling module 802 may base the nature or extent of its adaptations on the indicated nature or extent of the forthcoming mitigation.

[0114] Notably, modifications and other embodiments of the disclosed invention(s) will come to mind to one skilled in the art having the benefit of the teachings presented in the foregoing descriptions and the associated drawings. Therefore, it is to be understood that the invention(s) is / are not to be limited to the specific embodiments disclosed and that modifications and other embodiments are intended to be included within the scope of this disclosure. Although specific terms may be employed herein, they are used in a generic and descriptive sense only and not for purposes of limitation.

Claims

1. Attorney Docket No. 1009-6680 / P108970W001 CLAIMSWhat is claimed is:

1. A method (500) of operation by a Radio Unit (28), the method (500) comprising: receiving (502) two or more traffic streams (52) for concurrent radio transmission, wherein each traffic stream (52) is received from a respective scheduler (50), and wherein each traffic stream (52) individually contributes to an aggregate utilization of a shared operational resource (140, 142, 144) of the Radio Unit (28); forecasting (504) the aggregate utilization on an ongoing basis with respect to future time intervals; and determining (506) that the forecasted aggregate utilization will exceed a limit for the shared operational resource (140, 142, 144) at a respective one of the future time intervals and, in response, performing (508) a mitigation operation comprising: selecting one or more of the two or more traffic streams (52) according to stream priorities; and with respect to occurrence of the respective future time interval, reducing a perstream utilization of the shared operational resource by each selected traffic stream (52).

2. The method (500) according to claim 1, wherein, for each traffic stream (52), there is a corresponding nominal utilization of the shared operational resource (140, 142, 144) for the respective future time interval, and wherein reducing the per-stream utilization for each selected traffic stream (52) comprises modifying transmission or associated transmit processing of the selected traffic stream (52) such that actual utilization of the shared operational resource for the selected traffic stream (52) is below the corresponding nominal utilization.

3. The method (500) according to claim 1 or 2, further comprising receiving the stream priorities as incoming control information from one or more of the schedulers (50).

4. The method (500) according to any one of claims 1-3, wherein forecasting the aggregate utilization comprises predicting the aggregate utilization from per-stream demand forecasts, and wherein the per-stream demand forecasts are received from one or more of the schedulers (50) or forecasted by the Radio Unit (28).Attorney Docket No. 1009-6680 / P108970W0015. The method (500) according to claim 4, wherein the method (500) includes the Radio Unit (28) determining the per-stream demand forecasts based on current or previous characteristics of each traffic stream (52).

6. The method (500) according to any one of claims 1-5, wherein the shared operational resource (140, 142, 144) is a computing resource of the Radio Unit (28) that is used for processing the two or more traffic streams (52) to form a multicarrier radio frequency (RF) signal (56) comprising two or more radio carriers (54), each radio carrier (54) conveying a respective one of the two or more traffic streams (52).

7. The method (500) according to any one of claims 1-5, wherein the shared operational resource (140, 142, 144) is an operating temperature margin of the Radio Unit (28), and wherein processing or transmission of each traffic stream (52) reduces the operating temperature margin by contributing to an overall operating temperature of the Radio Unit (28).

8. The method (500) according to any one of claims 1-4, wherein the shared operational resource (140, 142, 144) is an overall transmit power of the Radio Unit (28), and wherein reducing the per-stream utilization of the overall transmit power of the Radio Unit (28) by each selected traffic stream (52) comprises reducing a transmit power used for transmission of the selected traffic stream (52).

9. The method (500) according to claim 8, wherein reducing the transmit power used for transmission of the selected traffic stream (52) comprises reducing an average transmit power.

10. The method (500) according to claim 8, wherein reducing the transmit power used for transmission of the selected traffic stream (52) comprises reducing an instantaneous transmit power.

11. The method (500) according to any one of claims 1-10, wherein the concurrent radio transmission comprises the Radio Unit (28) transmitting a multicarrier radio frequency (RF) signal comprising two or more RF carriers (54), with each RF carrier (54) modulated to convey a respective one of the two or more traffic streams (52) and with each traffic stream (52) serving targeted ones among a respective group (120) of users (12).Attorney Docket No. 1009-6680 / P108970W00112. The method (500) according to any of claims 1-11, wherein receiving the two or more traffic streams (52) comprises receiving each traffic stream (52) as a stream of digital waveform samples.

13. The method (500) according to claim 1 or 2, wherein receiving the two or more traffic streams (52) comprises receiving each traffic stream (52) via a respective communication interface (80) coupling the Radio Unit (28) with the respective one among the two or more schedulers (50).

14. The method (500) according to any one of claims 1-13, wherein the Radio Unit (28) operates as a transmission point in a wireless communication network (10) operating according to Third Generation Partnership Project (3GPP) specifications.

15. The method (500) according to any one of claims 1-14, further comprising providing feedback (100-1, 100-2) to one or more of the two or more schedulers (50), the feedback providing advance indication that the forecasted aggregate utilization will exceed the limit for the shared operational resource (140, 142, 144) at the respective future time interval and enabling the scheduler or schedulers that are responsible for the one or more selected traffic streams (52) to adjust scheduling operations with respect to the respective future time interval to compensate for the reduced per-stream utilization that will be imposed by the Radio Unit (28).

16. The method (500) according to claim 15, wherein the two or more traffic streams (52) include at least a first traffic stream (52-1) originating from a first scheduler (50-1) residing in a Digital Unit (26-1) of a first wireless communications network (10-1), and a second traffic stream (52-2) originating from a second scheduler (50-2) residing in a Digital Unit (26-2) of a second wireless communications network (10-2), such that feedback (100-1, 100-2) effects indirect coordination between the first and second wireless communications networks (10-1, 10- 2).

17. A method (700) of operation by a scheduler (50) of a Digital Unit (26), the method (700) comprising: forming (702) a traffic stream (52) based on making ongoing scheduling decisions with respect to per-user traffic associated with a plurality of users (12); outputting (704) the traffic stream (52) for radio transmission by a remote Radio Unit (28) that is shared by the scheduler (50) with one or more other schedulers (50);Attorney Docket No. 1009-6680 / P108970W001 receiving (706) feedback (100) generated by the Radio Unit (28), indicating that the Radio Unit (28) will impose a reduction in utilization by the traffic stream (52) of a shared operational resource (140, 142, 144) at the Radio Unit (28) for a future time interval, responsive to the Radio Unit (28) determining that an aggregate utilization associated with sharing of the Radio Unit (28) will exceed a limit for the shared operational resource (140, 142, 144); and making or updating (708) the scheduling decisions with respect to the future time interval in dependence on the imposed reduction in utilization.

18. The method (700) according to claim 17, wherein forming the traffic stream (52) comprises forming a packetized stream of digital waveform samples, for generation of a corresponding radio signal (54) at the Radio Unit (28).

19. The method (700) according to claim 17 or 18, wherein the feedback (100) comprises a flag indicating the forthcoming imposed reduction in utilization.

20. The method (700) according to claim 17 or 18, wherein the feedback (100) indicates an estimated amount of reduction to be imposed at the Radio Unit (28).

21. The method (700) according to any one of claims 17-20, wherein the future time interval is one or more ones among a continuing succession of timeslots, with the scheduler (50) making the scheduling decisions on a per-timeslot basis, and wherein the feedback (100) indicates the one or more ones of the future timeslots corresponding to the imposed reduction.

22. The method (700) according to any one of claims 17-20, wherein the method (700) further includes the scheduler (50) outputting demand information (92) for use by the Radio Unit (28) in forecasting aggregate utilization of the shared operational resource (140, 142, 144) at the Radio Unit (28), the demand information (92) indicating a demand placed directly or indirectly on the shared operational resource (140, 142, 144) in association with transmission of the traffic stream (52) by the Radio Unit (28).

23. The method (700) according to claim 22, wherein the demand information (92) comprises forecasted demand information and wherein the method (700) further comprises the scheduler (50) generating the forecasted demand information based on monitoring one or more characteristics of the traffic stream (52).Attorney Docket No. 1009-6680 / P108970W00124. The method (700) according to any one of claims 17-23, wherein making the scheduling decisions with respect to the future time interval in dependence on the imposed reduction in utilization comprises adapting scheduling to reduce or eliminate deleterious effects that would otherwise be caused by the imposed reduction in utilization.

25. A Radio Unit (28) comprising: transmission circuitry (132, 134, 136); interface circuitry (130) configured to receive two or more traffic streams (52) for concurrent radio transmission via the transmission circuitry (132, 134, 136), wherein each traffic stream (52) is received from a respective one among two or more schedulers (50) sharing the Radio Unit (28), and wherein each traffic stream (52) individually contributes to an aggregate utilization of a shared operational resource (140, 142, 144) of the Radio Unit (28); and processing circuitry (150) configured to: forecast the aggregate utilization on an ongoing basis with respect to future time intervals; and determine that the forecasted aggregate utilization will exceed a limit for the shared operational resource at a respective one of the future time intervals and, in response, performing a mitigation operation comprising: selecting one or more of the two or more traffic streams (52) according to stream priorities; and with respect to occurrence of the respective future time interval, reduce a per-stream utilization of the shared operational resource (140, 142, 144) by each selected traffic stream (52).

26. A scheduler (50) of a Digital Unit (26), the scheduler (50) comprising: processing circuitry (172) configured to form a traffic stream(52) based on making ongoing scheduling decisions with respect to per-user traffic associated with a plurality of users (12); and interface circuitry (170) configured to output the traffic stream (52) for radio transmission by a remote Radio Unit (28) that is shared by the scheduler (50) with one or more other schedulers (50) outputting corresponding traffic streams (52) for radio transmission by the Radio Unit (28); wherein the processing circuitry (172) is further configured to:Attorney Docket No. 1009-6680 / P108970W001 receive feedback (100) via the interface circuitry (170) that is generated by the Radio Unit (28), the feedback (100) indicating that the Radio Unit (28) will impose a reduction in utilization by the traffic stream (52) of a shared operational resource (140, 142, 144) at the Radio Unit (28) for a future time interval, responsive to the Radio Unit (28) determining that an aggregate utilization of the shared operational resource (140, 142, 144) will exceed a limit for the shared operational resource (140, 142, 144); and make the scheduling decisions with respect to the future time interval in dependence on the imposed reduction in utilization.

Images