WO2018002691A1

WO2018002691A1 - Dynamic radio downlink power amplifier control for base station energy efficiency

Info

Publication number: WO2018002691A1
Application number: PCT/IB2016/053902
Authority: WO
Inventors: Yu Li; Edwin Vai Hou Iun; Ping Liu
Original assignee: Telefonaktiebolaget Lm Ericsson (Publ)
Priority date: 2016-06-29
Filing date: 2016-06-29
Publication date: 2018-01-04

Abstract

A method and base station, i.e., network node, for adaptively controlling a power setting of a power amplifier to conserve power are disclosed. According to one aspect, a method for adaptively controlling a power amplifier in a base station is provided. The method includes calculating at least one digital power level in a first time interval based on at least one digital sample of a signal collected over the first time interval. The method further includes predicting at least one digital power level in a second time interval subsequent to the first time interval based on the at least one calculated digital power level of the first time interval. The method also includes adjusting an operation of the power amplifier of the base station based at least in part on the at least one predicted digital power level.

Description

DYNAMIC RADIO DOWNLINK POWER AMPLIFIER CONTROL FOR BASE STATION ENERGY EFFICIENCY

TECHNICAL FIELD

The disclosure relates to wireless communication and in particular, dynamic radio downlink power amplifier control for base station energy efficiency.

BACKGROUND

In a typical wireless communication system, wireless devices communicate via a radio access network (RAN) with other wireless devices and core networks. FIG. 1 is a block diagram of such a typical wireless communication system 10 including a core network 12, multiple base stations 14 and multiple wireless devices 16. The core network 12 may include intermediary devices such as a mobile management entity (MME) or serving gateway (S-GW) and provide access to the Internet and the Public Switched Telephone Network (PSTN). The base stations 14 may cover different geographic regions called cells which may overlap. Thus, a wireless device 16 may communicate over the air, i.e., wirelessly, with one or more base stations to communicate voice and data between the wireless device 16 and another wireless device, a landline telephone and/or the Internet. The base stations 14 may communicate with each other over an X2 interface and communicate with the core network over an SI interface.

The term wireless device, includes user equipment (UE), mobile terminal, etc., and may refer to any type of wireless device communicating with a network node and/or with another wireless device in a cellular or mobile communication system. Examples of a wireless device are target device, device to device (D2D) wireless device, machine type wireless device or wireless device capable of machine to machine (M2M) communication, PDA, iPAD, Tablet, mobile terminals, smart phone, laptop embedded equipped (LEE), laptop mounted equipment (LME), USB dongles etc.

The term base station, e.g. a Radio Base Station (RBS), sometimes may be referred to herein as, e.g., evolved NodeB "eNB", "eNodeB", "NodeB", "B node", or BTS (Base Transceiver Station), depending on the technology and terminology used. The base stations may be of different classes such as, e.g., macro eNodeB, home eNodeB or pico base station, based on transmission power and thereby also cell size. A cell is the geographical area where radio coverage is provided by the base station at a base station site. One base station, situated on the base station site, may serve one or several cells. Further, each base station may support one or several communication technologies. The base stations communicate over the air interface operating on radio frequencies with the wireless devices within range of the base stations.

In 3rd Generation Partnership Project (3GPP) Long Term Evolution (LTE), base stations may be directly connected to one or more core networks. Further, although embodiments are described with reference to base stations, it is understood that embodiments can be implemented in or across any suitable network node, of which base stations are a type.

The term "downlink" refers to transmission from the base station 14 to the wireless devices 16, and the term "uplink" refers to transmission from a wireless device 16 to the base station 14. Downlink power consumption refers to power consumed by the transmitter of the base station in the process of transmitting signals on the downlink. Downlink (DL) power consumption at base stations should be controlled to minimize costs and to minimize interference with communications in unlicensed spectrums. Also, the introduction of multiple input multiple output (MIMO) permits a higher antenna gain such that the DL power requirement at each antenna port will be lower. With the advent of fifth generation (5G) networks, network equipment should utilize 5G power saving features, while remaining backward compatible. In particular, the next generation of network equipment should be able to detect and adapt to its surroundings while avoiding protocol changes.

The need for high energy performance has become a factor for network infrastructure at least because:

(1) High network energy performance is crucial to reducing operational cost, and is a driver for better node and network dimensioning, which leads to reduced Total Cost of Ownership (TCO).

(2) High network energy performance allows for off-grid network deployment relying on solar panels of practical size as power supplies, enabling wireless connectivity to even the most remote areas. (3) High network energy performance is part of a general operator aim to provide wireless access in a sustainable and more resource-efficient way.

In addition, to conform to 5G requirements, the network equipment should conserve power whenever possible. In a typical network, traffic is unevenly distributed among base stations, with some base stations actively serving many wireless devices and other base stations actively serving relatively few wireless devices. Further, traffic varies greatly with time of day and day of week. For example, traffic is typically lower during the night time as compared to daytime usage.

Radio Frequency (RF) branch settings are typically calibrated in advance for heavy traffic, and these settings typically remain static. Networks are designed for continuous and highly reliable operation, which traditionally has been associated with an "always on" design, implying that nodes and components are always on in order to be immediately available when needed. The power level of the Power Amplifier (PA) in the Radio Base Station (RBS) is set based on the planned static base station configuration for the nominal power level of each carrier. The nominal power level is the carrier power level at a given nominal traffic condition. Once the carriers are configured and activated on the RBS, the power level of the PA will be biased to the target level and no further adjustment will be made when the baseband traffic level changes. Due to this static nature of configuration, it is not surprising that network energy remains high even though the customer traffic demand may be low.

Currently the RF branches are calibrated for the worst case nominal (i.e. heavy) traffic. However, field measurements indicate that the traffic variation is periodic in 24 hours. There are also weekly patterns (e.g., traffic is heavy during weekdays in a downtown base station) and yearly patterns (e.g., a base station near a college campus experiences lower traffic during the summer months). Further, studies show that only 5% of network cells are mostly loaded and about 95% of them are not fully loaded. Thus, tuning a PA to account for busy times causes a tax on lesser loaded base stations, and is inefficient during quiet periods. Also, in- phase/quadrature (IQ) traffic throughput may also depend on a type of data being transmitted (e.g., data, voice, image, or video). Further, excess power provided by the PA results in energy loss via heat dissipation which may be compensated by ventilation and heat sinks, thereby increasing the costs of deployment.

In one conventional approach to power conservation, a scheduler assigns user data to center subcarriers so that the physical resource blocks (PRB) at edges of a frequency spectrum are not used. In other words, the number of subcarriers is reduced. This approach saves power but user data may be delayed if in any specific time the number of PRBs requested exceeds the number of PRB available. This user delay is undesirable and may not meet the requirements of 5G networks.

In another conventional approach to power conservation, as much data as possible is batched in one sub frame, which is two time slots in LTE. When there is no user data in one sub frame, the PA can be turned off completely. Thus, this approach groups data in the time domain and also suffers from user data delay.

SUMMARY

Some embodiments advantageously provide a method and base station, i.e., network node for adaptively controlling a power setting of a power amplifier to conserve power. According to one aspect, a method for adaptively controlling a power amplifier in a base station is provided. The method includes calculating at least one digital power level in a first time interval based on at least one digital sample of a signal collected over the first time interval. The method further includes predicting at least one digital power level in a second time interval subsequent to the first time interval based on the at least one calculated digital power level of the first time interval. The method also includes adjusting an operation of the power amplifier of the base station based at least in part on the at least one predicted digital power level.

According to this aspect, in some embodiments, predicting at least one digital power level in the second time interval is further based in part on historical data. In some embodiments, the historical data is digital power levels measured during a particular period of time in the past, the particular period of time being one of a time period during a particular day, a number of days of a particular week and a number of weeks in a particular year. In some embodiments, the predicted at least one digital power level are computed recursively. In some embodiments, the method further includes computing an average of a plurality of predicted digital power levels, wherein adjusting the operation of the power amplifier is further based at least in part on the average of the plurality of predicted digital power levels. In some

embodiments, the method further includes computing an average of historical data of past recurring time intervals, wherein predicting at least one digital power level in the second time interval is further based at least in part on the average of historical data. In some embodiments, adjusting the operation of the power amplifier is further based at least in part on a margin of improvement of an error vector magnitude, EVM. In some embodiments, predicting at least one digital power level in the second time interval is further based at least in part on a margin of improvement of at least one key performance indicator. In some embodiments, adjusting the operation of the power amplifier is further based at least in part on a downlink power level at nominal traffic conditions. In some embodiments, adjusting the operation of the power amplifier is further based at least in part on a nominal power amplifier power level configured for nominal traffic conditions. In some embodiments, adjusting the operation of the power amplifier includes adjusting a bias parameter of the power amplifier to affect an amplification of the power amplifier. In some embodiments, the method includes overriding the adjustment step in response to a user command.

According to another aspect, a base station having a power amplifier is provided. The base station includes processing circuitry comprising a memory and a processor. The memory is configured to store digital power level data. The processor is configured to calculate at least one digital power level in a first time interval based on at least one digital sample of a signal collected over the current time interval, predict at least one digital power level in a second time interval subsequent to the first time interval based on the at least one calculated digital power level of the first time interval, and adjust an operation of the power amplifier of the base station based at least in part on the at least one predicted digital power level.

According to this aspect, in some embodiments, the base station further includes a link to a long term storage facility to store historical data, and wherein predicting at least one digital power level in the second time interval is further based in part on historical data. In some embodiments, the historical data is digital power levels measured during a particular period of time in the past, the particular period of time being one of a time period during a particular day, a number of days of a particular week and a number of weeks in a particular year. In some embodiments, the predicted at least one digital power level are computed recursively. In some embodiments, the processor is further configured to compute an average of a plurality of predicted digital power levels, wherein adjusting the operation of the power amplifier is further based at least in part on the average of the plurality of predicted digital power levels. In some embodiments, the processor is further configured to compute an average of historical data of past recurring time intervals, wherein predicting at least one digital power level in the second time interval is further based at least in part on the average of historical data. In some embodiments, the processor is further configured to adjust the operation of the power amplifier based at least in part on a margin of improvement of an error vector magnitude, EVM. In some embodiments, the processor is further configured to predict at least one digital power level in the second time interval based at least in part on a margin of improvement of at least one key performance indicator. In some embodiments, the processor is further configured to predict at least one digital power level in the second time interval based at least in part on a downlink power level at nominal traffic conditions. In some embodiments, the processor is further configured to adjust the operation of the power amplifier based at least in part on a nominal power amplifier power level configured for nominal traffic conditions. In some embodiments, adjusting the operation of the power amplifier further includes adjusting a bias parameter of the power amplifier to affect an amplification of the power amplifier. In some embodiments, the processor is further configured to receive an override command from an operator to override the adjusting by the processor.

According to yet another aspect, in some embodiments, a base station having a power amplifier is provided. The base station includes a downlink demand measurement module configured to calculate at least one digital power level in a first time interval based on at least one digital sample collected over the current time interval. The base station also includes a power demand prediction module configured to predict at least one digital power level in a second time interval subsequent to the first time interval based on the at least one calculated digital power level of the first time interval. The base station also includes an adjustment module configured to adjust an operation of the power amplifier of the base station based at least in part on the at least one predicted digital power level.

According to this aspect, in some embodiments, the base station further includes a link to a long term storage module configured to store historical data relating to power consumption by the power amplifier. In some embodiments, a short term storage module is configured to store the calculated at least one digital power level. In some embodiments, the base station further includes an operator overwrite module to override the adjustment of the adjustment module. In some embodiments, the base station further includes a key performance indicator, KPI, module configured to supply to the adjustment module a predetermined margin of improvement of the KPI.

BRIEF DESCRIPTION OF THE DRAWINGS

A more complete understanding of the present embodiments, and the attendant advantages and features thereof, will be more readily understood by reference to the following detailed description when considered in conjunction with the accompanying drawings wherein:

FIG. 1 is a block diagram of a conventional wireless communication network; FIG. 2 is a block diagram of a base station, core network and antenna system, the base station being configured to adaptively control a power amplifier of the base station;

FIG. 3 is a block diagram of a base station showing processing circuitry used to implement a baseband processing unit and a radio processing unit;

FIG. 4 is a block diagram of a base station configured to adaptively control a power amplifier of the base station;

FIG. 5 is a block diagram of an alternative embodiment of a base station configured to adaptively control a power amplifier of the base station;

FIG. 6 is a flowchart of an exemplary process for adaptively controlling a power amplifier; and

FIG. 7 is a flowchart of another exemplary process for adaptively controlling a power amplifier. DETAILED DESCRIPTION

Before describing in detail exemplary embodiments, it is noted that the embodiments reside primarily in combinations of apparatus components and processing steps related to dynamic radio downlink power amplifier control for base station energy efficiency. Accordingly, components have been represented where appropriate by conventional symbols in the drawings, showing only those specific details that are pertinent to understanding the embodiments so as not to obscure the disclosure with details that will be readily apparent to those of ordinary skill in the art having the benefit of the description herein.

As used herein, relational terms, such as "first" and "second," "top" and

"bottom," and the like, may be used solely to distinguish one entity or element from another entity or element without necessarily requiring or implying any physical or logical relationship or order between such entities or elements.

In some embodiments, the control of the power level of the PA is based on real time digital power level data of cells served by a radio transmission port, and may also be based on historical digital power level data at cells served by the radio transmission port. The historical digital power level data, referred to herein as historical data, is digital power levels measured during a time in the past. The time in the past may be one of a time period of a particular day, a number of days of a particular week and a number of weeks in a particular year. The period of time and day from which the historical data is taken may correspond to a next time interval for which anticipated digital power levels are to be predicted. For example, if the next time interval falls between the hours of 9am to 5am during weekdays, the historical data may be selected from the same time period on weekdays of previous weeks.

The predictions of future digital power levels may be performed at the base station which has the PA or may be based on predictions from one or more other base stations. The real time digital power level data and the historical data are used as an input to a model to derive a predicted average digital power level for a next time interval. This predicted average digital power level is in turn used as an input to another model to calculate the optimum power level of the PA based on other factors such as PA headroom (which is associated with, e.g., modulation scheme) and the available key performance indicators (KPI). In some embodiments, the KPI may include channel state information (CSI) or other channel quality measurements from wireless devices. The time interval may be based on the rate of change of predicted digital power level and historical data, taking into consideration the impact upon service when the PA power level is adjusted via a power level tuning procedure. Thus, for slow changes in predicted digital power levels, a larger time interval may be used, whereas for rapid changes in predicted digital power levels, a smaller time interval may be used. Since the analog power level of the PA is directly driven by the digital power level of the in phase/quadrature (IQ) data, using digital power levels to control PA power is accurate and efficient.

Some embodiments provide lower energy consumption which lowers cost of operation of the network equipment, since as much as 20% of power consumption can be saved based on energy conservation method studies. Reduction in energy consumption may result in operation of the PA at optimum power efficiency, thereby reducing heat dissipation. Reduction of heat dissipation improves mean time between failures due to lower temperature stress on electrical components. Also, risk of PA shutdown due to excessive high temperature will be reduced. As a countering factor, the PA size may sometimes be relaxed from the absolute minimum to avoid over- optimization, thereby reducing RF signal distortion and improving DL performance.

Also, lower power consumption facilitates implementation of fifth generation (5G) networks. 5G uses higher frequencies which results in decreased signal travel distance, which in turn increases base station density, which in turn increases network power consumption. Therefore, lower power consumption by PAs in the network allows off-grid base stations and lower heat dissipation allows for further integrated circuit and printed circuit board (PCB) miniaturization.

According to some embodiments, a PA is only active when and if needed.

This is achieved without baseband scheduler tempering which reduces carrier bandwidth to achieve power conservation. Thus, a difference between conventional methods for power conservation and embodiments described below is the lack of rescheduling. Further, no 3GPP protocol changes are required to implement the methods described herein. The methods described herein enable the base station to learn in real time and adjust the PA setting dynamically. A further advantage to embodiments described herein is that there may be no need to change hardware, but merely to use a new prediction methodology that can be configured in software executed by a processor that is currently used in a base station. Bulk data storage used to store historical data may be preexisting for storing KPI information for a network. This historical data may be provided to the processor using existing data transfer protocols. In some embodiments the dynamic adjustment of the PA setting is internal to the base station, although some or all of the functions performed to achieve such adjustment could be distributed to other network nodes. In some embodiments, the base station performing the PA adjustment knows the existing traffic conditions based on reports from wireless devices served by the base station. This knowledge can be used as inputs to a prediction module to aid in predicting the digital power level in a next time interval.

As mentioned, storage of historical data may be in a facility located away from the base stations since the base stations may not have enough storage capacity to store all the required data. Transfer of the historical data to and from the base station may be performed using existing protocols.

FIG. 2 is a block diagram of a base station 15 configured for dynamic radio downlink power amplifier control for base station energy efficiency. The base station 15 includes a baseband processing unit 20 and a radio processing unit 22 and an antenna system 24. In some embodiments, the antenna system 24 may be located remote from the base station 15. The base station 15 is coupled to a core network 26 having a long term storage facility 28. The long term storage facility 28 may consist of non-volatile memory that is readable and writable and may include disk memory, flash memory, etc. The base station 15 communicates with wireless devices 16 via the antenna system 24. The base station 15 also includes a power amplifier 30 that may be controlled for optimal power transfer based on digital power levels and historical data. In one embodiment, the core network 26 may be or include a high capacity communication facility that connects primary nodes so that they can exchange data and services. The long term storage facility 28 stores historical digital power levels for use in predicting future power levels used to adjust an operating point of the power amplifier 30. Traditionally, settings of the downlink power amplifier 30 are configured during installation of the base station as part of the planned static configuration 32 which may be stored in the baseband processing unit 20. These settings are applied initially and conventionally, and are never or seldom changed. The power amplifier setting stored in the planned static configuration 32 is fed to a power amplifier configuration unit 34 which facilitates setting of the power amplifier 30 via the PA adjustment unit 36. The PA adjustment module 36 determines appropriate register values for configuring the PA 30. These values are applied as a control signal to the PA 30. In some embodiments, the control signal adjusts a bias parameter to alter an operating point of the PA.

A drawback with a traditional static manner of setting the DL PA is that the setting is designed to handle a nominal scenario. Since the setting is static, during normal conditions there is wasted transmit power and unnecessary power consumption. To overcome this and other drawbacks, in some embodiments, the PA 30 may be set dynamically according to predicted digital power levels determined by a PA demand prediction unit 38 of a dynamic configuration block 40 in the radio processing unit 22.

The dynamic configuration block 40 includes a downlink demand

measurement unit 42, short term storage 46 and power demand prediction unit 38. The downlink demand measurement unit 42 captures digital power levels of data samples received from the DL data plane 44 in a current time interval. The DL data plane 44 receives the data samples from the core network 26 and encodes the data samples. The data samples may be from another wireless device or landline telephone or from the Internet, for example. The data samples from data plane 44 are also sent to the PA 30 where they are amplified and radiated by the antenna system 24.

The digital power levels of the current time interval are received from the downlink demand measurement unit 42 and stored in a short term storage unit 46 and is also periodically pushed to the long term storage facility 28. The short term storage unit 46 may consist of volatile or non-volatile memory that is readable and writable, and may include memory such as random access memory (RAM) and flash memory. The power demand prediction unit 38 receives input data from the short term storage unit 46, from the long term storage facility 28 and KPI feedback control unit 48. The power demand prediction unit 38 implements a model that uses the digital power levels received from the short term storage unit 46 combined with historical data from the long term storage facility 28 and KPI data from unit 48 to forecast an amount of anticipated downlink digital power levels. This forecasted data is considered by the PA adjustment unit 36 to influence the setting of the PA 30 during the next time interval.

Note that the current time interval is an interval of time during which samples are taken and the next time interval is an interval of time subsequent to the current time interval for which power levels are predicted.

The KPI feedback control 48 sends the IQ modulation scheme, which may be, for example, quadrature phase shift keying (QPSK) or quadrature amplitude modulation (QAM), and measured KPIs to the PA demand prediction unit 38 to alter the power prediction to include a margin of headroom to safeguard any over- optimization by the PA demand prediction unit 38 based on digital power levels and historical data. Thus, for example, if the IQ modulation scheme is of a high order, a margin may be added to the predicted digital power levels and historical digital power levels to increase the optimal power predicted for the next time interval.

An operator overwrite switch 50 enables the operator to bypass the PA power demand prediction unit 38. In one embodiment, the overwrite switch 50 may be a logical switch responsive to input from the operator via a graphical user interface, for example. This enables the operator to respond to a natural disaster, for example. Note that changing the operating point of the PA 30 may also require changes to settings along the entire downlink chain as is well established base station functionality.

Note that the functions described as being performed by the base station 15 may, in other embodiments, be distributed among more than one network node, i.e., they can be performed remote from the base station 15. Similarly, some of the functions described as being performed in the baseband processing unit 20 may instead be performed in the radio processing unit 22. Conversely, some of the functions described as being performed in the radio processing unit 22 may instead be performed in the baseband processing unit 20. The downlink demand measurement unit 42 may measure digital power levels by one of several methods. One method includes calculating the digital power levels on each of a plurality of filter branches based on an amplitude root mean square (ARMS) of all digital samples collected over a current time interval such as a radio frame or time slot according to the following formula:

N

DPM = [∑(I_k ² + Q_k ²)] / N implemented by a digital power meter (DPM), where k is the index of samples during the time interval, I_k is the in-phase component of the digital signal, Q_k is the quadrature component of the digital signal, N is the number of samples collected. The I and Q (in phase and quadrature) samples may be obtained from the data plane 44. In some embodiments, the computed value of the digital power meter (DPM) is amplified per carrier to achieve a desired preliminary power level. The amplified digital power from all filter branches are summed and corrected using digital pre- distortion to compensate for the non-linearity of the PA 30, as is known in the art. The digital pre-distorted signal is converted to analog before being input to the PA 30. Thus, the digital power levels of the I and Q channels dictate the predicted digital power levels by the PA power demand prediction unit 38.

In some embodiments, a system transfer function and the least squares system identification method are employed for digital power level prediction. The system transfer function may be given as follows:

Y(z blZ^n→ + b2Z^n~2 + - + bn

= G(z) =

U(z) Zⁿ - alZ^n→ - a2Z^n~2 an where U(z) is the system input (time) and Y(z) is the system output (the digital power level for the DL branch supported by the PA). The n^th order difference equation, also known as the auto regression moving average model, that describes the prediction model can be represented as:

y(/ ) = al y(k - 1) + a2 y(k - 2) + · ·· + any(k - n) + bl u(k - 1)

+ b2 u{k— 2) + · ·· + bnu(k— n) where u(k) is the time at a prediction point k in a next time interval, and y(k) is the predicted downlink digital power level at the prediction point k in the next time interval. Thus, the predicted digital power levels may be computed recursively.

The coefficients a_n and b_n , which in vector form can be expressed as:

Θ = (al a2 ... an bl b2 ... bn)^T can be determined from measurements of the input-output sequence u(k) and y(k) using a set of N+l measurement pairs collected by the downlink demand measurement unit 42 from a current time interval:

{ u(0), y(0)}, {u(l), y(l)}, ... { u(N), y(N)} with N≥n≥0

Using the least square system identification method, the least square estimation of the coefficients a_n and b_n can be determined using the following matrix calculation based on the N+l measurement pairs:

Θ = [F^T(N F(N ]-¹F^T(N). y(N

where

and

fT(k) = [y(/c - 1) y k - 2) ... y(/ - ri) u(k - 1) u(k - 2) u(k - n)].

In words, the system identification process (learning process) will be carried out by calculating the difference between the predicted DL digital power level and the measured DL digital power level. The sum of the square of absolute difference of the sample points collected in a fixed time interval is computed and the coefficients are then adjusted to make the sum of the square of absolute error a minimum. Once the coefficient vector is calculated, a certain number of predicted digital power level values can be calculated using a formula for different prediction points in the next time interval. The average of a certain number of predicted digital power level values calculated for the next time interval will then be calculated using the following formula:

The average of certain numbers of historical digital power level values for the next time interval may than be calculat the following formula:

The historical data may be averaged over a number of past recurring time intervals. These past recurring time intervals may correspond to a next time interval for which digital power levels are predicted. The predicted downlink digital power level for the next time interval may then be calculate as the average of Tp and Th, as follows:

Tp + Th

T^{= J}^-

The predicted downlink digital power level for the next time interval may then be used to determine the optimal PA power level for the next time interval using the following formula:

p = Pn x (—) + Pm + Pk

XTn)

where

f=Target PA Power Level Setting (in Watt or dBm)

Pn= Nominal PA power level configured for nominal traffic condition;

T= The predicted downlink digital power level for the next time interval; Tn= DL digital power level at nominal traffic condition;

Pm= PA margin reserved for error vector magnitude (EVM) improvement, which will be determined, based on the RAN types and the IQ data modulation schema received from the baseband unit for instance; and Pk= PA margin reserved for KPI improvement, which will be determined based on the RAN types and the signal error rate that is measured by the wireless device and received from the baseband unit.

Some or all of these parameters may be omitted from the calculation of P. Once the optimal power level is determined using these formulas, the power amplifier can be adjusted to operate so that the average or peak power of the power amplifier equals the predicted power. The adjustment can be made by varying a bias parameter of the power amplifier.

Note that in some embodiments, the optimal power level for a power amplifier may be based on predicted digital power levels and historical data from another base station. For example, historical data measured at a first base station may be used in the prediction of future power usage by a second base station.

Note also that the PA margin reserved for EVM improvement may depend on the order of modulation implemented. For example, a higher margin may be used for higher orders of modulation.

As noted above, the operator may override this computed power setting manually. Note also that base stations deployed bear a business district, which has high usage during the day and much lower usage at night, may be adjusted differently than base stations near major entertainment areas such as sports stadiums which may have higher usage at nights when games are played. These differences in historical usage are accounted for in the above formula by the variable, Th. As a further example, downtown base stations may have different historical data than base stations covering a college campus.

Also, different quality of service provided by different base stations demand different signal strengths. Low latency data traffic requires higher order modulation so that more data can be transmitted in the same period of time. Higher order modulation results in higher peak to average power ratio (PAR). Often higher order modulation requires larger PAR headroom. However, when the baseband processing unit 20 detects that the quality is not high, for example, the baseband processing unit may lower the order of modulation to reduce PAR headroom to allow PA operation at a more efficient range based on need. FIG. 3 is a block diagram of an alternative view of the base station 15 showing that the baseband processing unit 20 and radio processing unit 22 may be

implemented by processing circuitry 52 which includes a memory 54. In addition to a traditional processor and memory, processing circuitry 52 may comprise integrated circuitry for processing and/or control, e.g., one or more processors and/or processor cores and/or FPGAs (Field Programmable Gate Array) and/or ASICs (Application Specific Integrated Circuitry).

Processing circuitry 52 may comprise and/or be connected to and/or be adapted for accessing (e.g., writing to and/or reading from) memory 54, which may comprise any kind of volatile and/or non-volatile memory, e.g., cache and/or buffer memory and/or RAM (Random Access Memory) and/or ROM (Read-Only Memory) and/or optical memory and/or EPROM (Erasable Programmable Read-Only

Memory). Such memory 54 may be configured to store code executable by control circuitry and/or other data, e.g., data pertaining to communication, e.g., configuration and/or address data of nodes, etc. Processing circuitry 52 may be adapted to control any of the methods described herein and/or to cause such methods to be performed, e.g., by processor 22. Corresponding instructions may be stored in the memory 54, which may be readable and/or readably connected to the processing circuitry 52. In other words, processing circuitry 52 may include a controller, which may comprise a microprocessor and/or microcontroller and/or FPGA (Field-Programmable Gate

Array) device and/or ASIC (Application Specific Integrated Circuit) device. It may be considered that processing circuitry 52 includes or may be connected or connectable to memory, which may be configured to be accessible for reading and/or writing by the controller and/or processing circuitry 52.

A more detailed view of the radio processing unit 22 within the base station 15 is shown in FIG. 4. Baseband processing unit 20 is not shown in FIG. 4 only for the sake of simplicity and to aid understanding of the features and functionality of the radio processing unit 22. Referring to FIG. 4, the short term storage 46 is configured to store digital power level data 56 captured by a DL demand measurement (digital power meter (DPM)) unit 42 of the processor 58. In some embodiments, the processor 58 may also execute software to implement a power demand prediction unit 38. The power demand prediction unit 38 may perform the calculations set forth above for determining the optimal power level P for the next time interval and make an adjustment to the power amplifier 30 via power adjustment unit 36. The long term storage facility 28 supplies historical data to the power demand prediction unit 38 and receives short term digital power level data from the downlink demand measurement unit 42.

FIG. 5 is a block diagram of an alternative embodiment of the base station 15 showing a memory module 46 and software modules 66, 68 and 70 that, when executed by a processor, cause the processor to perform the functions of downlink demand measurement 66, power demand prediction 68 and power adjustment 70 as discussed above with respect to units 42, 38 and 36 of FIG. 4, respectively. Fig. 5 also shows a long term storage module 29 corresponding to the long term storage facility 28 of FIG. 4.

FIG. 6 is a flowchart of an exemplary process to dynamically control a power amplifier at a base station to achieve energy efficiency. The process includes calculating at least one digital power level in a first time interval based on at least one digital sample collected by the downlink demand measurement unit 42 over the first time interval (block S100). At least one digital power level in a second time interval is predicted by the PA power demand prediction unit 38 based on at least one calculated digital power level of the first time interval (block S102). Operation of the power amplifier 30 is adjusted by the PA power adjustment unit 36 based at least in part on the at least one predicted digital power level (block S104). The process then continues for next subsequent time intervals. It will be understood that the terms first and second refer to one interval followed by a next subsequent time interval so that the process continues for subsequent time intervals. In other words, the terms "first" and "second" are used solely for reference of one time interval to another and not with respect to a specific value.

FIG. 7 is a flowchart of an exemplary process for adaptive control of the PA power level. First, the base station is active and the transmission of downlink traffic is commenced (block S106). The process includes measuring DL digital power level demand by the downlink demand measurement unit 42 during a first time interval (block S108). In particular, the digital power level of a plurality of filter branches sharing the same PA is measured. The measured DL digital power level demand is archived to short term memory 46 and long term storage (block SI 10). The DL digital power level demand for a next time interval is predicted by the PA power demand prediction unit 38 (block SI 12). This involves training the model for predicting the DL digital power level using real time DL digital power level measurements collected during the first interval (as in block S108). The first interval may be a fixed number of samples, for example 60 samples collected every minute or every hour.

Historical data is retrieved from the long term storage facility 28 (block SI 14). The retrieved historical data may correspond to the next time interval for which power is to be predicted. For example, if the next time interval is 7pm to 9pm on Saturday, the historical data to be retrieved may be the data collected from 7pm to 9pm on one or more previous Saturdays. The predicted digital power level demand is combined with the historical data to predict power amplifier power demand for the next time interval by the PA power demand prediction unit 38 (block SI 16). IQ data modulation, measured KPI feedback and optionally, operator overwrite from operator overwrite switch 50 are parameters used to determine terms to be added to the calculation of the power amplifier setting (block SI 18) which is applied to the power amplifier (block SI 20). A goal of the setting of the power amplifier is to eliminate surplus power which would otherwise be expended. The base station continues to monitor KPI via the KPI feedback control unit 48 to determine if there is any system degradation that would require increased power (block S122). The process continues at block s 106.

Thus, the power demand prediction unit 38 uses the digital power levels of a current time interval to forecast an amount of anticipated DL power which may be combined with historical data and margins reserved for EVM (which margin depends on modulation order) and margins reserved to improve KPIs.

Some embodiments described herein include closed loop real time adjustment of PA power based on analysis of a base station's historic and environmental data and real time predicted data to set the power amplifier operation to conserve power, and improve DL performance when throughput demand exceeds nominal conditions. This is done while maintaining backward compatibility with pre-5G networks. As will be appreciated by one of skill in the art, the concepts described herein may be embodied as a method, data processing system, and/or computer program product. Accordingly, the concepts described herein may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects all generally referred to herein as a

"circuit" or "module." Furthermore, the disclosure may take the form of a computer program product on a tangible computer usable storage medium having computer program code embodied in the medium that can be executed by a computer. Any suitable tangible computer readable medium may be utilized including hard disks, CD-ROMs, electronic storage devices, optical storage devices, or magnetic storage devices.

Some embodiments are described herein with reference to flowchart illustrations and/or block diagrams of methods, systems and computer program products. It will be understood that each block of the flowchart illustrations and/or block diagrams, and combinations of blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer (to thereby provide a special purpose computer), special purpose computer, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions/acts specified in the flowchart and/or block diagram block or blocks.

These computer program instructions may also be stored in a computer readable memory or storage medium that can direct a computer or other

programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer readable memory produce an article of manufacture including instruction means which implement the function/act specified in the flowchart and/or block diagram block or blocks.

The computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions/acts specified in the flowchart and/or block diagram block or blocks.

It is to be understood that the functions/acts noted in the blocks may occur out of the order noted in the operational illustrations. For example, two blocks shown in succession may in fact be executed substantially concurrently or the blocks may sometimes be executed in the reverse order, depending upon the functionality/acts involved. Although some of the diagrams include arrows on communication paths to show a primary direction of communication, it is to be understood that

communication may occur in the opposite direction to the depicted arrows.

Computer program code for carrying out operations of the concepts described herein may be written in an object oriented programming language such as Java® or C++. However, the computer program code for carrying out operations of the disclosure may also be written in conventional procedural programming languages, such as the "C" programming language. The program code may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer. In the latter scenario, the remote computer may be connected to the user's computer through a local area network (LAN) or a wide area network (WAN), or the connection may be made to an external computer (for example, through the Internet using an Internet Service Provider).

Many different embodiments have been disclosed herein, in connection with the above description and the drawings. It will be understood that it would be unduly repetitious and obfuscating to literally describe and illustrate every combination and subcombination of these embodiments. Accordingly, all embodiments can be combined in any way and/or combination, and the present specification, including the drawings, shall be construed to constitute a complete written description of all combinations and subcombinations of the embodiments described herein, and of the manner and process of making and using them, and shall support claims to any such combination or subcombination.

It will be appreciated by persons skilled in the art that the embodiments described herein are not limited to what has been particularly shown and described herein above. In addition, unless mention was made above to the contrary, it should be noted that all of the accompanying drawings are not to scale. A variety of modifications and variations are possible in light of the above teachings without departing from the scope of the following claims.

Claims

What is claimed is:

1. A method for adaptively controlling a power amplifier (30) in a base station (15), the method comprising:

calculating at least one digital power level in a first time interval based on at least one digital sample of a signal collected over the first time interval (SI 00); predicting at least one digital power level in a second time interval subsequent to the first time interval based on the at least one calculated digital power level of the first time interval (SI 02); and

adjusting an operation of the power amplifier (30) of the base station (15) based at least in part on the at least one predicted digital power level (SI 04).

2. The method of Claim 1 , wherein predicting at least one digital power level (38) in the second time interval is further based in part on historical data.

3. The method of Claim 2, wherein the historical data is digital power levels measured during a particular period of time in the past, the particular period of time being one of a time period during a particular day, a number of days of a particular week and a number of weeks in a particular year.

4. The method of any of Claims 1-3, wherein the predicted at least one digital power level are computed recursively.

5. The method of Claim 4, further comprising:

computing an average of a plurality of predicted digital power levels; and wherein adjusting the operation of the power amplifier is further based at least in part on the average of the plurality of predicted digital power levels.

6. The method of any of Claims 2-5, further comprising:

computing an average of historical data of past recurring time intervals; and wherein predicting (38) at least one digital power level in the second time interval is further based at least in part on the average of historical data.

7. The method of any of Claims 2-6, wherein adjusting the operation of the power amplifier (30) is further based at least in part on a margin of improvement of an error vector magnitude, EVM.

8. The method of any of Claims 2-7, wherein predicting at least one digital power level (38) in the second time interval is further based at least in part on a margin of improvement of at least one key performance indicator.

9. The method of any of Claims 2-8, wherein adjusting the operation of the power amplifier (30) is further based at least in part on a downlink power level at nominal traffic conditions.

10. The method of any of Claims 2-9, wherein adjusting the operation of the power amplifier (30) is further based at least in part on a nominal power amplifier power level configured for nominal traffic conditions.

11. The method of Claim 1 , wherein adjusting the operation of the power amplifier (30) includes adjusting a bias parameter of the power amplifier to affect an amplification of the power amplifier.

12. The method of any of Claims 1-11, further comprising overriding the adjustment step in response to a user command.

13. A base station having a power amplifier (30), the base station (15) comprising:

processing circuitry (52) comprising:

a memory (54) configured to store digital power level data;

a processor (58) configured to:

calculate at least one digital power level in a first time interval based on at least one digital sample of a signal collected over the current time interval (S100); predict at least one digital power level in a second time interval subsequent to the first time interval based on the at least one calculated digital power level of the first time interval (SI 02); and

adjust an operation of the power amplifier (30) of the base station based at least in part on the at least one predicted digital power level (S104).

14. The base station of Claim 13, further comprising a link to a long term storage facility (28) to store historical data, and wherein predicting at least one digital power level in the second time interval is further based in part on historical data.

15. The base station of Claim 14, wherein the historical data is digital power levels measured during a particular period of time in the past, the particular period of time being one of a time period during a particular day, a number of days of a particular week and a number of weeks in a particular year.

16. The base station of any of Claims 13-15, wherein the predicted at least one digital power level are computed recursively.

17. The base station of Claim 16, wherein the processor (58) is further configured to:

compute an average of a plurality of predicted digital power levels; and wherein adjusting the operation of the power amplifier is further based at least in part on the average of the plurality of predicted digital power levels.

18. The base station of any of Claims 13-17, wherein the processor (58) is further configured to:

compute an average of historical data of past recurring time intervals; and wherein predicting at least one digital power level in the second time interval is further based at least in part on the average of historical data.

19. The base station of any of Claims 13-18, wherein the processor (58) is further configured to adjust the operation of the power amplifier (30) based at least in part on a margin of improvement of an error vector magnitude, EVM.

20. The base station of any of Claims 13-19, wherein the processor (58) is further configured to predict at least one digital power level in the second time interval based at least in part on a margin of improvement of at least one key performance indicator.

21. The base station of any of Claims 13-20, wherein the processor (58) is further configured to predict at least one digital power level in the second time interval based at least in part on a downlink power level at nominal traffic conditions.

22. The base station of any of Claims 13-21, wherein the processor (58) is further configured to adjust the operation of the power amplifier (30) based at least in part on a nominal power amplifier power level configured for nominal traffic conditions.

23. The base station of Claim 13, wherein adjusting the operation of the power amplifier (30) further includes adjusting a bias parameter of the power amplifier (30) to affect an amplification of the power amplifier (30).

24. The base station of any of Claims 13-23, wherein the processor (58) is further configured to receive an override command from an operator to override the adjusting by the processor (58).

25. A base station having a power amplifier, the base station (15) comprising:

a downlink demand measurement module (66) configured to calculate at least one digital power level in a first time interval based on at least one digital sample collected over the current time interval; a power demand prediction module (68) configured to predict at least one digital power level in a second time interval subsequent to the first time interval based on the at least one calculated digital power level of the first time interval; and

an adjustment module (70) configured to adjust an operation of the power amplifier of the base station based at least in part on the at least one predicted digital power level.

26. The base station of Claim 25, further comprising:

a link to a long term storage module (29) configured to store historical data relating to power consumption by the power amplifier (30).

27. The base station of any of Claims 25 and 26, further comprising: a short term storage module (46) configured to store the calculated at least one digital power level.

28. The base station of any of Claims 25-27, further comprising an operator overwrite module (50) to override the adjustment of the adjustment module (70).

29. The base station of any of Claims 25-28, further comprising a key performance indicator, KPI, module (48) configured to supply to the adjustment module (70) a predetermined margin of improvement of the KPI.