CROSS-REFERENCE TO RELATED APPLICATION

[0001] This application claims the benefit of priority to International Application No. PCT/US2022/11997, entitled “SYSTEM-ON-CHIP IMPLEMENTING DROOP COMPENSATION, APPARATUS, AND METHOD THEREOF,” filed on January 11, 2022, which is incorporated by reference herein in its entirety.

BACKGROUND

[0002] Embodiments of the present disclosure relate to apparatuses and methods for wireless communication.

[0003] Wireless communication systems are widely deployed to provide various telecommunication services such as telephony, video, data, messaging, and broadcasts. At a receiver of a wireless communication system, the circuits at the analog front end and/or the circuits at the digital front end may introduce droops, which degrades the signal quality. As a result, the baseband signal level cannot be maintained within an acceptable range for correct signal processing.

SUMMARY

[0004] Embodiments of an apparatus, a system, and a method, implementing droop calibration, are provided.

[0005] In one aspect, the present disclosure provides some embodiments of the apparatus implementing droop calibration. The apparatus may include a receiver associated with a radio frequency (RF) band, a processor, and memory coupled to the processor and storing instructions. When executed by the processor, the instructions may cause the processor to tune a frequency of the RF band of the receiver to a selected band. A transmitter may be configured to transmit signals in the selected band. Samples may be received in a time domain by the receiver based on the signals and the RF band. A plurality of frequency tones in a frequency domain may be obtained based on the samples in the time domain. Each of the plurality of frequency tones may correspond to a power level. A droop profile in a logarithmic (log) domain may be modeled based on a log conversion of the power levels of the frequency tones. A representation of an undrooping profile in the log domain may be obtained based on the droop profile in the log domain. The undrooping profile may be an inverse of the droop profile.

[0006] In another aspect, the present disclosure provides some embodiments of the system implementing droop calibration. The system may include a transmitter, a first apparatus, and a second apparatus. The first apparatus may include a receiver associated with a radio frequency (RF) band, a processor, and memory coupled to the processor and storing instructions. The instructions may cause the processor to tune a frequency of the RF band of the receiver to a selected band for estimating droop. A transmitter may be configured to transmit signals in the selected band. Samples may be received in a time domain based on the signals and the RF band. A plurality of frequency tones in a frequency domain may be obtained based on the samples in the time domain. Each of the plurality of frequency tones may correspond to a power level. A droop profile in a logarithmic (log) domain may be modeled based on a log conversion of the power levels of the frequency tones. A representation of an undrooping profile in the log domain may be obtained and stored based on the droop profile in the log domain. The undrooping profile may be an inverse of the droop profile. The second apparatus may be configured to retrieve the representation of the undrooping profile in the log domain. At least one of droop estimation or droop compensation may be performed based on the representation of the undrooping profile.

[0007] In still another aspect, the present disclosure provides some embodiments of the method implementing droop calibration. The method may include tuning a frequency of a radio frequency (RF) band of a receiver to a selected band. Signals in the selected band may be transmitted by a transmitter. Samples in a time domain may be received by the receiver based on the signals and the RF band. The samples in the time domain may be transformed into a plurality of frequency tones in a frequency domain. Each of the plurality of frequency tones may correspond to a power level. A droop profile in a logarithmic (log) domain may be modeled based on a log conversion of the power levels of the frequency tones. A representation of an undrooping profile in the log domain may be obtained based on the droop profile in the log domain. The undrooping profile may be an inverse of the droop profile.

BRIEF DESCRIPTION OF THE DRAWINGS

[0008] The accompanying drawings, which are incorporated herein and form a part of the specification, illustrate embodiments of the present disclosure and, together with the description, further serve to explain the principles of the present disclosure and to enable a person skilled in the pertinent art to make and use the present disclosure.

[0009] FIG. 1 illustrates an exemplary droop profile with respect to the frequency at a receiver.

[0010] FIG. 2 illustrates an exemplary wireless network, in which certain aspects of the present disclosure may be implemented, according to some embodiments of the present disclosure. [0011] FIG. 3 illustrates a block diagram of an exemplary node, according to some embodiments of the present disclosure.

[0012] FIG. 4 illustrates a block diagram of an exemplary system, including a radio frequency (RF) chip and a baseband chip, implementing the droop calibration, according to some embodiments of the present disclosure.

[0013] FIG. 5 illustrates a flow diagram of an exemplary method implementing droop calibration, according to some embodiments of the present disclosure.

[0014] Embodiments of the present disclosure will be described with reference to the accompanying drawings.

DETAILED DESCRIPTION

[0015] Although specific configurations and arrangements are discussed, it should be understood that this is done for illustrative purposes only. A person skilled in the pertinent art will recognize that other configurations and arrangements can be used without departing from the spirit and scope of the present disclosure. It will be apparent to a person skilled in the pertinent art that the present disclosure can also be employed in a variety of other applications.

[0016] It is noted that references in the specification to “one embodiment,” “an embodiment,” “an example embodiment,” “some embodiments,” “certain embodiments,” etc., indicate that the embodiment described may include a particular feature, structure, or characteristic, but every embodiment may not necessarily include the particular feature, structure, or characteristic. Moreover, such phrases do not necessarily refer to the same embodiment. Further, when a particular feature, structure, or characteristic is described in connection with an embodiment, it would be within the knowledge of a person skilled in the pertinent art to effect such feature, structure, or characteristic in connection with other embodiments whether or not explicitly described.

[0017] In general, terminology may be understood at least in part from usage in context. For example, the term “one or more” as used herein, depending at least in part upon context, may be used to describe any feature, structure, or characteristic in a singular sense or may be used to describe combinations of features, structures, or characteristics in a plural sense. Similarly, terms, such as “a,” “an,” or “the,” again, may be understood to convey a singular usage or to convey a plural usage, depending at least in part upon context. In addition, the term “based on” may be understood as not necessarily intended to convey an exclusive set of factors and may, instead, allow for the existence of additional factors not necessarily expressly described, again, depending at least in part on context.

[0018] Various aspects of wireless communication systems will now be described with reference to various apparatuses, systems, and methods. These apparatuses and methods will be described in the following detailed description and illustrated in the accompanying drawings by various blocks, modules, units, components, circuits, steps, operations, processes, algorithms, etc. (collectively referred to as “elements”). These elements may be implemented using electronic hardware, firmware, computer software, or any combination thereof. Whether such elements are implemented as hardware, firmware, or software depends upon the particular application and design constraints imposed on the overall system.

[0019] The techniques described herein may be used for various wireless communication networks, such as code division multiple access (CDMA) system, time division multiple access (TDMA) system, frequency division multiple access (FDMA) system, orthogonal frequency division multiple access (OFDMA) system, single-carrier frequency division multiple access (SC- FDMA) system, and other networks. The terms “network” and “system” are often used interchangeably. A CDMA network may implement a radio access technology (RAT), such as Universal Terrestrial Radio Access (UTRA), evolved UTRA (E-UTRA), CDMA 2000, etc. A TDMA network may implement a RAT, such as the Global System for Mobile Communications (GSM). An OFDMA network may implement a RAT, such as Long-Term Evolution (LTE) or New Radio (NR). The techniques described herein may be used for the wireless networks and RATs mentioned above, as well as other wireless networks and RATs.

[0020] In a wireless communication system, data is transmitted in a data channel between two nodes of the network, such as between a base station and a user equipment (UE). Downlink is part of the data transmission, sending the data out or downwards from a higher level or a part of the network to the UE. In the downlink, the antenna at the UE may transmit received RF signals to the RF chip where certain functions may be performed, such as filtering, down conversation, sample-rate conversions of the RF signals, etc. Subsequently, the received RF signals are transformed into suitable low-frequency digital signals (i.e., baseband signals) that can be processed by a baseband chip. Accordingly, the RF chip may include RF filters (e.g., analog antialiasing filters), RF amplifiers, local oscillators, mixers, re-sampler, decimator, other RF circuits, and/or digital circuits for various signal processing functions. These components at the RF chip, however, may introduce drooping of various levels, for which the frequency response does not show a wide, flat pass band as expected.

[0021] In some scenarios, drooping may be viewed as part of the channel and thus may be incorporated into the channel effect without any compensation and/or demodulation. In other cases, however, in order to estimate the signal strength or quality with fewer errors, it may become necessary to estimate and compensate the drooping so that the receiver can have more accurate information about the received signal properties.

[0022] FIG. 1 illustrates an exemplary droop profile 100 with respect to the frequency at a receiver. The droop profile 100 in FIG. 1 is plotted using a power level as the vertical axis in response to a given frequency at the horizontal axis. As shown in FIG. 1, the power gradually decreases, away from an RF center frequency towards its two tails (or the band edges). To some extent, the edge frequency response is less than the center frequency response. In an ideal situation, a wider and flatter profile may be expected over a receiver signal band. That is, without drooping, an amount of the power at measurement may be the same (or at least similar if measurement errors are being considered) over the same signal bandwidth. Due to the impact of drooping, however, power loss over the frequency may be observed. In addition, the power profile may not be symmetric about the center RF frequency.

[0023] The undesired decrease of the power is referred to as a drooping issue, which is generally inevitable in a receiver (RX) system with RF circuits and digital front end (DFE) components. As described above, with the drooping issue, the edge tones of the power profile may experience more attenuation compared to the center tones thereof. It results that the signals fed to the baseband chip cannot be maintained within the desired power range. That is, the baseband signal processing may be distorted. Consequently, the subsequent operations, such as channel estimation, cell search, received signal strength indicator (RS SI), or reference signal received power (RSRP) measurement, and/or demodulation, cannot be performed properly as desired. The signal strength and quality estimation may be necessary for some scenarios. For example, the RSRP/RSSI measurement may be an essential index for a system to determine an optimal cell to which the communication can be switched to ensure high communication quality. In other words, the droop estimation and compensation may need to be done eventually.

[0024] To overcome the droop issues, certain technical approaches proposed employing some external instrument to generate calibration waveforms so as to explore and characterize how a receiver in a wireless communication system responds to signals. The receiver may be configured to receive the calibration waveforms generated by the external instrument, and based on the received signals, the receiver can be calibrated concerning the drooping. The external instrument can be expensive. Further, the external instrument may generate and transmit only pure tones over different frequencies, each tone with a frequency spectrum limited to a narrow band. As drooping is a band and frequency-dependent issue, a bandwidth of interest may need to be exhaustively scanned in order to construct a complete droop profile. In view of the time-consuming procedures, these approaches may be limited to exploring and characterizing only certain important parts in the wireless communication system. Subsequently, an average of the calibrated results on the important parts may be obtained and applied to all parts, thereby simplifying the procedures. For that reason, these approaches may not produce satisfactory results.

[0025] In view of these and other drawbacks in these approaches, some embodiments of the present disclosure provide inventive and efficient methods for droop calibration that may be implemented to various wireless communication systems. The proposed system, according to some embodiments of the present disclosure, is a self-calibrated system in regard to droop without the usage of the external calibration instrument. In some embodiments, through compressing an undrooping profile to a log domain, memory storage can be significantly reduced. Subsequently, in expanding log-domain undrooping gains, the implementation can be significantly simplified and can provide support for both single carrier scenarios and carrier aggregation (CA) scenarios. Meanwhile, the complexity of the DFE control can also be reduced.

[0026] Moreover, in some embodiments, a scheme of various dynamic ranges for coefficients of the undrooping profile is provided. It can enhance the accuracy of the undrooping profile and also enables simple linear interpolation, e.g., through hardware implementation, for obtaining the undrooping gains at a relatively low cost of power, timing, and implementation complexity. In some embodiments, when the droop over a predetermined frequency band is relatively smooth, the linear interpolation may be suited. Therefore, the overall cost may be significantly reduced. In addition, a lookup table (LUT) for locating the undrooping gains is no longer needed. [0027] Reference will now be made in detail to exemplary embodiments of the present disclosure in the following, which may be illustrated in the accompanying drawings. It is apparent that the embodiments are some but not all of the embodiments of the present disclosure, and the drawings are used for illustration but not for limitation.

[0028] FIG. 2 illustrates an exemplary wireless network 200, in which certain aspects of the present disclosure may be implemented, according to some embodiments of the present disclosure. As shown in FIG. 2, wireless network 200 may include a network of nodes, such as a user equipment (UE) 202, an access node 204, and a core network element 206. UE 202 may be any terminal device, such as a mobile phone, a desktop computer, a laptop computer, a tablet, a vehicle computer, a gaming console, a printer, a positioning device, a wearable electronic device, a smart sensor, or any other device capable of receiving, processing, and transmitting information, such as any member of a vehicle to everything (V2X) network, a cluster network, a smart grid node, or an Internet-of-Things (loT) node. It is understood that UE 202 is illustrated as a mobile phone simply by way of illustration and not by way of limitation.

[0029] Access node 204 may be a device that communicates with UE 202, such as a wireless access point, a base station (BS), a Node B, an enhanced Node B (eNodeB or eNB), a next-generation NodeB (gNodeB or gNB), a cluster master node, or the like. Access node 204 may have a wired connection to UE 202, a wireless connection to UE 202, or any combination thereof. Access node 204 may be connected to UE 202 by multiple connections, and UE 202 may be connected to other access nodes in addition to access node 204. Access node 204 may also be connected to other user equipments. It is understood that access node 204 is illustrated by a radio tower by way of illustration and not by way of limitation.

[0030] Core network element 206 may serve access node 204 and UE 202 to provide core network services. Examples of core network element 206 may include a home subscriber server (HSS), a mobility management entity (MME), a serving gateway (SGW), or a packet data network gateway (PGW). These are examples of core network elements of an evolved packet core (EPC) system, which is a core network for the LTE system. Other core network elements may be used in LTE and in other communication systems. In some embodiments, core network element 206 includes an access and mobility management function (AMF) device, a session management function (SMF) device, or a user plane function (UPF) device, of a core network for the NR system. It is understood that core network element 206 is shown as a set of rack-mounted servers by way of illustration and not by way of limitation. [0031] Core network element 206 may connect with a large network, such as Internet 208, or another Internet Protocol (IP) network, to communicate packet data over any distance. In this way, data from UE 202 may be communicated to other user equipments connected to other access points, including, for example, a computer 210 connected to Internet 208, for example, using a wired connection or a wireless connection, or to a tablet 212 wirelessly connected to Internet 208 via a router 214. Thus, computer 210 and tablet 212 provide additional examples of possible user equipments, and router 214 provides an example of another possible access node.

[0032] A generic example of a rack-mounted server is provided as an illustration of core network element 206. However, there may be multiple elements in the core network including database servers, such as a database 216, and security and authentication servers, such as an authentication server 218. Database 216 may, for example, manage data related to user subscription to network services. A home location register (HLR) is an example of a standardized database of subscriber information for a cellular network. Likewise, authentication server 218 may handle the authentication of users, sessions, and so on. In the NR system, an authentication server function (AUSF) device may be the specific entity to perform user equipment authentication. In some embodiments, a single server rack may handle multiple such functions, such that the connections between core network element 206, authentication server 218, and database 216, may be local connections within a single rack.

[0033] Each element in FIG. 2 may be considered a node of wireless network 200. More detail regarding the possible implementation of a node is provided by way of example in the description of a node 300 in FIG. 3. Node 300 may be configured as UE 202, access node 204, or core network element 206 in FIG. 2. Similarly, node 300 may also be configured as computer 210, router 214, tablet 212, database 216, or authentication server 218 in FIG. 2. As shown in FIG. 3, node 300 may include a processor 302, a memory 304, and a transceiver 306. These components are shown as connected to one another by a bus, but other connection types are also permitted. When node 300 is UE 202, additional components may also be included, such as a user interface (UI), sensors, and the like. Similarly, node 300 may be implemented as a blade in a server system when node 300 is configured as core network element 206. Other implementations are also possible.

[0034] Transceiver 306 may include any suitable device for sending and/or receiving data. Node 300 may include one or more transceivers, although only one transceiver 306 is shown for simplicity of illustration. An antenna 308 is shown as a possible communication mechanism for node 300. Multiple antennas and/or arrays of antennas may be utilized. Additionally, examples of node 300 may communicate using wired techniques rather than (or in addition to) wireless techniques. For example, access node 204 may communicate wirelessly to UE 202 and may communicate by a wired connection (for example, by optical or coaxial cable) to core network element 206. Other communication hardware, such as a network interface card (NIC), may be included as well.

[0035] As shown in FIG. 3, node 300 may include processor 302. Although only one processor is shown, it is understood that multiple processors can be included. Processor 302 may include microprocessors, microcontroller units (MCUs), digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), programmable logic devices (PLDs), state machines, gated logic, discrete hardware circuits, and other suitable hardware configured to perform the various functions described throughout the present disclosure. Processor 302 may be a hardware structure having one or more processing cores. Processor 302 may execute software. Software shall be construed broadly to mean instructions, instruction sets, code, code segments, program code, programs, subprograms, software modules, applications, software applications, software packages, routines, subroutines, objects, executables, threads of execution, procedures, functions, etc., whether referred to as software, firmware, middleware, microcode, hardware description language, or otherwise. Software can include computer instructions written in an interpreted language, a compiled language, or machine code. Other techniques for instructing hardware are also permitted under the broad category of software.

[0036] As shown in FIG. 3, node 300 may also include memory 304. Although only one memory is shown, it is understood that multiple memories can be included. Memory 304 can broadly include both memory and storage. For example, memory 304 may include random-access memory (RAM), read-only memory (ROM), static RAM (SRAM), dynamic RAM (DRAM), ferroelectric RAM (FRAM), electrically erasable programmable ROM (EEPROM), CD-ROM or other optical disk storage, hard disk drive (HDD), such as magnetic disk storage or other magnetic storage devices, Flash drive, solid-state drive (SSD), or any other medium that can be used to carry or store desired program code in the form of instructions that can be accessed and executed by processor 302. Broadly, memory 304 may be embodied by any computer-readable medium, such as a non-transitory computer-readable medium.

[0037] Processor 302, memory 304, and transceiver 306 may be implemented in various forms in node 300 for performing wireless communication functions. In some embodiments, processor 302, memory 304, and transceiver 306 of node 300 are implemented (e.g., integrated) on one or more SoCs. In one example, processor 302 and memory 304 may be integrated on an application processor (AP) SoC (sometimes known as a “host,” referred to herein as a “host chip”) that handles application processing in an operating system environment, including generating raw data to be transmitted. In another example, processor 302 and memory 304 may be integrated on a baseband processor (BP) SoC (sometimes known as a “modem,” referred to herein as a “baseband chip”) that converts the raw data, e.g., from the host chip, to signals that can be used to modulate the carrier frequency for transmission, and vice versa, which can run a real-time operating system (RTOS). In still another example, processor 302 and transceiver 306 (and memory 304 in some cases) may be integrated on an RF SoC (sometimes known as a “transceiver,” referred to herein as an “RF chip”) that transmits and receives RF signals with antenna 308. It is understood that in some examples, some or all of the host chip, baseband chip, and RF chip may be integrated as a single SoC. For example, a baseband chip and an RF chip may be integrated into a single SoC that manages all the radio functions for cellular communication.

[0038] Any suitable node of wireless network 200, as shown in FIG. 2, which receives signals to another node (e.g., from access node 204 to UE 202 via downlink) may implement the schemes of droop calibration as described below with reference to FIGs. 4 and 5. Accordingly, compared with the known solutions in other approaches, a wireless communication system with better performance, less memory storage, simpler hardware structures, and more flexibility can be obtained.

[0039] FIG. 4 illustrates a block diagram of an exemplary system 400, including an RF chip 402 and a baseband chip 404, implementing the droop calibration, according to some embodiments of the present disclosure. Consistent with the spirit of the present disclosure, RF chip 402 may include a plurality of functional modules or units collectively configured to perform the offline droop calibration according to some embodiments. These implementations for the droop calibration may be transparent to an upper layer or base station, so the interaction between layers can be established based on the implementations. In some embodiments, at least a portion of the function modules or units, in part performing the droop calibration, may be implemented by processor 302 and memory 304, as shown in FIG. 3. Processor 302 may be a hardware device having one or more processing cores, or processor 302 may execute software stored in memory 304 on the hardware device to control the hardware device, the choice between which may depend upon the particular application and design constraints imposed on the overall system. For example, RF chip 402 may include an RF processor executing instructions stored in a local memory 4026. The RF processor may be a generic processor, such as a central processing unit, not dedicated to the droop calibration. That is, the RF processor is also responsible for any other functions of RF chip 402 and can be interrupted during the droop calibration being performed due to another process with higher priority. Some or all of the functional modules or units in RF chip 402 may be implemented as software modules executed by the RF processor to perform the respective functions described below in regard to FIGs. 4 and 5. In some embodiments, at least a portion of the modules or units may be implemented as an electronic circuit, such as ASIC, dedicated to implementing certain functions in regard to the droop calibration.

[0040] System 400 may be an example for any suitable node, such as UE 202, in wireless network 200 in FIG. 2. As shown in FIG. 4, system 400 may include an RF chip 402, a baseband chip 404, and an antenna 406. It is apparent that system 400 may further include other chips or modules configured for certain other functions, e.g., a host chip, which is not shown for simplicity of illustration. RF chip 402, baseband chip 404, and the other chips may be connected or coupled with an external memory 408. RF chip 402 and baseband chip 404 are shown in FIG. 4 as two discrete chips, for example, whereas in other instances, RF chip 402 and baseband chip 404 can be integrated into a single SoC. Further, as described above with reference to FIG. 3, RF chip 402 may be embodied by processor 302, memory 304, and transceiver 306, while baseband chip 404 may be embodied by processor 302 and memory 304.

[0041] As shown in FIG. 4, RF chip 402 may further include a receiver (RX), an RF front end (RF FE), an analog-to-digital converter (ADC), and a digital front-end (DFE) in its downlink. The RX may be configured to receive RF signals, through an antenna 406, from another node on the network. The RF FE may include any suitable elements between antenna 406 and ADC that are configured to process the RF signals of an analog form. In some instances, RF FE may include, e.g., RF filters, RF amplifiers, local oscillators, and mixers. ADC atRF chip 402 may be configured to convert a stream of symbols (i.e., samples, e.g., orthogonal frequency division multiplexing, OFDM, symbols) in the analog form (e.g., radio signals) to a digital form (e.g., digital signals). DFE may include any suitable elements that are configured to process the digital signals converted by ADC, such as filtering or sample-rate conversion. It can be understood that RF chip 402 in FIG. 4 merely illustrates some elements for ease of illustration but not for limitation. In addition to those elements shown in FIG. 4, RF chip 402 may include any other suitable elements, such as elements in its uplink.

[0042] FIG. 5 illustrates a flow diagram of an exemplary method 500 implementing droop calibration, according to some embodiments of the present disclosure. In the following, some embodiments of the present disclosure will be described with respect to FIGs. 4 and 5.

[0043] Some embodiments of the present disclosure may provide an efficient method implementing droop calibration, in which a frequency of RF chip 402 may be tuned based on a selected band. The selected band may be used for estimating droop under a plan. The method may proceed to 502 in FIG. 5. RF chip 402 may be associated with an RF band within which signals can be received. RF chip 402 in FIG. 4 may include an RF tuning module 4022. According to some embodiments, RF tuning module 4022 may be configured to adjust the frequency of RF chip 402 according to a chosen RF band to match with the selected band. The chosen RF band for RF chip 402 may include a band of interest for estimating droop. In some embodiments, the chosen RF band may include a time division duplex (TDD) band.

[0044] TDD is a spectrum-usage technique for emulating full-duplex communication over a half-duplex communication link. In TDD, both uplink and downlink use the same spectrum frequency at various times. More specifically, TDD applies the same frequency band by assigning alternating time slots for transmitting and receiving operations. As a consequence, TDD is more spectrum friendly, implying that the spectrum usage is increased. In TDD, frequencies can be tuned for operations anywhere in a band. Because a single frequency is assigned to a user for both directions, diplexers are not required in TDD. That is, the cost can be reduced for the transmitter, and the receiver uses the same components. In view of the above, some embodiments of the present disclosure may use a TDD band for tuning the frequency of RF chip 402 at 502 and thus can take advantage of it. Despite so, considering that drooping is an analog baseband response, that is, the droop calibration schemes according to some embodiments can apply equally well with an FDD band.

[0045] In some embodiments, RF tuning module 4022 may be further configured to perform a fine-tuning to enhance the performance of the droop calibration. In one example of finetuning, the chosen RF band of RF chip 402 may include relatively flat frequency responses around a center frequency, and a bandwidth of the chosen RF band may be larger than or equal to a selected bandwidth. The selected bandwidth may be a bandwidth of interest for estimating droop under a plan (or termed “droop bandwidth” in the following). For instance, a system may transmit signals having 50 MHz bandwidth or a larger bandwidth (60 MHz) to measure droop response of 40 MHz bandwidth. The term “relatively” may be used to describe that frequency responses around a central frequency may be flat, with a substantially zero slope, compared to frequencies of other tones away from the center frequency.

[0046] In certain examples, frequency responses having wider bandwidth and a more even profile around its center frequency may be obtained so as to serve as ideal reference power. Depending on the design of the RF chip and/or the application constraints, however, these requirements may be optional. For example, through a curve-fitting with a large number of frequency samples and by exploiting an envelope according to some embodiments of the present disclosure, uneven frequency responses and/or rippling around a center frequency can be mitigated. That is, even in the absence of flat frequency responses around a center frequency, the proposed schemes can still perform fairly well with the benefits of the smoothing techniques in the frequency domain.

[0047] RF tuning module 4022 may include one more functional units to perform the frequency tuning operations. For example, RF tuning module 4022 may include at least one of an antenna tuning circuit configured to tune antenna 406, one or more band pass filters arranged at the RF FE and configured to reject unwanted frequencies, or one or more tuned amplifiers configured to select desired frequency range. In FIG. 4, RF tuning module 4022 is depicted as a standalone module. However, it can be understood that RF tuning module 4022 can also be integrated into one RF analog component of RF chip 402 (e.g., the RF FE) so that the tuning operations can be performed by an existing element of RF chip 402, to which the present disclosure does not place limitations.

[0048] Turning to FIG. 5, the method may proceed to 504, where signals covering the selected bandwidth in the selected band may be transmitted and subsequently received by RF chip 402 through antenna 406. In some embodiments, system 400 may include a transmitter configured to generate and transmit the signals. In one instance, the transmitter may include a signal generator 410 configured for the purposes. In order to cover the frequencies of interest, the signals may include wideband signals over the droop bandwidth so that droop can be properly calibrated, estimated, and compensated. As described above, the chosen RF band of RF chip 402 may include relatively flat frequency responses around a center frequency of the RF band, and the bandwidth of the chosen RF band may be at least equal to the droop bandwidth.

[0049] In one instance, signal generator 410 may be configured to generate and transmit white-noise-like signals that are random signals with equal intensity at different frequencies in a band, and their responses with droop can be measured at the receiver side. The frequency responses can serve to estimate reference power. The reference power may be used later for actual droop estimation. In other instances, other waveforms readily generated at the transmitter may also be used for the droop calibration, e.g., the pseudorandom binary sequence (PRBS) signals for cell search or the orthogonal frequency division multiplexing (OFDM) symbols. The PRBS signals have white-noise-like properties and are similar to those used for 5G NR narrowband cell search. [0050] In some instances, the transmitted signals may have a wider bandwidth than the droop bandwidth so that droop near the edge tones of the band (as shown in FIG. 1) can still be estimated. The droop near the edge tones may provide information in the guard band if a large frequency offset that rotates the signal out of the band is present. For that purpose, the OFDM symbols with a wider band may be superior to the PRBS signals for the reason that the PRBS signals are similar to those used in cell search/measurement over a narrow band. As a result, the edge frequency tones may not be covered. In some instances, the transmitter may generate the OFDM symbols at various bandwidths for 5G NR, for example, 5 MHz, 10 MHz, 15 MHz, 20 MHz, 25 MHz, 30 MHz, 40 MHz, 50 MHz, 60 MHz, 70 MHz, 80 MHz, 90 MHz, and 100 MHz. In some instances, in a 4G LTE network, the symbols at 1 ,4MHz, 3MHz, 5 MHz, 10 MHz, 15MHz, and 20MHz bandwidth may be transmitted. It can be understood that the guard bands can be different for the same bandwidth of symbols in a 4G LTE band or a 5G NR band. In addition, for carrier aggregation scenarios, the carrier bandwidth may be not limited to the above-mentioned bandwidths as they are for the single carrier transmission. The bandwidth for carrier aggregation scenarios, for example, can be 200MHz if two carriers are intra-band contiguous where each carrier has 100MHz bandwidth. In other embodiments, the above exemplary 200MHz channel bandwidth may cover intra-band non-contiguous carrier aggregation scenarios limited to the maximum bandwidth of 200MHz, for example, with two carriers with bandwidths of 100MHz and 40MHz respectively, or three carriers with bandwidths of 30MHz, 40MHz, and 50MHz respectively, four carriers with bandwidths of 20MHz, 20MHz, 40MHz, and 40MHz respectively, or the like.

[0051] Under some circumstances, the transmitter may have large power leakage, which may contaminate the received signals at RF chip 402. Therefore, in some embodiments, a transmitter power level of the transmitter may be tuned properly so as to ensure that the power leakage to a receiver path is at a reasonable level. As a consequence, interference from the transmitter to the receiver side may be relatively small and thus can be ignored such that the received signals will be suited to the droop calibration. [0052] The method may proceed to 506 in FIG. 5, where the receiver path may be turned on at RF chip 402 to receive samples from the transmitter. The received samples may be transmitted through an RF and DFE chain of RF chip 402 (e.g., including the RF RE, ADC, DFE, etc.). The received samples may be taken in a time-domain at the receiver side. As shown in FIG. 4, RF chip 402 may further include a sample collector 4024 and local memory 4026. At 506 in FIG. 5, the received samples may be captured at DFE output by sample collector 4024. Further, at 508, the received samples may be stored in local memory 4026. In one instance, each sample may be stored with its associated timing information at a given sampling rate, so that transformation to a frequency domain can be accurately performed based on the timing information.

[0053] Local memory 4026 may include on-chip memory of RF chip 402, such as a buffer, a cache, or non-volatile memory. In some instances, local memory 4026 may include fast, low- latency static RAM (SRAM). In one embodiment, the size of local memory 4026 may be determined at least by the size of the required samples. In some embodiments, local memory 4026 may be identical to memory that stores instructions to enable processor 302 to perform certain operations in regard to the droop calibration at RF chip 402. In other embodiments, however, RF chip 402 may include multiple on-chip memory, and local memory 4026 may be different from memory stored the instructions for configuring processor 302.

[0054] Turning back to FIG. 5, at 510, it may be determined whether the signal transmission of signal generator 410 and the sample collection of sample collector 4024 are required to be continued. In some embodiments, a threshold that depends on at least the estimation accuracy may be preset. In some embodiments, the threshold may be determined by a sample size configured for covering a selected bandwidth for estimating droop. In some embodiments, the threshold may be determined based on both the estimation accuracy and the sample size covering the selected bandwidth.

[0055] The threshold may be determined, based on the estimation accuracy, in terms of the size of required samples. In one instance, a counter may be preset and accumulated when the samples are buffered in local memory 4026. The counter may be compared to the threshold to determine whether the signal transmission and the sample collection are required to be continued. In response to the counter being greater than the threshold, the signal transmission and the sample collection may be terminated, and the method may go to the next step. Otherwise, another iteration of the data transmission and collection may be continued. Through this manner, even the size of the samples that local memory 4026 can buffer is smaller than the threshold, based on the counter, some of the buffered samples that have been used may be flushed to empty the space of local memory 4026 for buffering new samples. Accordingly, the size of samples used in the estimation can be increased so as to meet the required estimation accuracy.

[0056] The threshold may be predetermined based on a sample size that can cover the whole droop bandwidth. The threshold can be channel-dependent, that is, technology-dependent. As described above, if the transmitted signals are white-noise signals or white-noise-like signals (e.g., the OFDM symbols) over the droop bandwidth, it can be assured that the received samples will capture the whole droop profile covering the selected band of interest.

[0057] The method may proceed to 512 in FIG. 5, where the fast Fourier transform (FFT) may be performed on the samples by an FFT module 4028, at a sampling rate much larger than the selected bandwidth. The samples may be fetched from local memory 4026. The operation of the FFT is known for being fast and being able to be implemented by hardware components. In principle, signals can be described in either a time domain or a frequency domain. Correspondingly, the droop can also be estimated in a time domain or a frequency domain. In the OFDM designs, considering that most of the baseband processing is in the frequency domain, droop estimation and compensation in a frequency domain may be beneficial. That means, additional time-domain processing is not required, and thus various sorts of filters may not be required to handle different scenarios to achieve good performance in the frequency response. Each filter may require tens of coefficients to be implemented. The implementation of the filters may increase the cost of the hardware implementation. In view of the above, some embodiments of the present disclosure may employ the FFT transformation to convert the time-domain samples to the frequency domain, although the present disclosure does not limit thereto. In some embodiments, the FFT module 4028 may be implemented through hardware, such as an FFT circuit.

[0058] Further, it is also known that the Nyquist-Shannon sampling theorem establishes a sufficient condition for a sampling rate that permits a discrete sequence of samples to capture all the information from a continuous-time signal of finite bandwidth. Therefore, based on the Nyquist-Shannon sampling theorem, the received samples in the time-domain, covering the band, can be translated to the frequency domain by the FFT with an oversampling rate to capture discrete power properties (for each tone) associated with the samples.

[0059] Returning back to FIG. 4, RF chip 402 may include a power and droop measuring module(s) 4030 and a droop and undrooping modeling module(s) 4032. At the output of the FFT module 4028, the time-domain samples may be translated into the frequency-domain samples, each corresponding to one tone. Based on the frequency-domain samples, at 514 in FIG. 5, a level of power corresponding to each tone in the band may be measured, and further at 516, power decreasing, i.e., the droop, can be obtained. These operations may be performed by power and droop measuring module(s) 4030 in FIG. 4. Power and droop measuring module(s) 4030 is illustrated as a single module configured to perform the power measurement and the droop measurement in serial operations. In some instances, the power measuring module and droop measuring module may be two separate modules, one configured to perform the power measurement and the other configured to perform the droop measurement, respectively.

[0060] Based on the droop corresponding to each tone, a droop profile in the frequency domain, covering the tones in the band, can be constructed. As to undrooping, a process of drooping compensation is a reverse process to cancel out droop impact; therefore, an undrooping profile can be achieved based on the droop profile. In the present disclosure, the terms “droop profile” and “droop model” may be used interchangeably. Likewise, the terms “drooping profile” and “drooping model” may be used interchangeably.

[0061] At 514, in some embodiments, for each tone over frequency (or termed “each frequency tone”) for estimating the droop, FFT outputs with multiple sample sets may be obtained, and average power for the tone may be calculated based on the FFT outputs so as to have a more precise droop measurement. The term “sample set” may be used to refer to a set of received samples used to estimate the droop on the frequency domain individually. Through this manner, an impact of random disturbance (such as instantaneous disturbance, thermal noise, fluctuation of voltage, etc.) may be significantly reduced, and therefore, the droop profile estimation can be improved at 516 and 518 in the method.

[0062] It can be observed that saving an undrooping profile based on frequency tones may demand a huge amount of memory when the selected bandwidth is large. Therefore, in order to save storage space, a formula or a mathematical model to represent the undrooping profile may be exploited. For example, this may be done by a curve-fitting using the Least-Mean-Square (LMS) method based on coefficients of a selected model. In some embodiments, a predetermined threshold may be obtained. For optimizing the system performance, a distortion after the droop compensation may be smaller than the threshold.

[0063] Consistent with the spirit of the present disclosure, a scheme for reducing a dimension of coefficients for representing an undrooping model is provided, in which the undrooping profile may be represented in a log domain to help compress a dynamic range of the coefficients and simplify the system control logic. The method may proceed to 518 in FIG. 5. The droop for each frequency tone may be first converted to a logarithmic domain (or “log domain” in short) and the droop in the log domain for each tone may be modeled using, e.g., the LMS method, to obtain the droop profile. That is, the droop profile over the frequency tones may be converted into the log domain to obtain the droop profile in the log domain. The droop for a frequency tone may include or be associated with a power level corresponding to that frequency tone. In some embodiments, a power level for each tone in the frequency domain may be taken by a log function, i.e., log (), which may translate the power level from an original domain associated with the raw data (or termed “original frequency domain” in short) to the log domain to obtain the logarithmic conversion of the power level for each tone. In other words, the log domain may refer to a domain constructed by a logarithmic function over frequency. As a result, a power profile mapped and projected to the log domain can be obtained, and memory space for storing the coefficients for representing the undrooping model can be reduced.

[0064] It is known that the log function, logb(z), has a property, in which when z=x*y, log _b (z) = log _b (x * y) = log _b (x) + log _b (y) , where x and y are positive numbers, and represents a multiplication operation. Based on the property, by using a log function, a multiplication operation can be translated into an addition operation. As a result, the range of the droop profile may be substantially compressed. On the other hand, the inverse function of log _b (x) = t is an exponential function b ¹ , which can be used for subsequent expansion/decompression to calculate the droop compensation gains back to its original frequency domain. In other words, by taking advantage of the log conversion, the undrooping profile is an inverse of the droop profile. As a consequence, fewer coefficients can also be used to model the undrooping profile. That is, memory space for storing the undrooping model can be reduced.

[0065] The log function may have a base of any positive number, e.g., a Euler’s number e (~ 2.71828...) or a prime integer (e.g., 2, 3, 5, ...), but not limited thereto. In some embodiments, the log function may use 2 as the base, i.e., a binary logarithm log ₂ (x'). In view of Institute of Electrical and Electronics Engineers (IEEE) floating-point standard and implementation, the binary logarithm can be related to a binary numeral system and can be easily implemented.

[0066] By the application of the log function, signals in the original frequency domain can be compressed into a smaller range. For example, signals with the range [1,4] in the original frequency domain can turn into the range [0,2] in the binary logarithm domain, where the size of the range is reduced from 3 (=4-1) to 2 (=2-0). The smaller dynamic range, under the same size of storage, can provide more solutions for the number of the same bit width per solution. When the ranges [1,4] and [0,2] of the same size of storage are being quantized, respectively, it can be expected that the smaller range [0,2] has a smaller error. In other words, with the same amount of bit width, the droop model can be more accurate but simpler. On the other hand, the undrooping profile, as an inverse of the droop profile, may be smoother in the log domain due to a decrease in the slope. As a result, the modeling, even over a large bandwidth, does not need segmented processing as those performed in the original frequency domain. That is, only one model is required if the droop profile is symmetric to the center tone. Even for an asymmetric droop profile, only two droop models may be in need. These properties can simplify the processing of undrooping gain calculation and storage, significantly simplify the control (e.g., the DEF control), and reduce the online implementation with less memory and computation.

[0067] Now turning back to FIG. 4, as described, RF chip 402 may include droop and undrooping modeling module(s) 4032, and the droop profile may be modeled through droop and undrooping modeling module(s) 4032. This may be offline modeling in the calibration process. Droop and undrooping modeling module(s) 4032 are illustrated in FIG. 4 as a single module configured to perform the droop modeling and the undrooping modeling in serial operations. In some instances, the droop modeling module and undrooping modeling module may be two separate modules, one configured to perform the droop modeling, and the other configured to perform the undrooping modeling.

[0068] In some embodiments, several candidate functions may be applied to approximate the undrooping profile over a closed interval. A polynomial function can be an option from those for the approximation. According to the Weierstrass approximation theorem, every continuous profile defined on a closed interval among [a, b] can be uniformly approximated as closely, depending on the highest order, as desired by a polynomial function. One reason to apply the theorem is based on the consideration that the polynomial function is one of the simplest functions. Further, the computer system may evaluate the polynomial function directly and easily by an iteration approach, and thus polynomial functions are easy to be implemented with the hardware structures.

[0069] Although the polynomial is taken as an example for the description, the present disclosure does not place limitations thereto. In some cases, other functions may also be applied. For example, a rational function having the format R(t) = P(t)/Q(t) may be used for the approximation, where P(t) and Q(t) are polynomials. P(t) and Q(t) herein may have different orders. [0070] In some embodiments, the coefficients of the desired droop profile may be achieved with various methods, such as the LMS method. In the method, a weight may be adjusted onto one frequency tone if an approximation error on the tone is assumed relatively large and intended to minimize the error more substantially. In some embodiments, weights may be added to the droop profile over or corresponding to one or more tones to achieve a better or predetermined approximation error in building the undrooping model with the coefficients associated with the weights. For example, more weights may be added to the edge tones for better droop compensation onto the edge tones. Although the LMS method is taken as an example for the modeling description, the present disclosure does not place limitations thereto. In other embodiments, the Least-Square method or Weighted-Least-Square method may also be employed to model the droop profile.

[0071] At 520, droop and undrooping modeling module(s) 4032 at RF chip 402 may be configured to obtain a representation for the undrooping profile in the log domain. As described above, through the log conversion, the undrooping profile may be an inverse of the droop profile. Therefore, in building the undrooping model, from the droop gains in the log domain to obtain the undrooping gains in the log domain, only a sign flipping operation may be required. This can also be considered as an advantage of using the log domain. By contrast, in the original frequency domain, the drooping and undrooping are mutually inverse, and thus one or more expensive division operations must be applied to obtain the undrooping gains from the drooping gains for each tone. In view of that, the modeling cost in the log domain is substantially smaller.

[0072] Consistent with the spirit of the present disclosure, to enhance droop modeling, a profile for describing a dynamic range of the coefficients of the droop curve is provided. This method can be applied to various communication systems, such as an OFDM-based wireless communication system (e.g., 4G LTE, 5GNR, or Wi-Fi).

[0073] In some embodiments, the modeled undrooping profile in the log domain with a chosen base, e.g., base 2 (it can be understood that other bases are also applicable), may be approximated and expressed in terms of a polynomial function as: where a(i) denotes the ith coefficient in the polynomial function for the term f to approximate the modeled undrooping profile X(t), t denotes a tone index difference between a given tone and the tone in the center frequency, z denotes a power index inside the polynomial function, and N denotes the maximum order of the polynomial function. The indices t and z may be non-negative integers, i.e., t=0, 1, 2...and z=0, 1, 2...The index t is an independent variable for the polynomial function. If a frequency difference between two adjacent tones is determined to be, e.g., 15KHz, the frequency difference between the given tone and the tone in center frequency can be expressed as t times 15KHz.

[0074] In one instance, the tone index t for the center frequency can be set as 0. That is, the tone in the center frequency may be taken as a reference point, and the undrooping profile may be approximated, in a half-profile manner, based on an index difference (i.e., f) over the frequency from the tone (e.g., 0) in the center frequency. In some embodiments, another tone other than the tone in the center frequency, such as a tone in the band edge, may be taken as a reference point, and the undrooping profile may be approximated in a whole-profile manner. In some instances, the i ^th coefficient a(i) in the polynomial function may be expressed in an array form that has a size of N+1, such as a(i) =[rz(O), rz(l), ..., a(N)], where a(0) may correspond to the constant coefficient in the polynomial function, rz(l) may correspond to the first-order coefficient in the polynomial function, and so on.

[0075] For better understanding, in an example of a third-order polynomial profile, Equation (1) can be expanded and expressed as, e.g., X(t) ≈ a(0) + a(1) * t + a(2) * t ² , where t may start at 0 that is used to represent the tone index corresponding to the center frequency. Based on Equation (1), the un drooping profile X(f) in the log domain can be approximated to the polynomial function.

[0076] The polynomial coefficients a(i) can be used as a representation of the undrooping profile in the log domain. Therefore, in the present disclosure, the terms “coefficients” and “representation” may be used to refer to a(i) in the polynomial function of Equation (1) and can be used interchangeably.

[0077] By storing the representation to local memory 4026 or external memory 408, the undrooping profile in the log domain can be reconstructed later in droop estimation and compensation processing through a droop compensation unit 4042. Droop compensation unit 4042 may be configured to retrieve the representation of the undrooping profile and perform droop estimation and compensation based on the representation. Droop compensation unit 4042 may be implemented at baseband chip 404 as shown in FIG. 4, whereas in other embodiments, droop compensation unit 4042 may be implemented locally at RF chip 402. The present disclosure does not limit thereto.

[0078] In Equation (1), the coefficients a(i) in the polynomial function may belong to floating-point numbers, with which the hardware structure may hardly cope. Therefore, in some embodiments, the coefficients a(i) can be further approximated by a set of integer numbers. Therefore, corresponding hardware may be easily embodied. For example, the coefficients a(i) in the modeled undrooping profile X(t) may be approximated by two integer numbers aINT(i) and expINT(i) and can be expressed as: where aINT(i) is a mantissa of an approximation of the coefficients a(i), and expINT(i) is an exponent of the approximation of the coefficients a(i), and both aINT(i) and expINT(i) belong to integer numbers. [0079] It can be understood that Equation (2) is only given as an example to describe a manner about how to translate the floating-point coefficients a(i) into integers. In other embodiments, the floating-point coefficients a(i) may be approximated by another form of different integers. The present disclosure does not place limitations thereto. [0080] In some embodiments, instead of storing the floating-point coefficients a(i), the integer mantissa aINT(i) and the integer exponent expINT(i) may be stored in, e.g., local memory 4026. In the subsequent procedures, the integer mantissa aINT(i) and the integer exponent expINT(i) may be retrieved and used to reconstruct the undrooping profile for droop estimation and compensation. It can be understood that the integer mantissa aINT(i) and the integer exponent expINT(i) construct a pseudo-floating-point representation corresponding to the floating-point coefficients a(i). Therefore, in the present disclosure, the term “representation of the undrooping profile” may be also used to refer to the integer mantissa aINT(i) and the integer exponent expINT(i). [0081] As described, the modeling of the undrooping profile is a dynamic design process. That is, the modeling is associated with various dynamic ranges and different coefficients to determine an applied model. Therefore, through Equation (2), the floating-point numbers representing the coefficients a(i) can be manipulated in a manner using additional bits to indicate the dynamic range in terms of integers so as to obtain simpler implementation and faster processing time. [0082] For example, a polynomial function with a coefficient array a(i) = [-6611.1034 5282.4816 -1461.0197 176.2256 -0.5946 0.0011565568] may be used to approximate an undrooping polynomial function X(t). It can be obtained that a(0)/a(5) = -6611.1034/0.0011565568 = -5.7e ⁶ = -2 ^22.44 , which represents a shared range of the six coefficients. If a sufficient accuracy for a(0:5) with the shared range is intended to be maintained, roughly 30 bits can be required with the original coefficients a(i) = [-6611.1034 5282.4816 -1461.0197 176.2256 -0.5946 0.0011565568], This means that an exceptionally large dynamical range for the floating-point number coefficients a(i) is demanded. However, a(0) and a(5), at the boundaries of the array, may have many least significant bits (LSBs) equal to 0 and many most significant bits (MSBs) equal to 0, respectively; therefore, the spaces are wasted. By contrast, based on the floating-point implementation using Equation (2), a(0:5) can be represented by 16-bit aINT(i) and 4-bit expINT(i) : aINT(i) = [-26444 21130 -23376 22557 -19485 38] (3), and expINT(i)= [2 2 4 7 15 15] (4), where aINT(i)/2 ^expINT (i) is an approximation for the coefficients a(i) (or the pseudo-floating-point representation) that can be expressed as: a(i)« aINT(i)/2 ^expINT (i) = [-6611.0 5282.5 -1461.0 176.2266 0.5946 0.00115966] (5).

[0083] The accuracy of the coefficients and the approximation by aINT(i) and expINT(i) can be higher than 70dB with the overall 20 bits (16 bits for aINT(i) and 4 bits for expINT(i)) and the results of the above pseudo-floating-point representation show satisfactory accuracy on an average between the floating-point implementation and the fixed-point implementation.

[0084] With the dynamic range scheme, the memory cost for saving the dynamic range for each coefficient can be apparent. In view of the above case, 6 (the amount of the coefficients) x [16 bits (for aINT) + 4 bits (for expINT)] = 120 bits. In reality, some coefficients may have a smaller bit width, and thus the required memory may be further reduced. By comparison, an identical range for all the floating-point coefficients may demand more memory space. That is, it will be larger than 6 (number of coefficients) x 30 (bits per coefficient) = 180 bits, while the corresponding accuracy cannot reach the same level that aINT and expINT can support. To make it worse, the representation with the larger bit width may cause higher expenses in its implementation due to more cycles or more power required.

[0085] Consistent with the spirit of the present disclosure, the various dynamic range for the coefficients enhances the accuracy of the undrooping profile. It also enables simple linear interpolation, e.g., through hardware implementation, to obtain the undrooping gains at a relatively low cost of power, timing, and implementation complexity. Consequently, the overall cost is reduced significantly. The additional parameters (aINT and expINT) indicating the various dynamic range can be absorbed into the number of bits when calculating the undrooping gains. In turn, the cost for these parameters may be rewarded by simpler implementation, in addition to enhancing the accuracy. In addition, a LUT for finding the undrooping gains is no longer needed. Only the order and coefficients of the undrooping profile and the dynamic range that corresponds to bit shifting information for calculating the undrooping gains are required to be saved.

[0086] Consistent with the spirit of the present disclosure, through compressing the undrooping profile to the log domain, the memory storage can be significantly reduced. In comparison with the segmented piecewise polynomial approaches, a reduction in the memory storage by multiple times can be obtained. Roughly, if an N-segmented droop profile is used, there may be N-time reduction. In expanding the log-domain undrooping gains, the hardware implementation can be significantly simplified and provide support for both single carrier scenarios and CA scenarios. Meanwhile, the complexity of the DFE control can also be reduced.

[0087] Embodiments of an apparatus, a system, and a method, implementing droop calibration, are provided. In one aspect, the present disclosure provides some embodiments of the apparatus implementing droop calibration. The apparatus may include a receiver associated with a radio frequency (RF) band, a processor, and memory coupled to the processor and storing instructions. When executed by the processor, the instructions may cause the processor to tune a frequency of the RF band of the receiver to a selected band. A transmitter may be configured to transmit signals in the selected band. Samples may be received in a time domain by the receiver based on the signals and the RF band. A plurality of frequency tones in a frequency domain may be obtained based on the samples in the time domain. Each of the plurality of frequency tones may correspond to a power level. A droop profile in a logarithmic (log) domain may be modeled based on a log conversion of the power levels of the frequency tones. A representation of an undrooping profile in the log domain may be obtained based on the droop profile in the log domain. The undrooping profile may be an inverse of the droop profile.

[0088] In some embodiments, the instructions may further cause the processor to fine-tune the frequency of the RF band in the selected band. The RF band may include frequency responses that include a substantially zero slope around a center frequency of the RF band.

[0089] In some embodiments, the RF band may include a bandwidth that is larger than or equal to a selected bandwidth for estimating droop. In some embodiments, the RF band may include a time division duplex (TDD) band. In some embodiments, the transmitter may be configured to transmit white-noise random signals or white-noise-like random signals.

[0090] In some embodiments, the instructions may further cause the processor to model the droop profile in the log domain based on the log conversion of the power levels of the frequency tones using a least-mean-square (LMS) method.

[0091] In some embodiments, in response to a size of the samples being greater than a threshold, a fast Fourier transform (FFT) circuit may perform an FFT on the samples in the time domain to transform the samples to the plurality of frequency tones in the frequency domain. The threshold may be determined by at least one of a sample size configured for covering a selected bandwidth for estimating droop or an estimation accuracy.

[0092] In some embodiments, the instructions may further cause the processor to obtain a plurality of FFT outputs corresponding to each frequency tone based on the samples. An average of the FFT outputs corresponding to each frequency tone may be obtained. A droop, corresponding to each frequency tone, in the log domain may be obtained based on the average of the FFT outputs for each frequency tone. The droop profile in the log domain may be modeled based on the droop, corresponding to each frequency tone, in the log domain.

[0093] In some embodiments, the instructions may further cause the processor to obtain droops, corresponding to the plurality of frequency tones, in the log domain based on the log conversion of the power levels of the plurality of tones. A curve-fitting may be performed on the droops, corresponding to the plurality of frequency tones, in the log domain to obtain a polynomial function. The undrooping profile may include the polynomial function.

[0094] In some embodiments, the instructions may further cause the processor to approximate the undrooping profile using a polynomial function with an i ^th coefficient a(i), i indicating a coefficient index in the polynomial function, and i being a non-negative integer. In some embodiments, the processor may be a first processor. A second processor may be configured to perform at least one of droop estimation or droop compensation based on the coefficient a(i).

[0095] In some embodiments, the coefficient a( ) may be a floating-point coefficient. The instructions may further cause the processor to obtain an approximation of the floating-point coefficient a(i) based on operations of a set of integers.

[0096] In another aspect, the present disclosure provides some embodiments of the system implementing droop calibration. The system may include a transmitter, a first apparatus, and a second apparatus. The first apparatus may include a receiver associated with a radio frequency (RF) band, a processor, and memory coupled to the processor and storing instructions. The instructions may cause the processor to tune a frequency of the RF band of the receiver to a selected band for estimating droop. A transmitter may be configured to transmit signals in the selected band. Samples may be received in a time domain based on the signals and the RF band. A plurality of frequency tones in a frequency domain may be obtained based on the samples in the time domain. Each of the plurality of frequency tones may correspond to a power level. A droop profile in a logarithmic (log) domain may be modeled based on a log conversion of the power levels of the frequency tones. A representation of an undrooping profile in the log domain may be obtained and stored based on the droop profile in the log domain. The undrooping profile may be an inverse of the droop profile. The second apparatus may be configured to retrieve the representation of the undrooping profile in the log domain. At least one of droop estimation or droop compensation may be performed based on the representation of the undrooping profile.

[0097] In some embodiments, the first apparatus may include a radio frequency chip. The second apparatus may be the radio frequency chip. In some embodiments, the first apparatus may include a radio frequency chip. The second apparatus may include a baseband chip.

[0098] In still another aspect, the present disclosure provides some embodiments of the method implementing droop calibration. The method may include tuning a frequency of a radio frequency (RF) band of a receiver to a selected band. Signals in the selected band may be transmitted by a transmitter. Samples in a time domain may be received by the receiver based on the signals and the RF band. The samples in the time domain may be transformed into a plurality of frequency tones in a frequency domain. Each of the plurality of frequency tones may correspond to a power level. A droop profile in a logarithmic (log) domain may be modeled based on a log conversion of the power levels of the frequency tones. A representation of an undrooping profile in the log domain may be obtained based on the droop profile in the log domain. The undrooping profile may be an inverse of the droop profile.

[0099] In some embodiments, tuning the frequency of the RF band of the receiver may include fine-tuning the frequency of the RF band in the selected ban. The RF band may include frequency responses that comprise a substantially zero slope around a center frequency of the RF band. In some embodiments, the method may further include in response to a size of the samples being greater than a threshold, performing an FFT on the samples in the time domain to transform the samples into the plurality of frequency tones in the frequency domain. The threshold may be determined by at least one of a sample size configured for covering a selected bandwidth for estimating droop or an estimation accuracy.

[0100] In some embodiments, the method may further include approximating the undrooping profile using a polynomial function in a form of: where a(i) may denote an i ^th coefficient, corresponding to a power index z in the polynomial function, t may denote a tone index difference between a given frequency tone and a tone in a center frequency of the RF band, and N may denote a maximum order of the polynomial function. t and N may be non-negative integers.

[0101] In some embodiments, the coefficient a(i) may be a floating-point coefficient. The method may further include obtaining an approximation of the floating-point coefficient a(i) based on operations of a set of integers.

[0102] The foregoing description of the specific embodiments will so reveal the general nature of the present disclosure that others can, by applying knowledge within the skill of the art, readily modify and/or adapt for various applications such specific embodiments, without undue experimentation, without departing from the general concept of the present disclosure. Therefore, such adaptations and modifications are intended to be within the meaning and range of equivalents of the disclosed embodiments, based on the teaching and guidance presented herein. It is to be understood that the phraseology or terminology herein is for the purpose of description and not of limitation, such that the terminology or phraseology of the present specification is to be interpreted by the skilled artisan in light of the teachings and guidance.

[0103] Embodiments of the present disclosure have been described above with the aid of functional building blocks illustrating the implementation of specified functions and relationships thereof. The boundaries of these functional building blocks have been arbitrarily defined herein for the convenience of the description. Alternate boundaries can be defined so long as the specified functions and relationships thereof are appropriately performed.

[0104] The Summary and Abstract sections may set forth one or more but not all exemplary embodiments of the present disclosure as contemplated by the inventor(s), and thus, are not intended to limit the present disclosure and the appended claims in any way.

[0105] Various functional blocks, modules, and steps are disclosed above. The particular arrangements provided are illustrative and without limitation. Accordingly, the functional blocks, modules, and steps may be re-ordered or combined in diverse ways than in the examples provided above. Likewise, certain embodiments include only a subset of the functional blocks, modules, and steps, and any such subset is permitted.

[0106] The breadth and scope of the present disclosure should not be limited by any of the above-described exemplary embodiments but should be defined only in accordance with the following claims and their equivalents.