Multi-packet crystal offset estimation

The Kalman filter-based method for estimating carrier frequency offset across multiple packets addresses accuracy and power efficiency issues in UWB systems, enhancing range and reliability by refining estimates over repeated packets.

WO2026235771A1PCT designated stage Publication Date: 2026-11-12QORVO US INC
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
PCT/US2026/016847
Authority / Receiving Office
WO · WO
Patent Type
Applications
Current Assignee / Owner
Priority Date
2025-05-09
Filing Date
2026-02-26
Publication Date
2026-11-12

AI Technical Summary

Technical Problem

Current wireless communication systems face challenges in achieving accurate crystal frequency offset estimation, particularly in narrowband assist (NBA) systems, due to factors like packet length, signal-to-noise ratio (SNR), multipath fading, and doppler shifts, which limit the range and power efficiency of UWB systems.

Method used

A method and system for estimating carrier frequency offset (CFO) using a Kalman filter-based frequency estimator that updates frequency estimation parameters across multiple packets, improving accuracy by leveraging previous estimates and error signals to correct for frequency and phase offsets.

Benefits of technology

Enhances crystal offset estimation accuracy, reducing power consumption and extending the range of UWB systems by enabling coherent combining of millisecond UWB fragments, suitable for applications like Apple's 'Find My' and secure keyless-entry systems.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure US2026016847_12112026_PF_FP_ABST
    Figure US2026016847_12112026_PF_FP_ABST
Patent Text Reader

Abstract

Aspects of the disclosure include a method of estimating carrier frequency offset, the method including receiving first and second transmitted packets, the first and second transmitted packets each including a plurality of symbols, repeatedly updating a set of frequency estimation parameters based on a set of previous values of the set of frequency estimation parameters, and based on each of the plurality of symbols of the first and second transmitted packets, with a frequency estimator, determining an updated frequency offset estimate based on each updated set of frequency estimation parameters, and programming an offset correction circuit based on each updated frequency offset estimate, the offset correction circuit configured to generate frequency offset corrected data based on received data and based on the updated frequency offset estimates.
Need to check novelty before this filing date? Find Prior Art

Description

Docket No. QID241632 / 62306.204 WOO 1MULTI-PACKET CRYSTAL OFFSET ESTIMATIONInventor: Andrew FortRELATED APPLICATIONS

[0001] The present application claims the benefit of U.S. Provisional application no.63 / 803,272, filed May 09, 2025, which is hereby expressly incorporated by reference herein in its entirety.TECHNICAL FIELD

[0002] The present disclosure relates generally to wireless communication devices having receivers, and in particular, relates to receivers which estimate carrier frequency and phase offsets.BACKGROUND

[0003] Communication devices, such as radios and cell phones have receivers which receive, for example, RF signals from an antenna, and process the RF signals to generate digital data, for example, which is then provided to other components of the devices, such as a processor. Naturally, there is an offset between carrier frequencies of the received signals and the local oscillator of the receiver. Systems and methods for measuring the offsets are used to mitigate errors caused by the offsets. Therefore, there is a need to improve performance of offset measurement in receivers.SUMMARY

[0004] Embodiments of the present disclosure include carrier frequency offset (CFO) estimation and correction systems and techniques.

[0005] One inventive aspect is a method of estimating carrier frequency offset, the method including receiving first and second transmitted packets, the first and second transmitted packets each including a plurality of symbols, repeatedly updating a set of frequency estimation parameters based on a set of previous values of the set of frequency estimation parameters, and based on each of the plurality of symbols of the first and second transmitted packets, with a frequency estimator, determining an updated frequency offset estimate based on each updated set of frequency estimation parameters, and programming an offset correction circuit based on each updated frequency offset estimate, the offset correction circuit configured to generateDocket No. QID241632 / 62306.204 WOO 1frequency offset corrected data based on received data and based on the updated frequency offset estimates.

[0006] In some embodiments, the method further includes receiving additional transmitted packets, the additional transmitted packets each including a plurality of symbols, repeatedly updating the set of frequency estimation parameters based on a set of previous values of the set of frequency estimation parameters updated based on the symbols of the first and second transmitted packets, and based on each of a plurality of subsequent symbols of the additional transmitted packets, and with the frequency estimator, determining an updated frequency offset estimate based on each set of frequency estimation parameters updated based on the set of previous values of the set of frequency estimation parameters and based on the subsequent symbols of the additional transmitted packets.

[0007] In some embodiments, the method further includes generating an error signal based on each symbol of the first and second transmitted packets and based on a corresponding frequency offset estimate, where the set of frequency estimation parameters are repeatedly updated based on the error signals.

[0008] In some embodiments, the method further includes the frequency estimation parameters include a Kalman filter gain, and where the frequency estimator includes a Kalman Filter.

[0009] In some embodiments, the method further includes synchronizing to first and second preambles of the first and second transmitted packets, respectively, to respectively generate first and second initial phase offset estimates for the first and second transmitted packets, where the first transmitted packet is received from a first external device, and where the second transmitted packet is received from a second external device.

[0010] In some embodiments, the method further includes repeatedly updating a first subset of the frequency estimation parameters based on previous values of the first subset of frequency estimation parameters, based on each of a plurality of subsequent symbols of the first transmitted packet, and based on subsequent symbols of one or more additional packets received from the first external device, and repeatedly updating a second subset of the frequency estimation parameters based on previous values of the second subset of frequency estimation parameters, based on each of a plurality of subsequent symbols of the secondDocket No. QID241632 / 62306.204 WOO 1transmitted packet, and based on subsequent symbols of one or more additional packets received from the second external device.

[0011] In some embodiments, the method further includes determining the updated frequency offset estimate is based on an updated set of phase estimation parameters.

[0012] In some embodiments, the method further includes the phase estimation parameters include a Kalman filter gain.

[0013] Another inventive aspect is a receiver, including analog circuitry including an input and one or more outputs, one or more analog to digital converters (ADCs) coupled to the one or more outputs and including one or more digital outputs, and a processor configured to receive first and second transmitted packets, the first and second transmitted packets each including a plurality of symbols, repeatedly update a set of frequency estimation parameters based on a set of previous values of the set of frequency estimation parameters and based on each of a plurality of symbols of the first and second transmitted packets, determine an updated frequency offset estimate based on each updated set of frequency estimation parameters, and program an offset correction circuit based on each updated frequency offset estimate, the offset correction circuit configured to generate frequency offset corrected data based on data received from the ADCs and based on the updated frequency offset estimates.

[0014] In some embodiments, the receiver further includes the processor is further configured to receive additional transmitted packets, the additional transmitted packets each including a plurality of symbols, repeatedly update the set of frequency estimation parameters based on a set of previous values of the set of frequency estimation parameters and based on each of a plurality of subsequent symbols of the additional transmitted packets, and determine an updated frequency offset estimate based on each set of frequency estimation parameters updated based on the set of previous values of the set of frequency estimation parameters and based on the subsequent symbols of the additional transmitted packets.

[0015] In some embodiments, the receiver further includes the processor is further configured to generate an error signal based on each symbol of the first and second transmitted packets and based on a corresponding frequency offset estimate, where the set of frequency estimation parameters are repeatedly updated based on the error signals.Docket No. QID241632 / 62306.204 WOO 1

[0016] In some embodiments, the receiver further includes the frequency estimation parameters include a Kalman filter gain.

[0017] In some embodiments, the receiver further includes the processor is further configured to synchronize to first and second preambles of the first and second transmitted packets, respectively, to respectively generate first and second initial phase offset estimates for the first and second transmitted packets, where the first transmitted packet is received from a first external device, and where the second transmitted packet is received from a second external device.

[0018] In some embodiments, the receiver further includes the processor is further configured to repeatedly update a first subset of the frequency estimation parameters based on previous values of the first subset of frequency estimation parameters, based on each of a plurality of subsequent symbols of the first transmitted packet, and based on subsequent symbols of one or more additional packets received from the first external device, and repeatedly update a second subset of the frequency estimation parameters based on previous values of the second subset of frequency estimation parameters, based on each of a plurality of subsequent symbols of the second transmitted packet, and based on subsequent symbols of one or more additional packets received from the second external device.

[0019] In some embodiments, the receiver further includes determining the updated frequency offset estimate is based on an updated set of phase estimation parameters.

[0020] In some embodiments, the receiver further includes the phase estimation parameters include a Kalman filter gain.

[0021] Another inventive aspect is a wireless device, including an antenna input port, an antenna switch coupled to the antenna input port, and a receiver, including analog circuitry including an input coupled to the antenna switch and one or more outputs, one or more analog to digital converters (ADCs) coupled to the one or more outputs and including one or more digital outputs, and a processor configured to receive first and second transmitted packets, the first and second transmitted packets each including a plurality of symbols, repeatedly update a set of frequency estimation parameters based on a set of previous values of the set of frequency estimation parameters and based on each of a plurality of symbols of the first and second transmitted packets, determine an updated frequency offset estimate based on each updated set of frequency estimation parameters, and program an offset correction circuit based on eachDocket No. QID241632 / 62306.204 WOO 1updated frequency offset estimate, the offset correction circuit configured to generate frequency offset corrected data based on data received from the ADCs and based on the updated frequency offset estimates.

[0022] In some embodiments, the wireless device further includes the frequency estimation parameters include a Kalman filter gain.

[0023] In some embodiments, the wireless device further includes the processor is further configured to synchronize to first and second preambles of the first and second transmitted packets, respectively, to respectively generate first and second initial phase offset estimates for the first and second transmitted packets, where the first transmitted packet is received from a first external device, and where the second transmitted packet is received from a second external device, repeatedly update a first subset of the frequency estimation parameters based on previous values of the first subset of frequency estimation parameters, based on each of a plurality of subsequent symbols of the first transmitted packet, and based on subsequent symbols of one or more additional packets received from the first external device, and repeatedly update a second subset of the frequency estimation parameters based on previous values of the second subset of frequency estimation parameters, based on each of a plurality of subsequent symbols of the second transmitted packet, and based on subsequent symbols of one or more additional packets received from the second external device.

[0024] In some embodiments, the wireless device further includes determining the updated frequency offset estimate is based on an updated set of phase estimation parameters including a Kalman filter gain.

[0025] Additional aspects, embodiments, implementations, features, and advantages of the present disclosure will become apparent from the following detailed description.BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Illustrative embodiments of the present disclosure will be described with reference to the accompanying drawings, of which:

[0027] Figure 1 illustrates a receiver circuit, according to aspects of the present disclosure.

[0028] Figure 2 illustrates a schematic block diagram of a carrier frequency offset system according to aspects of the present disclosure.Docket No. QID241632 / 62306.204 WOO 1

[0029] Figure 3 illustrates a flowchart for a method of estimating carrier frequency offset, according to aspects of the present disclosure.

[0030] Figure 4 illustrates a schematic block diagram of a carrier frequency offset system according to aspects of the present disclosure.

[0031] Figure 5 illustrates a flowchart for a method of estimating carrier frequency offset, according to aspects of the present disclosure.

[0032] Figure 6 illustrates a flowchart for a method of correcting for a carrier frequency offset of a receiver, according to aspects of the present disclosure.

[0033] Figure 7 illustrates a flowchart for a method of correcting for a carrier frequency offset of a receiver, according to aspects of the present disclosure.

[0034] Figure 8 illustrates a communication device, according to aspects of the present disclosure.DETAILED DESCRIPTION

[0035] For the purposes of promoting an understanding of the principles of the present disclosure, reference will now be made to the embodiments illustrated in the drawings, and specific language will be used to describe the same. It is nevertheless understood that no limitation to the scope of the disclosure is intended. Any alterations and further modifications to the described devices, systems, and methods, and any further application of the principles of the present disclosure are fully contemplated and included within the present disclosure as would normally occur to one skilled in the art to which the disclosure relates. In particular, it is fully contemplated that the features, components, and / or steps described with respect to one embodiment may be combined with the features, components, and / or steps described with respect to other embodiments of the present disclosure. For the sake of brevity, however, the numerous iterations of these combinations will not be described separately.

[0036] Current UWB systems are limited to short ranges due, for example, to regulations capping transmitted energy within 1ms to prevent interference with other wireless systems. Much of this energy is used for searching and synchronizing with pulse trains (e.g., called “fragments”), leaving less energy available for, for example, secure ranging. To extend the range of UWB, the emerging IEEE 802.15.4ab standard has introduced a technique referred to as “Narrowband Assist” or “NBA”. Regulations allow narrowband systems to transmit at much higher power levels and at lower frequencies leading to potentially longer range than UWB. Using a narrowband system for initial synchronization to estimate the crystal frequency offset between devices and as an initial coarse timing estimate allows all energy in a 1ms fragmentDocket No. QID241632 / 62306.204 WOO 1to be used for secure ranging. This improves link budget and reduces power consumption. Furthermore, with an accurate enough crystal offset estimate, multiple 1ms fragments can be combined together maximizing the range of the system. This is referred to as “multimillisecond” ranging.

[0037] In IEEE 802.15.4ab NBA multi-millisecond ranging, transmissions are typically organized into ranging rounds which consist of initial narrowband packet exchanges used for timing and crystal offset synchronization, followed by one or more UWB fragment exchanges for secure ranging. In many applications, such as “find my” applications and “digital keyless entry” applications, ranging rounds are repeated periodically to track the motion and position of devices.

[0038] Current issues with crystal offset estimation include: 5 ppb accuracy is needed, for example, for optimal coherent combining of the energy in large numbers of UWB fragments but is difficult to obtain under all circumstances; and actual accuracy can depend on a variety of “run-time” factors and environmental conditions.

[0039] In general, the more accurately the NBA modem can determine the crystal frequency offset between devices, the less correlations across different frequencies are needed by the UWB system in synchronizing to a UWB fragment. This leads to lower power consumption. Furthermore, the more accurate the crystal offset estimate, the more millisecond UWB fragments can be combined. With an accurate enough estimate, many of these fragments can be combined at least partially coherently. This ultimately leads to better UWB range. At least for these reasons, NBA demands a much higher crystal offset estimation accuracy than traditional communication systems and more advanced CFO estimation strategies will be needed in the future to accommodate this.

[0040] Embodiments discussed herein extend the traditional single-packet crystal offset estimation scheme used in conventional radios to a continuous frequency offset estimate across multiple packets. By combining information from multiple packets, embodiments enable a more accurate crystal offset estimates which is less dependent on factors like packet length, SNR, multipath fading, doppler shifts, or worst-case remote device’s phase noise specifications. Therefore, these embodiments have benefits which result in lower-power, more reliability, and longer-range NBA-UWB performance.

[0041] Embodiments are relevant, for example, for any NBA multi-millisecond ranging application where ranging rounds are repeated periodically to track the motion of a device. This includes such important applications as Apple’s “Find My” application and secure keyless-entry applications.Docket No. QID241632 / 62306.204 WOO 1

[0042] Figure 1 illustrates a receiver circuit 100 connected to antenna 110, according to aspects of the present disclosure. Receiver circuit 100 includes antenna switch 105, RF front end 115, IF circuit 125, ADC circuit 135, CFO / phase correction system 145, processor 155, and memory 165. Receiver circuit 100 is configured to perform or partially perform the functions discussed elsewhere herein, for example, as related to the beneficial aspects discussed herein.

[0043] Receiver circuit 100 is configured to receive RF signals from antenna 110. The RF signals may include a baseband transmission signal which modulate an RF carrier signal.

[0044] Antenna switch 105 is configured to selectively couple RF front end 115 to an antenna input port 107. In the illustrated embodiment, the antenna input port 107 is connected to antenna 110. When coupled to antenna 110, RF front end 115 receives the RF signals from antenna 110.

[0045] The RF front end 115 may, for example, be configured to process the RF signal so as to generate in-phase (I) and quadrature (Q) signals at an intermediate frequency. For example, the RF front end 115 may comprise a low-noise amplifier and a filter configured to amplify and remove broadband interference from the received RF signal. In some embodiments, the RF front end 115 comprises a low-noise amplifier circuit known in the art. In some embodiments, the RF front end 115 comprises a filter circuit known in the art. In some embodiments, the RF front end 115 comprises a mixer circuit configured to generate the intermediate frequency I and Q signals. In some embodiments, the RF front end 115 comprises a mixer circuit known in the art. Accordingly, RF front end 115 is configured to generate intermediate frequency I and Q signals which correspond with the RF signal. For example, the intermediate frequency I and Q signals may preserve and represent or carry the information of the baseband transmission signal modulating the RF carrier signal of the RF signal received from antenna 110.

[0046] The IF circuit 125 may, for example, be configured to process the intermediate frequency I and Q signals so as to generate I and Q baseband signals at a baseband frequency. For example, the IF circuit 125 may comprise amplifier and / or filtering circuits configured to amplify and remove noise from the received intermediate frequency I and Q signals. In some embodiments, the IF circuit 125 comprises an amplifier circuitry known in the art. In some embodiments, the IF circuit 125 comprises filter circuitry known in the art. In some embodiments, the IF circuit 125 comprises a mixer circuit configured to generate the baseband frequency I and Q signals. In some embodiments, the IF circuit 125 comprises a mixer circuit known in the art. Accordingly, IF circuit 125 is configured to generate baseband frequency IDocket No. QID241632 / 62306.204 WOO 1and Q signals which correspond with the RF signal. For example, the baseband frequency I and Q signals may preserve and represent or carry the information of the baseband transmission signal modulating the RF carrier signal of the RF signal received from antenna 110. The RF front end 115 and IF circuit 125 may together be referred to as analog circuitry or an analog front-end. In some embodiments, the analog circuitry or analog front end may comprise additional circuits. There may be other configurations for converting an RF signal to baseband and the delay determination or estimation described herein is applicable to those configurations as would be understood in the art.

[0047] ADC circuit 135 is configured to receive the baseband frequency I and Q signals from IF circuit 125. In addition, ADC circuit 135 is configured to digitize the baseband frequency I and Q signals to generate I and Q digital signals. In some embodiments, ADC circuit 135 comprises ADC circuitry known in the art. In some embodiments, ADC circuit 135 comprises a first ADC circuit to convert I signals from IF circuit 125 and a second ADC circuit to convert Q signals from IF circuit 125.

[0048] CFO / phase correction system 145 is configured to receive the digital signals from ADC circuit 135 and to generate CFO and phase corrected signals for processor 155.Embodiments of CFO / phase correction systems are discussed in more detail elsewhere herein. In some embodiments, CFO / phase correction system 145 comprises a first CFO / phase correction system to correct CFO and phase for I digital signals from ADC 135, and a second CFO / phase correction system to correct CFO and phase for Q digital signals from ADC 135.

[0049] Processor 155 is configured to receive the CFO and phase corrected digital signals from CFO / phase correction system 145. In addition, processor 155 is configured to process the corrected digital signals according to the functionality of the application or system using receiver circuit 100. For this purpose, processor 155 may include digital circuitry configured to Process the corrected digital signals.

[0050] Processor 155 may be configured to perform or partially perform the functions discussed elsewhere herein, for example, as related to the beneficial aspects discussed herein. In some embodiments, processor 155 is configured to perform or partially perform the functions of CFO / phase correction system 145. Accordingly, in some embodiments, CFO / phase correction system 145 is integrated into processor 155. In some embodiments, processor 155 is configured to transmit signals to other components of receiver circuit 100 which cause the other components to perform or partially perform the functions discussed elsewhere herein, for example, as related to the beneficial aspects discussed herein.Docket No. QID241632 / 62306.204 WOO 1

[0051] Processor 155 may be configured to cause data generated based at least partly on the corrected digital signals to be stored in memory 165. Processor 155 may be configured to receive data from memory 165 and to process the corrected digital signals based in part on the received data. Processor 155 may be configured to receive instruction codes from memory 165 and to process the corrected digital signals based in part on the received instruction codes.

[0052] Figure 2 illustrates a schematic block diagram of a carrier frequency offset (CFO) / phase correction system 200 according to aspects of the present disclosure. CFO / phase correction system 200 is an embodiment of a narrowband wireless communication system control loop for estimating and tracking carrier frequency offsets between a received RF signal and the receiver having the CFO / phase correction system 200. CFO / phase correction system 200 includes CFO / phase error correction block 210, correlator bank 220, phase and CFO error estimator 230, and frequency estimator 240. CFO / phase correction system 200 includes blocks which may be typical for 802.15.4 and 802.15.4ab NBA system. However, alternative embodiments may be used in other communication systems as well, and other variations are contemplated.

[0053] Digital correction of the frequency and phase of the received digital signals may be employed to correct for differences between the received signal and the local oscillator.

[0054] CFO / phase error correction block 210 is configured to receive digital signals which may be subject to CFO / phase errors. In addition, CFO / phase error correction block 210 is configured to correct each next received symbol according to CFO correction and Phase correction signals generated based on previously received symbols. Accordingly, CFO / phase error correction block 210 is configured to generate corrected digital data based on received digital signals and based on CFO correction and Phase correction signals received from frequency estimator 240. Error correction block may have any architecture and is not limited.

[0055] In the illustrated embodiment, CFO / phase error correction block 210 is configured to correct each next received symbol according to CFO correction and Phase correction signals generated based on previously received symbols. In some embodiments, the systems and methods are modified to correct each next received symbol according to CFO correction and Phase correction signals generated based on previously received symbols and additionally based on the next received symbol. For example, in some embodiments, the CFO correction and Phase correction signals are modified based on the next received symbol, and the nextDocket No. QID241632 / 62306.204 WOO 1received symbol is subsequently corrected based on the modified CFO correction and Phase correction signals.

[0056] In the context of spread-spectrum systems, such as 802.15.4 and 802.15.4ab NBA, the corrected signals may be de-spread by correlator bank 220, where each of the correlators of correlator bank 220 is matched to one of a set of expected chip sequences or symbols. In 802.15.4ab NBA 250kbps mode, the correlators each correspond to one of 16 different chip sequences or symbols, each representing 4 information bits. For coherent systems, the symbol corresponding with the correlator generating the complex output having the largest real part is chosen as the most likely received symbol. In non-coherent systems, the symbol corresponding with the correlator generating the complex output having the largest magnitude is chosen as the most likely received symbol. Correlator bank 220 may also be configured to generate a digital output corresponding to the most likely received symbol, for example, for a processor, such as processor 155.

[0057] In some embodiments, the phase and CFO error estimator 230 selects the angle of the complex output of the correlator bank 220 of the most likely received symbol as an error signal an for the frequency estimator 240, which generates CFO correction and phase correction signals.

[0058] In some embodiments, frequency estimator 240 includes a Kalman filter. For example, in some embodiments, the Kalman filter acts to update its current estimates of the carrier frequency offset and phase offset with the new information contained in error signal an. In the Kalman filter terminology, error signal an may therefore be referred to as the innovation.

[0059] Throughout this description, frequency estimator 240 is described as a Kalman filter. In addition, differences over a traditional Kalman filter tracking loop will be presented. The high accuracy, flexibility, and recursive nature of a Kalman filter make it particularly well-suited for some embodiments. However, alternative embodiments may be based on different estimation systems. For example, a phase-locked loop (PLL) can be used, for example, as having properties similar to a Kalman filter. In some embodiments, a PLL may be considered a special case of the more general Kalman filter.

[0060] The Kalman filter may beneficially estimate parameters of a system based on repeated measurements, for example, in Gaussian noise.

[0061] The details of a Kalman filter algorithm are beyond the scope of this document, but the following equations summarize the operation of a Kalman filter as it applies to CFO estimation in various embodiments.Docket No. QID241632 / 62306.204 WOO 1

[0062] Equation (1) predicts an updated phase (0n\n- 1) and frequency (<Sr!|r!-i) estimate from the previous symbol’s estimate of the phaseand frequencyThe predicted estimate is then updated in equation 2 with the innovation, an, scaled by the Kalman Gain factor, Kn. These equations are implemented in the frequency estimator 240, and the output is the most recent phase and frequency estimate (0n\nandwhich are used by CFO / phase error correction block 210 for correcting the subsequent symbol, and by correlator bank 220 and phase and CFO error estimator 230 for determining the next innovation.

[0063] The Kalman Gain factor, Kn, is a 1 by 2 matrix which effectively weights much of the new estimate is based on the previous phase / frequency estimates, and how much of it is based on the most recent error signal. Typically, the magnitude of the Kalman Gain is updated on each symbol and decreases overtime as more confidence is obtained in the phase / frequency estimate compared to the error signal. It can be shown that, under the assumption of zero-mean Gaussian noise and a linear process, the following set of update equations, when used for computing the Kalman gain each symbol (Kn), result in an unbiased phase and frequency estimate with the minimum mean squared error:

[0064] For the specific case of CFO estimation in wireless systems the terms from the above equation may be as follows: F = [1 l;0 1], H =

[0010] , R = co2, P0|0 =[ cp,init2 0;0 of,init2], and Q = 0. In some embodiments, these parameters may be altered to account for practical differences in the implementation.

[0065] F and H are constants specific to this situation. Pn|n from equation (3) can be interpreted as the accuracy of the current estimate at symbol n. Pn|n represents the 2 by 2Docket No. QID241632 / 62306.204 WOO 1covariance matrix of the error for the current phase and frequency estimates, which may be updated recursively or incrementally based on the predicted estimation accuracy of the previous symbol, Pn|n-1. The term co2 represents how noisy the innovations or the error signal an, are expected to be in the system. Equations (1) - (3) are recursive, or iterative, or incremental in the sense that new estimates, gains, and accuracies are calculated from the previous values. Thus, for the very first symbol, an initial estimate of the phase (0O|o)and frequency (eoO|o) is generated and an initial accuracy P0|0 is assigned thereto. The initial estimate of the phase and frequency may be generated, for example, using correlation against a known preamble sequence transmitted at the start of a packet. Other methods for generating the initial estimate may be used. The error variance of the initial preamble-based phase estimate is op,init2, and the frequency estimate is of,init2. These estimates may be uncorrelated. The above equations represent a Kalman filter system and variations may be used, for example, to account for system delays, introduce “forget factors”, or include different types of process errors via the Q term (set to 0, for example,). However, the specific Kalman filter variation is not limited.

[0066] Figure 3 illustrates a flowchart for a method 300 of estimating carrier frequency offset, according to aspects of the present disclosure. Method 300 illustrates a Kalman filter CFO estimation algorithm as used in a wireless communication system. Method 300 may be performed, for example, by receiver circuit 100.

[0067] At block 310, the wireless communication system synchronizes to a preamble to obtain an estimate of the phase and frequency offset between the receiver and the received RF signal. The wireless communication system may use any technique for synchronizing. For example, the wireless communication system may synchronize to a preamble via correlation against a known preamble at multiple frequency offsets and select the correlator and corresponding frequency with largest magnitude.

[0068] At block 320, the registers of the Kalman filter which implement equations (1) - (3) are initialized with the initial phase estimate op,init2 and the initial frequency estimate of, init2 as determined at block 310 based on the preambleand the initial accuracy of the estimate P0|0.

[0069] At block 330, the phase and frequency estimates may be refined for each of a number of successively received symbols by computing the innovation or error signal an and updating the estimate according to equations (1) and (2), for example, as discussed above with reference to figure 2. The Kalman gain of equation (2) may also be updated each symbol according toDocket No. QID241632 / 62306.204 WOO 1equation (3). In some embodiments, the wireless communication system pre-computes a fixed Kalman gain table in advance, where the Kalman gain table stores Kn up to the maximum packet length (N). In some embodiments, the iterative formula depends on the signal statistics: the observation error variance (oo2), the initial preamble-based phase estimation variance (op,init2), and the initial preamble-based frequency estimation variance (of,init2). In some embodiments, these signal statistics are fixed values. In some embodiments, the Kalman gain table is updated after each received symbol. In some embodiments, the Kalman gain table is updated after each received packet. In some embodiments, the Kalman gain table is updated at other intervals, for example, to tradeoff computational complexity and memory.

[0070] The general approach outlined in the above figures and equations has the following characteristics with respect to obtaining an optimal CFO estimate critical for NBA-UWB systems: The final accuracy of the CFO estimate at the end of each packet depends on the initial accuracy of the preamble-based synchronization (P0|0) and the length of the packet; and The accuracy of Kalman filter calculations depends on the accuracy of the signal statistics ( GO2, op,init2, and of,init2).

[0071] Because the Kalman filter iteratively refines an initial CFO estimate, the final estimation accuracy depends on the initial estimate. Unfortunately, initial synchronization and phase and frequency estimation may be computationally intensive, involving large numbers of preamble correlations against multiple frequencies. Accordingly, some embodiments of low-cost NBA-UWB loT systems use coarse initial frequency and phase estimates. Depending on packet length, the Kalman filter may have time to generate a final estimate initial which is affected by the coarse initial frequency and phase estimates. Unfortunately, it is desirable to keep the packet lengths as short as possible in NBA systems to avoid interference with other narrowband systems and to reduce power as much as possible. In some IEEE 802.15.4ab ranging embodiments, the NBA packet is set to a minimum length, carries little or no information, and may be used exclusively for synchronization to the UWB fragments.

[0072] Furthermore, because Kalman filter performance is based on the accuracy of the signal statistics, and the signal statistics may depend on many factors that are unknown and changing over time including the following: SNR, Phase noise characteristics of the remote device, Multipath channel, Speed of motion relative to other device (doppler effects). Some embodiments deal with this by calculating gain tables based on worst-case signal statistics. This typically causes the CFO estimation accuracy to converge slowly to the correct value and can result in sub-optimal estimates, particularly for the short packet lengths used in IEEE 802.15.4ab. While a sub-optimal CFO estimate may be sufficient for some wirelessDocket No. QID241632 / 62306.204 WOO 1communication systems, low-power NBA-UWB systems deploying coherent multimillisecond UWB combining benefit from higher accuracies using techniques discussed below.

[0073] A drawback of CFO estimation schemes used in nearly all current wireless systems and standards is that the estimation scheme starts over for each packet. Therefore, the accuracy of CFO estimates in these systems depends on the packet length, initial preamblesynchronization accuracy, and the signal statistics (typically worst-case).

[0074] In embodiments discussed herein, an iterative estimation across multiple packets is performed. Rather than re-estimating the frequency from the preamble of each packet, the frequency estimator simply continues from where it left off from in the previous packet. In this way, the estimation accuracy continues to improve each packet and each ranging round. Therefore, even if many of the individual packets are quite short and the frequency estimator is initialized for the first packet with parameters chosen for worst-case signal statistics, highly accurate CFO estimates are achieved.

[0075] Figure 4 illustrates a schematic block diagram of a CFO / phase correction system 400 according to aspects of the present disclosure. In some embodiments, all or part of CFO / phase correction system 400 is implemented in digital hardware. In some embodiments, all or part of CFO / phase correction system 400 is implemented as part of programmable processor 155. CFO / phase correction system 400 may be configured to perform the features and aspects regarding CFO and phase estimation and correction discussed herein. For example, CFO / phase correction system 400 may be configured to perform method 500, discussed below.

[0076] CFO / phase correction system 400 includes a programmable CFO / phase correction unit 410, a modem DSP 420, a Kalman filter 430, a programmable Kalman gain table 440, and a controller 450.

[0077] In some embodiments, the Kalman Filter 430 is implemented with another type of control loop, such a phase lock loop (PLL) control loop, or a frequency lock loop (FLL) control loop having a programmable loop bandwidth or similar programmable characteristic stored in the gain table 440 rather than the Kalman gain, and which can be programmed according to a loop error signal.

[0078] Programmable CFO / phase correction unit 410 is configured to receive digital signals which may be subject to CFO / phase errors. In addition, programmable CFO / phase correction unit 410 is configured to correct each next received symbol according to CFO correction and Phase correction signals generated based on previously received symbols. Accordingly, programmable CFO / phase correction unit 410 is configured to generate corrected digital data based on received digital signals and based on CFO correction and Phase correction signalsDocket No. QID241632 / 62306.204 WOO 1received from Kalman filter 430. Error correction block may have any architecture and is not limited.

[0079] In the illustrated embodiment, programmable CFO / phase correction unit 410 is configured to correct each next received symbol according to CFO correction and Phase correction signals generated based on previously received symbols. In some embodiments, the systems and methods are modified to correct each next received symbol according to CFO correction and Phase correction signals generated based on previously received symbols and additionally based on the next received symbol. For example, in some embodiments, the CFO correction and Phase correction signals are modified based on the next received symbol, and the next received symbol is subsequently corrected based on the modified CFO correction and Phase correction signals.

[0080] In some embodiments, CFO / phase correction unit 410 and Kalman filter gain table 440 may be initialized each symbol and / or each packet by controller 450. In some embodiments, the controller 450 is implemented in firmware. The controller 450 may be configured to generate or select the initialization parameters based on, for example, previous CFO and phase estimates and associated loop parameters it has stored. In some embodiments, initialization parameters may be generated or selected based on a particular device from which a packet is expected or is received. Furthermore, the controller 450 may adjust initialization parameters based on other factors, such as current frequency drift, channel conditions, device characteristics, or other characteristics and conditions the controller 450 has access to.

[0081] Programmable CFO / phase correction unit 410 is configured to receive a downconverted and digitized signal from, for example, one or more analog to digital converters. In addition, programmable CFO / phase correction unit 410 is configured to modify the received signal according to a CFO / phase correction algorithm based on CFO correction and phase correction parameters received from Kalman filter 430.

[0082] Modem DSP 420 receives the modified signal from programmable CFO / phase correction unit 410 and generates an error signal an for Kalman filter 430 and for controller 450. For example, to generate the error signal an, modem DSP 420 may comprise a plurality of correlators in a correlator bank having features similar or identical to correlator bank 220, discussed above, and an error generator having features similar or identical to phase and CFO error estimator 230, discussed above. Modem DSP 420 may also be configured to generate a digital output corresponding to the most likely received symbol, for example, for a processor, such as processor 155.Docket No. QID241632 / 62306.204 WOO 1

[0083] Kalman filter 430 may have features similar or identical to frequency estimator 240, discussed above. For example, Kalman filter 430 may be configured to generate CFO correction and phase correction data based on the error signal an using principles discussed above. In the illustrated embodiment, Kalman filter 430 is configured to receive Kalman filter gain parameters from programmable Kalman gain table 440, for example, as updated by controller 450.

[0084] In some embodiments, the Kalman filter gain parameters related to frequency for CFO correction are updated by controller 450 as part of or as a result of processing each of a series of received symbols. In some embodiments, the Kalman filter gain parameters related to frequency for CFO correction are updated by controller 450 as part of or as a result of processing each of a series of received packets.

[0085] In some embodiments, the Kalman filter gain parameters related to phase for phase correction are updated by controller 450 as part of or as a result of processing each of a series of received symbols. In some embodiments, the Kalman filter gain parameters related to phase for phase correction are updated by controller 450 as part of or as a result of processing each of a series of received packets.

[0086] In some embodiments, controller 450 is configured to store previous CFO estimates and / or previous phase estimates.

[0087] In some embodiments, controller 450 is configured to compare a current calculated CFO estimate with an expected CFO estimate range determined, for example, based on a statistical analysis of previous CFO estimates. In addition, controller 450 may be configured to prevent a current calculated CFO estimates from being used for updated CFO correction parameters transmitted to programmable CFO / phase correction unit 410.

[0088] In some embodiments, controller 450 is configured to compare a current calculated phase estimate with an expected phase estimate range determined, for example, based on a statistical analysis of previous phase estimates. In addition, controller 450 may be configured to prevent a current calculated phase estimates from being used for updated phase correction parameters transmitted to programmable CFO / phase correction unit 410.

[0089] Figure 5 illustrates a flowchart for a method 500 of estimating carrier frequency offset, according to aspects of the present disclosure. Method 500 may be performed, for example, by receiver circuit 100.

[0090] At block 510, the wireless communication system synchronizes to a preamble to obtain an estimate of the phase and frequency offset between the receiver and the received RF signal. The wireless communication system may use any technique for synchronizing. ForDocket No. QID241632 / 62306.204 WOO 1example, the wireless communication system may synchronize to a preamble via correlation against a known preamble at multiple frequency offsets and select the correlator and corresponding frequency with largest magnitude.

[0091] At block 520, the registers of the Kalman filter which implement equations (1) - (3) above, are initialized with the initial phase estimate op,init2 and the initial frequency estimate of, init2 as determined at block 510 based on the preamble (0O|o ■> ^o|o) and the initial accuracy of the estimate P0|0.

[0092] At block 530, the phase and frequency estimates may be refined for each of a number of successively received symbols by computing the innovation or error signal an and updating the estimate according to equations (1) and (2), for example, as discussed above with reference to figure 2. The Kalman gain of equation (2) may also be updated each symbol according to equation (3). In some embodiments, the wireless communication system pre-computes a fixed Kalman gain table in advance, where the Kalman gain table stores Kn up to the maximum packet length (N). In some embodiments, the iterative formula depends on the signal statistics: the observation error variance (oo2), the initial preamble-based phase estimation variance (op,init2), and the initial preamble-based frequency estimation variance (of,init2). In some embodiments, these signal statistics are fixed values. In some embodiments, the Kalman gain table is updated after each received symbol. In some embodiments, the Kalman gain table is updated after each received packet. In some embodiments, the Kalman gain table is updated at other intervals, for example, to tradeoff computational complexity and memory.

[0093] At 540, after the frequency and / or phase estimate is updated, the Kalman filters frequency estimate is initialized for the next packet to the final frequency estimate obtained at the end of the most recent packet (m0|0->mW|W). Accordingly, the CFO / phase correction block of the receiver, such as CFO / phase correction unit 410 of figure 4, is configured to correct for the latest frequency offset and / or phase offset for the next packet.

[0094] In some embodiments, for example, for power and / or performance reasons, frequency and / or phase correction may be alternatively performed in RF analog hardware of the receiver by configuring the carrier frequency synthesizer appropriately. In some embodiments, frequency and / or phase correction may be alternatively performed with a combination of digital and analog techniques. Furthermore, in some embodiments, a following packet arrives on another frequency channel, and the frequency and / or phase estimate may be adjusted for the crystal offset and the frequency channel.Docket No. QID241632 / 62306.204 WOO 1

[0095] At 550, the Kalman gain table for the next packet is calculated. In some embodiments, the Kalman gain table for the next packet is calculated with consideration of the fact that the current Kalman filter already has very accurate frequency and or phase estimate. Accordingly, in some embodiments, the gain values for the Kalman gain table decrease over successive estimates. For coherent estimation systems, while the crystal frequency offset is expected to be roughly the same at the start of each packet, the phase offset is expected to be uncorrelated. Thus, in some embodiments, the following initialization metric is used for subsequent packets:Po|o -> [ Op, ini 0; 0 PN| N,2,2] (4)

[0096] For a next packet, at 550, the gain table (Kn for n=l...N) can then be calculated according to equation (3) using this P0|0 as a starting point. Consequently, that element in the covariance matrix corresponding to the frequency estimation error (denoted PN|N,2,2) is carried over from the previous packet since the frequency estimate refining continues with the next packet.

[0097] In some embodiments, the element corresponding to the estimation error of the phase is not carried over to the next packet because the phase is understood to be substantially uncorrelated from one packet to the next. Instead, the phase estimate is set to be equal to the expected error variance of the preamble-based phase estimate (op,init2) so the phase correction may be re-estimated for the next packet. In some embodiments, this is not a problem because the phase is used for coherent detection of the symbols and is not otherwise used by the NBA-UWB system for ranging.

[0098] Wireless standards are designed to ensure that preambles are dimensioned sufficiently to provide a good-enough phase estimate for typical wireless communication applications. It is the crystal frequency offset accuracy which is important for NBA-UWB ranging and that parameter can indeed be refined by combining information from multiple packets. Note that, in the illustrated example, the diagonal elements (1,2) and (2,1) are equal to 0 in equation (4) under the understanding that the phase and frequency estimation error are sufficiently un-correlated. However, in alternative embodiments, the diagonal elements (1,2) and (2,1) are set to non-zero values for specific implementations.Docket No. QID241632 / 62306.204 WOO 1

[0099] At 560, the preamble of the next packet arrives, and, for coherent applications, the wireless communication system synchronizes to the preamble of the next packet to obtain an estimate of the phase offset between the receiver and the received RF signal. The wireless system already has an accurate estimate of the frequency offset from the previous packet. The wireless communication system may use any technique for synchronizing. For example, the wireless communication system may synchronize to a preamble via correlation against a known preamble at multiple frequency offsets and select the correlator and corresponding frequency with largest magnitude.

[0100] The frequency of the next packet is known quite accurately from the frequency estimate of block 550 obtained in the previous packet, so, in some embodiments, multiple correlations are not performed across the frequency uncertainty. This results in additional power savings and more robust performance.

[0101] At block 530, the Kalman filter continues to refine the frequency estimate using the most recent set of Kalman gain values and the cycle repeats for each packet. In this way, the carrier frequency offset estimate continually improves each ranging round.

[0102] In some embodiments, carrier frequency offset estimate accuracies of up to 1.5 ppb (parts-per-billion) are possible. This not only substantially reduces the search-space (and hence power / complexity) of the UWB synchronization systems but also enables coherent multimillisecond ranging by using large numbers of UWB fragments. This significantly improves the sensitivity and robustness of the system when compared to conventional non-coherent combining. Various embodiments benefit from the operation of the Kalman filter at least because the longer the Kalman filter has to refine the estimate, the less the specific signal statistics used to derive Kalman gains influence the accuracy of the result. Thus, after several ranging rounds, statistics used to initialize the system for the first packet are no longer significant. Therefore, in some embodiments, the signal statistics parameters (oo2, op,init2, of, init2, and the process noise Q) are set to conservative values which will work under worstcase conditions without compromising accuracy. Accordingly, these embodiments are more robust and have superior performance compared to conventional systems using only a single packet.

[0103] In some embodiments, at block 550 of method 500, the predicted estimation accuracy covariance matrix P0|0 is initialized based on the final accuracy of the estimate of the previous packet, which we denoted PN|N,2,2. This would be the optimal value under the assumption that crystal offset remains the same between packet transmissions. However, it is understood that in practice the crystal offset will not be perfectly the same due to, for example,Docket No. QID241632 / 62306.204 WOO 1the following effects: Changes in the crystal temperature of the device can cause frequency drift, Doppler effects due to the motion of devices can cause significant frequency fluctuations between packets even at typical walking speeds, and Very low cost devices may have very unstable references or large amounts of 1 / f noise. To help deal with these effects, some embodiments may modify equation (4) as follows:Po|o-> [ CFp.init20; 0 PN|N,2,2’], where PN|N,2,2’ = max(PN|N,2.2,drift_limit) (4)

[0104] Accordingly, in some embodiments, once the accuracy of the frequency estimate is better than the expected frequency drift between packets due to the various effects outlined above (the “drift limit”), the Kalman gains are calculated according to the drift limit rather than the accuracy of the most recent estimate. The drift limit may be determined based on factors including: how close in time packets are transmitted, how quickly a device is moving, a rate of temperature change, etc. Thus, some implementations may also include higher-level control that monitors stability of the frequency estimates across multiple packets and set the drift limit, and hence the Kalman gain table, to values which are appropriate for the current circumstances. Similarly, some embodiments predict the CFO estimate of the next packet based on a rate of frequency drift determined based on CFO estimates of previous packets, rather than using the CFO estimate of the previous packet.

[0105] Equation (3) depends on a term, Q, called the process noise. In some embodiments, Q = 0. In some embodiments, however, Q may be used to account for non-linearities in the system. In some embodiments, the Kalman filter yields optimal results in the event that the system is more linear. In some embodiments, the phase increases substantially linearly throughout each packet. However, the frequency drift effects listed above can introduce modelling errors which are unaccounted for in the linear estimation process. To account for this, some embodiments set Q to a non-zero value when calculating the gain table. In some embodiments, the value Q is modified based on monitored stability of the frequency estimates over time.

[0106] In some implementations, a doppler shift due to motion between devices may result in “outliers” in the frequency estimate. This may happen, for example, when a device in motion moves through a “null” in the channel response in the middle of a packet, causing a sudden and rapid change in both amplitude and phase. The Kalman filter algorithm may not cope very wellDocket No. QID241632 / 62306.204 WOO 1with this effect because the phase is assumed to increase linearly throughout a packet, and this assumption is invalid when the propagation channel changes too quickly. Thus, in some embodiments, the frequency estimates of devices in motion are monitored, and a current frequency estimate may be set to a previous frequency estimate and / or the Kalman gain tables may be set to previous values when outliers are detected by determining that a next frequency estimate is outside of an expected range. The expected range may be generated based on previous frequency estimates. In some embodiments, a linearity of phase measurements during a packet may be monitored, and a current phase measurement being different from an expected phase measurement range may be used as an indication that the current estimate is an outlier. The expected range may be generated based on previous phase linearity measurements.

[0107] The discussion has emphasized use of a Kalman filter; however, other filters, and other frequency estimators may be used. For example, other forms of frequency and / or phase tracking control loops may be used instead of the Kalman Filter in blocks 240, 430 / 440, and 540 / 550. For example, a phase-locked-loop (PLL) may be used. In some embodiments, a Kalman filter varies the loop gain (Kn) during packets to achieve an estimate with the lowest mean squared error. In contrast, a PLL may use a fixed loop gain which is chosen as a compromise between how quickly the control loop can respond to a frequency error, and how accurate its final estimate is. For better performance, in some embodiments, a PLL algorithm may use an approach referred to as “gear-shifting” which effectively varies the loop gain to a few discrete values. Typically, a large discrete loop gain is used at the start of the packet, so the control loop responds quickly to a large initial error, then a lower value (or values) is gradually introduced as the error gets smaller allowing for a more accurate final result. In some embodiments, a gear-shifting PLL is used as the frequency estimator, and its loop gain is modified using techniques discussed herein as applied to Kalman filter parameters.

[0108] In general, any phase and / or frequency estimation control loop can be used. For example, one or more parameters having the same basic tradeoff: how quickly the control loop can remove a large error and how accurate the final result. This tradeoff may be controlled at each symbol or packet via a gain table. However, systems that use other phase and / or frequency digital tracking-loops can also benefit from multi-packet CFO estimation by using a methodology having aspects similar or identical to the features discussed herein with reference to a Kalman filter. For example, using method 500 of figure 5, block 550 may be modified to vary loop-bandwidth or loop-gain parameters instead of gain parameters. For example, a conventional digital PLL or (gear-shifting PLL) may start with wider loop-bandwidthDocket No. QID241632 / 62306.204 WOO 1parameters for the initial frequency estimation of the preamble-based synchronization. And then, for subsequent packets, the system may program registers of the PLL based on the last frequency estimate at the end of a packet at block 540 and set increasingly narrower loop bandwidth parameters for subsequent packets at block 550.

[0109] Various embodiments are beneficial at least because the systems methods are connection oriented. For example, two wireless communication devices may initially coordinate with each other, for example, about frequency hopping patterns and transmission timing. Therefore, when a receiving device starts receiving at a particular frequency and time from a particular transmitting device, the identity of the particular transmitting device is identified. Accordingly, the receiving device may configure its local CFO compensation unit and relevant control loop parameters to values related to transmission packets previously received from the particular transmitting device. Similarly, when the receiving device starts receiving at a particular frequency and time from a second transmitting device, the identity of the second transmitting device is identified. Accordingly, the receiving device may configure its local CFO compensation unit and relevant control loop parameters to values related to transmission packets previously received from the second transmitting device.

[0110] In some embodiments, a receiver device repeatedly makes ranging measurements with each of a number of transmitting devices. Therefore, some embodiments store phase and / or frequency estimates and associated control loop parameters for each particular transmitting device, based on packets received from the particular transmitting device. The receiver device selects the stored phase and / or frequency estimate and associated control loop parameters for each received transmission packet or symbol according to the identity of the transmitting device. In addition, the receiver device processes the selected phase and / or frequency estimate and associated control loop parameters according to principles discussed elsewhere herein.

[0111] Figure 6 illustrates a flowchart for a method 600 of correcting for a carrier frequency offset of a receiver, according to aspects of the present disclosure. Method 600 may be performed, for example, by receiver circuit 100.

[0112] At block 610, first and second transmitted packets are received, for example, from another wireless communication device.

[0113] At block 620, a set of frequency estimation parameters are updated based on previous values of the set of frequency estimation parameters and based on each of a pluralityDocket No. QID241632 / 62306.204 WOO 1of subsequent symbols of the first and second transmitted packets. For example, the set of frequency estimation parameters may be updated using techniques similar or identical to techniques discussed elsewhere herein.

[0114] At block 630, a frequency estimator may determine an updated frequency offset estimate based on each of the sets of frequency estimation parameters updated at 620. For example, the frequency estimator may determine an updated frequency offset estimate using techniques similar or identical to techniques discussed elsewhere herein.

[0115] At block 640, a carrier frequency offset correction circuit is programmed based on each frequency offset estimate updated at block 630. For example, the carrier frequency offset correction circuit may be programmed using techniques similar or identical to techniques discussed elsewhere herein. In addition, the carrier frequency offset correction circuit is configured to generate frequency offset corrected data based on received data and based on the updated frequency offset estimates.

[0116] Figure 7 illustrates a flowchart for a method 700 of correcting for a carrier frequency offset of a receiver, according to aspects of the present disclosure. Method 700 may be performed, for example, by receiver circuit 100.

[0117] At block 710, a first preamble of a first packet is received, for example, from another wireless communication device.

[0118] At block 720, the wireless communication system synchronizes to the first preamble to obtain an estimate of the phase offset and the frequency offset between the receiver and the received RF signal. The wireless communication system may use any technique for synchronizing.

[0119] At block 730, a communication is received. The communication may, for example, be a symbol of a current packet or a next preamble of a next packet. The received communication may be processed with an offset correction circuit to correct for offset based on current estimates of frequency offset and phase offset, as determined at block 720 or 760, or as updated at block 750.

[0120] At block 740, if the communication is a symbol of the current packet, at block 750, the symbol is processed, and the frequency offset estimate, and the phase offset estimate are updated based on the symbol. If the communication is a preamble of a next packet, at block 760, the wireless communication system synchronizes to the preamble of the next packet to obtain an estimate of the phase offset. In some embodiments, at block 760, the wireless communication system does not determine an estimate of the frequency offset.Docket No. QID241632 / 62306.204 WOO 1

[0121] Figure 8 illustrates a communication device 800 having features similar or identical to receiver circuit 100, and being configured to perform the methods discussed herein, according to aspects of the present disclosure. The communication device 800 can be a communication device such as a mobile terminal, a smart watch, a tablet, a computer, a navigation device, an access point, a base station (e.g., eNB or gNB), and any other type of wireless communication device that supports wireless communications, such as cellular, wireless local area network (WLAN), Bluetooth, ultra-wideband (UWB), and near-field communications. The communication device 800 generally includes a control system 50, a baseband processor 52, transmit circuitry 54, receive circuitry 56, antenna switching circuitry 58, multiple antennas 60, and user interface circuitry 62. In a non-limiting example, the control system 50 can be a field-programmable gate array (FPGA). In this regard, the control system 50 can include one or more of at least a microprocessor, an embedded memory circuit, and a communication bus interface. In some embodiments, the receive circuitry 56 has features and capabilities similar or identical to receiver circuit 100. The receive circuitry 56 receives radio frequency signals via the multiple antennas 60 and through the antenna switching circuitry 58 from one or more base stations. A low-noise amplifier and a filter cooperate to amplify and remove broadband interference from the received signal for processing. Downconversion and digitization circuitry (not shown) will then downconvert the filtered, received signal to an intermediate or baseband frequency signal, which is then digitized into one or more digital streams using one or more analog-to-digital converters (ADCs).

[0122] The baseband processor 52 processes the digitized received signal to extract the information or data bits conveyed in the received signal. This processing typically comprises demodulation, decoding, and error correction operations, as will be discussed in greater detail below. The baseband processor 52 is generally implemented in one or more digital signal processors (DSPs) and application-specific integrated circuits (ASICs).

[0123] For transmission, the baseband processor 52 receives digitized data, which may represent voice, data, or control information, from the control system 50, which it encodes for transmission. The encoded data is output to the transmit circuitry 54, where a digital -to-analog converter (DAC) converts the digitally encoded data into an analog signal and a modulator modulates the analog signal onto a carrier signal that is at a desired transmit frequency or frequencies. A power amplifier amplifies the modulated carrier signal to a level appropriate for transmission and delivers the modulated carrier signal to the multiple antennas 60 through the antenna switching circuitry 58. The multiple antennas 60 and the replicated transmit circuitry 54 and receive circuitry 56 may provide spatial diversity. Modulation and processing detailsDocket No. QID241632 / 62306.204 WOO 1will be understood by those skilled in the art. In an embodiment, the RF receiver front-end 10 may be provided in any one or more of the circuitries in the communication device 800, such as the receive circuitry 56.

[0124] Some embodiments include a system performing a method of estimating the crystal frequency offset between two devices, the system and method including: a compensating hardware / software unit which compensates a current packet’s received signal for the crystal frequency offset between two devices; a hardware / software control loop which iteratively refines the crystal frequency offset between the two devices based on measurements of the phase or frequency error obtained on each data symbol in the packet; a hardware / software control unit which initializes the compensating unit’s compensating frequency with a crystal frequency offset estimate obtained by from a previously received packet; and a hardware / software control unit which initializes the parameters of the control loop to reflect the expected accuracy of this frequency estimate.

[0125] In some embodiments, the control unit stores previous crystal frequency offsets and corresponding accuracies between the local device and more than one other device and sets the compensating unit and the control loop parameters accordingly depending on which device it is communicating with.

[0126] In some embodiments, the control loop includes a Kalman filter, and the parameters of the control loop modified by the control unit are the Kalman gains.

[0127] In some embodiments, the Kalman gain is initialized using elements from a final estimation error covariance matrix of a previous packet.

[0128] In some embodiments, the control loop is a phase or frequency locked loop and the parameters of the control loop modified by the control unit is the loop bandwidth.

[0129] In some embodiments, the control unit stores one or more frequency estimates and corresponding accuracies from previously received packets of one or more devices and ignores frequency estimate outliers from the current packet and device that are significantly different from the expected accuracy for the current device. The control unit sets the compensating unit and control loop according to one or more of its previously stored values instead of the outliers.

[0130] In some embodiments, the control unit observes phase measurements used by the control loop and ignores outlier frequency estimates taken from the current packet when the phase measurements differ significantly from a straight-line estimate. The control unit sets the compensating unit and control loop according to one or more of its previously stored values instead of the outliers.Docket No. QID241632 / 62306.204 WOO 1

[0131] In some embodiments, the control unit predicts new parameters of the crystal frequency offset estimate and / or accuracy based on observed trends in previous estimates and accuracies reflecting current conditions.

[0132] In some embodiments, the current conditions represent a frequency drift, and the control unit estimates new parameters based on a linear combination of previous estimates predicting the current value.

[0133] In some embodiments, the current conditions represent a particular device’s phase noise and frequency characteristics, and the control unit estimates new parameters based on the observed characteristics of this device.

[0134] Persons skilled in the art will recognize that the apparatus, systems, and methods described above can be modified in various ways. Accordingly, persons of ordinary skill in the art will appreciate that the embodiments encompassed by the present disclosure are not limited to the particular exemplary embodiments described above. In that regard, although illustrative embodiments have been shown and described, a wide range of modification, change, and substitution is contemplated in the foregoing disclosure. It is understood that such variations may be made to the foregoing without departing from the scope of the present disclosure. Accordingly, it is appropriate that the appended claims be construed broadly, and in a manner consistent with the present disclosure.

Claims

Docket No. QID241632 / 62306.204 WOO 1CLAIMS1. A method of estimating carrier frequency offset, the method comprising:receiving first and second transmitted packets, the first and second transmitted packets each comprising a plurality of symbols;repeatedly updating a set of frequency estimation parameters based on a set of previous values of the set of frequency estimation parameters, and based on each of the plurality of symbols of the first and second transmitted packets;with a frequency estimator, determining an updated frequency offset estimate based on each updated set of frequency estimation parameters; andprogramming an offset correction circuit based on each updated frequency offset estimate, the offset correction circuit configured to generate frequency offset corrected data based on received data and based on the updated frequency offset estimates.

2. The method of claim 1, further comprising:receiving additional transmitted packets, the additional transmitted packets each comprising a plurality of symbols;repeatedly updating the set of frequency estimation parameters based on a set of previous values of the set of frequency estimation parameters updated based on the symbols of the first and second transmitted packets, and based on each of a plurality of subsequent symbols of the additional transmitted packets; andwith the frequency estimator, determining an updated frequency offset estimate based on each set of frequency estimation parameters updated based on the set of previous values of the set of frequency estimation parameters and based on the subsequent symbols of the additional transmitted packets.

3. The method of claim 1, further comprising generating an error signal based on each symbol of the first and second transmitted packets and based on a corresponding frequency offset estimate, wherein the set of frequency estimation parameters are repeatedly updated based on the error signals.

4. The method of claim 1, wherein the frequency estimation parameters comprise a Kalman filter gain, and wherein the frequency estimator comprises a Kalman Filter.Docket No. QID241632 / 62306.204 WOO 15. The method of claim 1, further comprising synchronizing to first and second preambles of the first and second transmitted packets, respectively, to respectively generate first and second initial phase offset estimates for the first and second transmitted packets, wherein the first transmitted packet is received from a first external device, and wherein the second transmitted packet is received from a second external device.

6. The method of claim 5, further comprising:repeatedly updating a first subset of the frequency estimation parameters based on previous values of the first subset of frequency estimation parameters, based on each of a plurality of subsequent symbols of the first transmitted packet, and based on subsequent symbols of one or more additional packets received from the first external device; and repeatedly updating a second subset of the frequency estimation parameters based on previous values of the second subset of frequency estimation parameters, based on each of a plurality of subsequent symbols of the second transmitted packet, and based on subsequent symbols of one or more additional packets received from the second external device.

7. The method of claim 6, wherein determining the updated frequency offset estimate is based on an updated set of phase estimation parameters.

8. The method of claim 7, wherein the phase estimation parameters comprise a Kalman filter gain.

9. A receiver, comprising:analog circuitry comprising an input and one or more outputs;one or more analog to digital converters (ADCs) coupled to the one or more outputs and comprising one or more digital outputs; anda processor configured to:receive first and second transmitted packets, the first and second transmitted packets each comprising a plurality of symbols;repeatedly update a set of frequency estimation parameters based on a set of previous values of the set of frequency estimation parameters and based on each of a plurality of symbols of the first and second transmitted packets;Docket No. QID241632 / 62306.204 WOO 1determine an updated frequency offset estimate based on each updated set of frequency estimation parameters; andprogram an offset correction circuit based on each updated frequency offset estimate, the offset correction circuit configured to generate frequency offset corrected data based on data received from the ADCs and based on the updated frequency offset estimates.

10. The receiver of claim 9, wherein the processor is further configured to:receive additional transmitted packets, the additional transmitted packets each comprising a plurality of symbols;repeatedly update the set of frequency estimation parameters based on a set of previous values of the set of frequency estimation parameters and based on each of a plurality of subsequent symbols of the additional transmitted packets; anddetermine an updated frequency offset estimate based on each set of frequency estimation parameters updated based on the set of previous values of the set of frequency estimation parameters and based on the subsequent symbols of the additional transmitted packets.

11. The receiver of claim 9, wherein the processor is further configured to generate an error signal based on each symbol of the first and second transmitted packets and based on a corresponding frequency offset estimate, wherein the set of frequency estimation parameters are repeatedly updated based on the error signals.

12. The receiver of claim 9, wherein the frequency estimation parameters comprise a Kalman filter gain.

13. The receiver of claim 9, wherein the processor is further configured to synchronize to first and second preambles of the first and second transmitted packets, respectively, to respectively generate first and second initial phase offset estimates for the first and second transmitted packets, wherein the first transmitted packet is received from a first external device, and wherein the second transmitted packet is received from a second external device.

14. The receiver of claim 13, wherein the processor is further configured to:Docket No. QID241632 / 62306.204 WOO 1repeatedly update a first subset of the frequency estimation parameters based on previous values of the first subset of frequency estimation parameters, based on each of a plurality of subsequent symbols of the first transmitted packet, and based on subsequent symbols of one or more additional packets received from the first external device; and repeatedly update a second subset of the frequency estimation parameters based on previous values of the second subset of frequency estimation parameters, based on each of a plurality of subsequent symbols of the second transmitted packet, and based on subsequent symbols of one or more additional packets received from the second external device.

15. The receiver of claim 14, wherein determining the updated frequency offset estimate is based on an updated set of phase estimation parameters.

16. The receiver of claim 15, wherein the phase estimation parameters comprise a Kalman filter gain.

17. A wireless device, comprising:an antenna input port;an antenna switch coupled to the antenna input port; anda receiver, comprising:analog circuitry comprising an input coupled to the antenna switch and one or more outputs;one or more analog to digital converters (ADCs) coupled to the one or more outputs and comprising one or more digital outputs; anda processor configured to:receive first and second transmitted packets, the first and second transmitted packets each comprising a plurality of symbols;repeatedly update a set of frequency estimation parameters based on a set of previous values of the set of frequency estimation parameters and based on each of a plurality of symbols of the first and second transmitted packets;determine an updated frequency offset estimate based on each updated set of frequency estimation parameters; andprogram an offset correction circuit based on each updated frequency offset estimate, the offset correction circuit configured to generate frequencyDocket No. QID241632 / 62306.204 WOO 1offset corrected data based on data received from the ADCs and based on the updated frequency offset estimates.f8. The wireless device of claim 17, wherein the frequency estimation parameters comprise a Kalman filter gain.

19. The wireless device of claim 17, wherein the processor is further configured to: synchronize to first and second preambles of the first and second transmitted packets, respectively, to respectively generate first and second initial phase offset estimates for the first and second transmitted packets, wherein the first transmitted packet is received from a first external device, and wherein the second transmitted packet is received from a second external device;repeatedly update a first subset of the frequency estimation parameters based on previous values of the first subset of frequency estimation parameters, based on each of a plurality of subsequent symbols of the first transmitted packet, and based on subsequent symbols of one or more additional packets received from the first external device; and repeatedly update a second subset of the frequency estimation parameters based on previous values of the second subset of frequency estimation parameters, based on each of a plurality of subsequent symbols of the second transmitted packet, and based on subsequent symbols of one or more additional packets received from the second external device.

20. The wireless device of claim 19, wherein determining the updated frequency offset estimate is based on an updated set of phase estimation parameters comprising a Kalman filter gain.