Wireless telecommunications base station and process for high mobility scenarios

Abstract

A wireless telecommunications base station that compensates for Doppler shift in each connected user equipment (UE). The base station deploys multiple parallel receivers, each with a given frequency offset above and below a carrier frequency. Each receiver performs a frequency shift on a common uplink signal, determines the quality of the frequency-shifted uplink signal, and demodulates the frequency-shifted uplink signal. A selector / combiner module generates the highest quality demodulated signal, which can be done by selecting the frequency-shifted uplink signal or soft combining a subset of the frequency-shifted uplink signals with sufficiently high quality.

Description

Technical Field

[0001] The present invention relates to a wireless telecommunication base station and related processes for high mobility scenarios, and more particularly, to a process and system for providing a high quality uplink to wireless devices in a high mobility setting. Background Art

[0002] According to complex telecommunication standards, such as LTE and 5G, given the large amount of data packed into a subframe of the RF spectrum, if a given user equipment (UE) moves quickly relative to the base station (eNodeB or gNodeB), the receiver decoding performance is severely degraded by the Doppler effect. This is due to the Doppler shift in the carrier frequency of the transmitting UE. Given enough Doppler shift, interference between the subcarriers of an orthogonal frequency division multiplexing (OFDM) system may occur, and the bit error rate (BER) experienced by the receiver may increase. The increase in BER may cause the base station to instruct the UE to gradually reduce the modulation mode to a lower rate.

[0003] The conventional solution to the Doppler shift problem involves predicting the UE speed to estimate the Doppler shift in the carrier frequency necessary to compensate and apply the appropriate carrier frequency offset. For these purposes, computationally intensive estimation and tracking algorithms are known. However, this approach still has disadvantages. First, in a typical high mobility scenario, different UEs may move at different speeds relative to the receiver. This makes it extremely difficult to estimate and track each UE connected to the receiver. Second, the quality of the link will only be good if the speed of each UE is estimated at a specific estimation moment. If the estimation is turned off, the connection can be severely degraded. Third, in a highly dynamic, high mobility environment, the UE may be subject to sudden Doppler transitions and unpredictable changes. This is particularly common in situations where the UE is located in a nearby vehicle, which may suddenly change speed or pass very quickly close to the receiver antenna.

[0004] Therefore, there is a need for a receiver that is able to account for and reduce Doppler shift in uplink transmissions from multiple UEs, each of which may be moving at different speeds in highly dynamic conditions, and wherein the reduction does not rely on computationally intensive and unreliable estimation and tracking techniques. Summary of the invention

[0005] The main object of the present invention is to provide a receiver capable of taking into account and reducing Doppler shift in uplink transmissions from multiple UEs, each of which may move at different speeds in highly dynamic conditions, wherein the reduction does not rely on computationally intensive and unreliable estimation and tracking techniques.

[0006] Another object of the present invention is to provide a wireless telecommunication procedure for high mobility scenarios.

[0007] The above objects are achieved by providing a parallel receiver architecture and a procedure for high mobility scenarios according to the features of claim 1 .

[0008] One aspect of the present invention relates to a wireless telecommunications base station. The base station includes a coordinator module that determines a plurality of frequency offsets relative to a carrier frequency; a plurality of receivers, wherein each of the plurality of receivers receives a different frequency offset within the plurality of frequency offsets, shifts an input signal at a carrier frequency by a corresponding frequency offset to generate a frequency-shifted signal, and determines a signal quality measurement value corresponding to the frequency-shifted signal thereof; and a selector / combiner module that receives the frequency-shifted signal and the signal quality measurement value from each of the plurality of receivers, generates a frequency-shifted signal of the highest quality, and transmits the frequency-shifted signal of the highest quality to an upper layer protocol stack.

[0009] Another aspect of the invention relates to a wireless telecommunications base station. The base station comprises a coordinator device for determining a plurality of frequency offsets relative to a carrier frequency, a plurality of receiver devices for receiving corresponding frequency shifts, receiving a signal, applying a corresponding frequency offset to the signal to generate a frequency shifted signal, and calculating a signal quality measurement corresponding to the frequency shifted signal, and a selector / combiner device for receiving the frequency shifted signal and the signal quality measurement from each of the plurality of receiver devices and generating a frequency shifted uplink signal of the highest quality.

[0010] Another aspect of the invention relates to a non-transitory computer readable memory encoded with instructions that, when executed by one or more processors, cause the one or more processors to perform a process. The process includes determining a plurality of frequency offsets; assigning each of the plurality of frequency offsets to a receiver among a plurality of receivers; shifting an input signal by each of the plurality of frequency offsets to generate a plurality of frequency shifted signals; calculating a signal quality measurement corresponding to each of the plurality of frequency shifted signals; and calculating an output signal based on the plurality of frequency shifted signals and the plurality of corresponding quality measurements. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] Other features and advantages of the present invention will be apparent from the description of preferred but not exclusive embodiments of a parallel receiver architecture and procedures for high mobility scenarios, which are illustrated in detail by way of example and not limitation in the accompanying drawings, in which:

[0012] Figure 1 shows a scenario where a base station is connected to multiple UEs that experience independent Doppler shifts;

[0013] Figure 2An exemplary system for reducing Doppler shift of multiple UEs is shown;

[0014] Figure 3 An exemplary receiver for processing uplink signals at multiple frequency offsets is shown;

[0015] Figure 4a An exemplary carrier frequency capture range without frequency offset is shown;

[0016] Figure 4b shows a number of exemplary frequency capture ranges for five receivers operating simultaneously at different frequency offsets;

[0017] Figure 4c Shows the Figure 4b Scenario Tighter Frequency Offset Multiple Exemplary Capture Ranges for Seven Receivers Operating Simultaneously;

[0018] Figure 5a An exemplary process for instantiating and de-instantiating a set of required receivers according to the present disclosure based on predetermined communication traffic conditions is shown;

[0019] Figure 5b An exemplary process for dynamically instantiating and de-instantiating receivers based on actual communication traffic conditions is shown. DETAILED DESCRIPTION

[0020] Figure 1 An exemplary base station deployment 100 is shown in which a wireless telecommunication base station 105 (e.g., an eNodeB) is connected to a plurality of UEs. As shown, UEs 130a, 130b, 130c, and 130d are located on a high-speed train 125 that is approaching and moving past the eNodeB 105 at a high speed. At the instant shown, each of the UEs 130a-d experiences a different Doppler shift relative to the carrier frequency of the eNodeB 105. In this example, at the instant shown, UE 130a experiences a large negative Doppler shift, and UE 130d experiences a large positive shift, as UE 130a and UE 130d move away from and toward the eNodeB 105, respectively. UE 130b and UE 130c experience a negative Doppler shift and a positive Doppler shift, respectively, but UE 130a and UE 130d experience shifts of smaller magnitudes.

[0021] Further shown is vehicle traffic, which involves several vehicles 135, each of which may have one or more UEs 140a, 140b and 140c. One or more of the UEs 140a-c may be integrated with the vehicle itself, or may be a user's mobile handheld terminal. One difference between the vehicle 135 and the train 125 is that the train 125 may move at a constant or near-constant speed, while each of the vehicles 135 may suddenly accelerate nonlinearly, which is common in urban traffic settings. The significance of the chaotic speed of the vehicles 135 is that, although the above-mentioned speeds and Doppler shifts still have many disadvantages, it is impossible to estimate the speed of a given vehicle 135 at a given time, while it is possible to estimate the speed of a given train 125 at a given time with reasonable accuracy. Traditional methods of channel modeling are defined in 3GPP technical specifications TS 36.101 and TS36.104.

[0022] like Figure 1 As shown, the eNodeB 105 may have a baseband processor 110 that performs radio access network (RAN) functions of a given standard, such as LTE. Further, the eNodeB 105 may be coupled to the Internet via a backhaul connection 120. Further, although the term eNodeB may be used in this example, it will be understood that the same description may also be used for 5GgNodeB.

[0023] An important aspect of the base station deployment 100 is that many UEs may be designed, each of which may move at different speeds relative to the eNodeB 105, and these speeds may vary independently and randomly.

[0024] Figure 2 An exemplary system 200 for reducing Doppler shift according to the present disclosure is shown. The system 200 can be basically deployed in Figure 1The system 200 may include an antenna 205 coupled to a low noise amplifier (LNA) 210. The LNA 210 may be coupled to a mixer 215 that mixes an amplified RF signal of a carrier frequency fc from the antenna 205 with an analog baseband signal 225 using a local oscillator (not shown). The combination of the LNA 210 and the mixer 215 may be referred to as an RF front end. The system 200 also includes a plurality of receivers 220, each of which is fed with an analog baseband signal 225. Each receiver 220 shifts the frequency of the baseband signal 225 by a specific frequency offset 235, which is provided to each receiver 220 by a coordinator module 230. The frequency offset 225 for one of the receivers 220 may be zero, and the frequency shift 225 may be increasingly negative and increasingly positive for other receivers 220. Each frequency offset 235 may correspond to a Doppler shift in the carrier frequency, thereby causing the UE to move relative to the eNodeB 105. This will be described further below.

[0025] Each receiver 220 performs a PHY layer function on its frequency-shifted baseband signal 225 and outputs a corresponding frequency-shifted signal 240. In particular, each frequency-shifted signal 240 is a UE signal set, which may include a physical uplink control channel (PUCCH), a physical uplink shared channel (PUSCH), a physical random access channel (PRACH), and a sounding reference signal (SRS) for each connected UE in the base station deployment 100. Each UE signal set 240 corresponds to a UE uplink (UL) signal received at a given Doppler offset of the carrier frequency and may include data indicating the quality of the signal received from the corresponding UE at the given Doppler frequency offset, which will be further described below. Each receiver 220 may be substantially similar and may be implemented as a pure software-based virtual subsystem that runs on general hardware, or they may be implemented in specific hardware or a mixture of dedicated hardware and software. It will be understood that such variations are possible and fall within the scope of the present disclosure. The specific functions of the receiver 220 will be described in more detail below.

[0026] The selector / combiner module 245 receives the UE signal set 240 from each receiver 220 and selects the signal set 240 with the highest signal quality (i.e., compensated for Doppler shift), or soft combines the respective PUCCH, PUSCH, and SRS from the receivers to calculate the best solution or a combination of two best solutions. In this case, the selector / combiner module 245 can populate multiple data arrays, one for each UE, with selection data from each UE signal set as a function of frequency offset. Due to this data, as described above, the selector / combiner module 245 can select the signal from the receiver 220 with the strongest signal, or combine some or all of the corresponding signals from a given UE associated with the frequency offset to calculate a soft combined signal. This point will be described in more detail below.

[0027] The selector / combiner module 245 outputs a single signal set 247 for each UE to the upper protocol layer 250 for further processing according to the appropriate communication protocol.

[0028] In particular, the single signal set 247 is the highest quality frequency shifted signal.

[0029] The generated UL signal is then delivered to the appropriate network operator core network 260 via the Internet 255 according to a given standard.

[0030] Figure 3 A system 200 is shown along with a plurality of exemplary receivers 220 according to the present disclosure. As shown, there are N receivers 220, each of which has a different frequency offset (including a frequency offset of a zero offset). As described above, each receiver 220 receives a baseband signal 225 from a mixer 215. The component blocks 305-355 within the receiver 220 can be viewed as functional and component modules. In the previous case, Figure 3The partitions within the receiver 220 in may be variations. In the case of a pure virtual software implementation of the receiver 220, the functions of the components 305-355 may be implemented in multiple threads running on one or more processor cores within one or more server boards of the baseband processor 110. In this case, the functions of the blocks 305-355 are encapsulated in a collection of machine-readable instructions that are stored on a non-transient memory and executed on computer hardware. As used herein, the term "non-transient memory" may refer to any tangible storage medium (as opposed to an electromagnetic or optical signal), and refers to the medium itself, but is not limited to data storage (e.g., RAM and ROM). For example, the non-transient memory may refer to an embedded memory that is encoded with instructions, whereby the memory may have to be reloaded with appropriate machine-readable instructions after a restart. Therefore, the receiver 220 may be described as a series of functional blocks or similarly described as a series of steps performed by one or more processors (hereinafter referred to as processors) within the computing environment of the baseband processor 110.

[0031] In block 305, the processor executes instructions to convert the analog baseband signal 225 into a digital signal. The A / D hard disk on the corresponding server board can be used to make the processor perform this. Optionally, the A / D function of block 305 can be performed in the mixer module 215 or between the mixer module 215 and the receiver 220, in which case the baseband signal 225 can be a digital baseband signal. Further with respect to block 305, the processor removes the cyclic prefix from the digital baseband signal according to the telecommunications standard. Therefore, the output of block 305 is a series of digitized symbols, each of which may have, for example, 2048 samples. The in-phase quadrature (I / Q) data of these 2048 samples represents the time domain signal received by the antenna 205.

[0032] In block 310, the processor may execute instructions to convert the digitized I / Q data to a 32-bit floating point representation. Doing so greatly increases the dynamic range of further processing in the processing chain of the receiver 220. Further, in the case of a pure software virtual baseband processor, converting the data to a 32-bit floating point representation has the additional advantage that the conversion to the floating point representation and subsequent processing can be completed by a floating point calculation engine in the server processor hardware. This may result in higher precision high-speed processing that can be dynamically provided between servers in, for example, a cloud computing environment. Although there are advantages to converting the digitized I / Q data to a floating point format, it will be understood that this step is optional and that the disclosed steps may also be performed on binary integer data.

[0033] In block 315, the processor executes instructions to frequency shift the 32-bit I / Q baseband signal corresponding to the frequency offset 235 provided by the coordinator module 230. This can be done as follows. The input 32-bit I / Q baseband data can be represented as r(k), where k is a discrete time index corresponding to a given I / Q baseband sample. In block 315, the processor can apply a frequency shift to the input I / Q baseband samples as follows:

[0034]

[0035] Among them, f n is the frequency offset 235 provided by the coordinator 230 to the nth receiver 220, and f s is the sampling frequency. Therefore, each receiver 220 calculates a series of discrete frequency-shifted baseband I / Q data, each of which is located at a different specified Doppler frequency offset. It will be understood that specific code implementations of the above equations are possible and fall within the scope of the present disclosure.

[0036] In block 320, the processor executes instructions to perform a fast Fourier transform (FFT) on a block of 2048 samples of the digitized frequency-shifted time-domain baseband I / Q data to convert the digitized frequency-shifted time-domain baseband I / Q data into digital frequency-time-domain data. Block 320 may be performed using conventional procedures in accordance with the 3GPP specification. As each subsequent set of 2048 time-domain samples is processed successively, the processor reconstructs a data frame including resource elements (REs).

[0037] In block 325, the processor executes instructions to extract the relevant designated REs from the frequency domain data about each connected UE. In this case, data about the physical uplink control channel (PUCCH), physical uplink shared channel (PUSCH) and sounding reference signal (SRS) of each UE are extracted for further baseband processing. For the purpose of this disclosure, the processing of PUSCH (in the context of LTE implementation) will be further described.

[0038] Block 330 illustrates processing of the PUSCH for a single UE in accordance with the present disclosure. As shown, there are M parallel processing blocks 330, each of the M parallel processing blocks corresponding to one of the M UEs. It will be appreciated that other processing in accordance with the 3GPP specifications is performed for various other channels (e.g., PUCCH, PRACH, etc.) for each of the UEs, and, although not shown or described herein, is part of the processing implemented by the baseband processor 110.

[0039] In block 335, the processor executes instructions to combine REs corresponding to the PUSCH of a particular UE and perform channel estimation corresponding to each UE data to generate regular parameter reports, such as channel state information (CSI), which includes a channel quality indicator (CQI), etc., as specified in 3GPP TS 36.213, and block 335 reports the regular parameter reports to the demodulation block 340 and the SNR estimation block 342, as will be further described below.

[0040] In block 340, the processor executes instructions to combine REs corresponding to the PUSCH of a particular UE and demodulate the combined 32-bit floating point I / Q data. Signal demodulation from the UE can be implemented using conventional procedures. However, more accurate (and therefore more robust) demodulation schemes can be provided using 32-bit floating point data. This is particularly common for high-order modulation schemes (e.g., 16-QAM, 64-QAM, and 256-QAM), whereby the increased dynamic range and improved precision of 32-bit floating point I / Q operations (with 16-bit integers) may reduce the bit error rate in these modulation schemes. The result of block 340 is a codeword with a "primary" cyclic redundancy check (CRC) bit set.

[0041] In block 342, the processor executes instructions to determine the quality of the UL signal from a given UE at a specific Doppler frequency offset. An example of quality assessment is to calculate the signal-to-noise ratio (SNR) of the demodulated digital signal in step 340. The SNR can be calculated using maximum likelihood estimation techniques and minimum mean square error equalization (MMSE). Both techniques are known in the art, and it will be understood that various implementations of these techniques are possible and fall within the scope of the present disclosure. In an example, the processor executes instructions to calculate the SNR using a data-assisted maximum likelihood algorithm on a frequency domain demodulation reference signal (DM-RS) present in the UL I / Q data from a given UE. The result of step 342 is an SNR estimate 344, which is then passed to the selector / combiner 245 in the UL signal set data 240 together with the corresponding reassembled packet data, which is extracted from the same resource block of the I / Q data as the DM-RS. As described above, the SNR corresponds to the quality of the UL signal received given the Doppler frequency compensation (frequency offset 235 provided to the receiver 220 by the coordinator 230).

[0042] Although the channel estimation block 335 and the SNR estimation block 342 are shown as distinct blocks, they may operate simultaneously on the PUSCH data combined by block 335. It will be appreciated that such variations are possible and fall within the scope of the present disclosure.

[0043] In block 345, the processor executes instructions to perform turbo decoding according to the present disclosure. This includes deinterleaving and descrambling demodulated code words to form a standard process of a code block set. The length of the code block in bytes, together with the coding rate, is a function of the modulation method adopted by the corresponding UE when transmitting. The detailed information of these blocks can be found in 3GPP TS 36.212. Block 345 continues the turbo decoding function, thereby generating a reconstructed and decoded code block. The processor performs a code block CRC in block 350, and performs a "main" CRC in block 355 and reassembles the transport block of the demodulated packetized data transmitted by the UE. The processor then sends the demodulated UE signal set 240 to the selector / combiner. This may include information about the number of CRC failures experienced in block 345.

[0044] The selector / combiner 245 receives the set of UE signals 240, each of which is located at each Doppler frequency offset (i.e., each block 330 from each receiver 220) along with a corresponding SNR estimate 344 (calculated in step 342), and may include the number of CRC failures determined in block 345. For the selector / combiner 245, the processor executes instructions to select the set of UE signals 240 from the receiver 220 on a UE-by-UE basis, which receiver 220 has the highest signal quality based on the SNR estimate 344 and the results of the CRC performed in each of the corresponding blocks 345 for a given UE. Optionally, the selector / combiner 245 may select all of the input UE signal sets 240 and soft combine the data to provide a high fidelity representation of the communicated UL data, which may include using the SNR estimate 344 as a weight to perform a weighted sum. Further turning to this alternative, only a subset of the input UE signal sets 240 may be summed in a soft combined form. In this case, the processor may execute instructions to select only those PUCCH data and PUSCH data (from the UE signal set 240) whose corresponding SNR estimates 344 are above a certain threshold. It will be apparent that such variations are significant and fall within the scope of the present invention.

[0045] The output of the selector / combiner 245 is individual Doppler compensated PUCCH and PUSCH data for each UE, which is input to the remaining upper protocol layers 250. The upper protocol layers 250 may be based on a standard 3GPP implementation, such as LTE, whereby the upper protocol layers 250 do not need to be aware of the presence of Doppler compensation to be performed in the receiver 220 and the selector / combiner 245, or that there are many receivers 220. The selector / combiner 245 may also provide state information to the coordinator 230, which may include all or some of the calculated SNR estimates 340, along with the soft combined sum of the selected receiver 220 or a subset of receivers 220, as a function of the UE and / or frequency offset.

[0046] Figure 4a An exemplary spectral capture range window 400 for an incoming OFDMA or SC-FDMA signal is shown. The x-axis corresponds to frequency, with frequency f c The y-axis is the number of successfully decoded input signals greater than and less than the carrier frequency f c 4. The frequency range 412 of FIG. 4 corresponds to the probability of successfully decoding a high modulation rate (e.g., 256-QAM) UL signal. This can be described as a capture probability curve 405 that spans a frequency capture range CR. As shown, there is a high probability band 410 having a frequency range 412 around the carrier frequency, whereby the high probability band 410 corresponds to a 95% probability of successfully decoding a high modulation rate (e.g., 256-QAM) UL signal. Exemplary frequency range 412 values ​​can be between 450 Hz and 500 Hz. The width of the CR, and thus the width of the frequency range 412, depends on the modulation scheme, and thus, the higher the modulation, the narrower the CR.

[0047] Figure 4b An exemplary set of spectral capture curves for a plurality of signals, each of which has a different frequency offset, is shown. As shown, the capture probability curve 405 has a zero offset and thus is centered at f c Centered (with Figure 4a The capture probability curve 415a has f n Frequency offset, f c +f n is centered and has a capture range CR 1a ; Capture probability curve 415b has -f n Frequency offset, f c -f n is centered and has a capture range CR 1b ; Capture probability curve 425a has 2f n Frequency offset, f c +2f nis centered and has a capture range CR 2a ; and the capture probability curve 425b has -2f n Frequency offset, f c -2f n is centered and has a capture range CR 2b Further illustrated by the overall capture probability curve 440, this depicts the probability of decoding a signal regardless of which frequency offset is used. As shown, the overall capture probability curve 440 shares the same y-axis probability scale as the individual capture probability curves 405, 415a / b, and 425a / b.

[0048] exist Figure 4b In the background, the functions of the disclosed system are described as follows. Figure 1 , a UE 130d located at the rear of a high-speed train 125 passing an eNodeB 105. In this scenario, the UE 130d approaches the eNodeB quickly, thereby generating a positive Doppler shift 450a ( Figure 4a The Doppler shift is in the high probability band 430a of the probability curve 425a. In this case, the receiver 220 has a 2f frequency band due to the highest probability of decoding the UL signal from the UE 130d. n Frequency offset 235 (provided by coordinator 230). As high speed train 125 passes eNodeB 105, the Doppler shift of UE 130d will quickly decrease towards zero. At some point in the method, the Doppler shift of UE 130d can be described by frequency offset 450b. In this case (Doppler shift at 450b), although still represented by a frequency offset of 2f n (and probability curve 425a) has a limited probability of successful decoding by receiver 220, but there is a higher probability that the receiver 220 with frequency offset f n 4. The receiver 220 with the highest quality estimate 344 (and probability curve 415a) will successfully decode the incoming signal and generate the highest quality estimate 344. The combined probability of successful decoding is shown by the total acquisition probability curve 440. As the UE 130d passes the eNodeB 105 and begins to move away from the eNodeB 105, its Doppler shift will become increasingly negative. Two negative Doppler shifts are shown as frequency offsets 450c and 450d. Although this example refers to four frequency offsets 450a-d, it will be understood that the frequency offsets are actually swept from a maximum positive offset to a maximum negative offset, and that the frequency offsets 450a-d are exemplary snapshot offsets.

[0049] Figure 4cAnother exemplary set of spectral capture ranges involving seven parallel receivers 220 is shown, six of which are operated with frequency offsets such that high probability frequency bands corresponding to the frequency offsets are substantially adjacent to each other. In this example, capture probability curve 405 corresponds to a zero offset center frequency and has a high probability frequency band frequency range 412. Capture probability curves 415a and 415b adjacent to capture probability curve 405 have center frequencies f c +f n and f c -f n . Frequency offset f n The capture probability curves 425a and 425b are adjacent to the capture probability curves 415a and 415b, respectively, and each of the capture probability curves 425a and 425b has a center frequency f c +2f n and f c -2f n Further, capture probability curves 435a and 435b are adjacent to the capture probability curves 425a and 425b, respectively, and each of the capture probability curves 435a and 435b has a center frequency f c +3f n and f c -3f n .

[0050] Figure 4c The advantage of the configuration is that the high probability coverage area can be substantially continuous over the expected full range of Doppler frequency offsets. This advantage requires the addition of two parallel receivers 220 to provide a center frequency f c Six different Doppler offsets are provided. It will be appreciated that more parallel receivers 220 may therefore be deployed, thereby further broadening the possible range of Doppler shift compensation.

[0051] An advantage of the disclosed baseband processor 110 is that a given UE can maintain high data rate modulation schemes (eg, 64-QAM or 256-QAM) because the UEs (such as UE 130d) traverse the eNodeB 105 at high speed, making it transparent to upper protocol layers.

[0052] It will be appreciated that these variations on the frequency offset are possible and fall within the scope of the present disclosure. For example, depending on the desired maximum frequency of the Doppler shift, corresponding to the maximum speed of the connected UEs associated with the eNodeB 105, more or less frequency offsets f n (and hence the receiver 220) is possible. Further, the frequency offset fn If spaced closer together, the total acquisition probability curve 440 will appear "flatter" and it is more likely that the UE will maintain a very high rate modulation scheme (e.g., 64-QAM or 256-QAM) over the entire range of Doppler shifts, but this may require more frequency offsets and thus more receivers 220, which may increase the demand on the computing environment of the baseband processor 110. It will be further understood that Figure 4a and 4b The curves shown in FIG. 5 are not drawn to scale but are for illustration purposes.

[0053] The process according to the present invention provides a high quality uplink connection to a wireless device in a high mobility setting.

[0054] In particular, according to a preferred embodiment, the non-transitory computer readable memory of the present invention is encoded with instructions that, when executed by one or more processors, cause the one or more processors to perform a process. The process steps may fall within the scope of the coordinator 230 and may be encoded into machine readable instructions that are stored in the non-transitory memory and executed on one or more processors associated with the baseband processor 110.

[0055] The process according to the invention comprises at least the following steps: determining a plurality of frequency offsets;

[0056] assigning each of the plurality of frequency offsets to a receiver among the plurality of receivers 220;

[0057] shifting an input signal by each of a plurality of frequency offsets to generate a plurality of frequency shifted signals 240;

[0058] calculating a signal quality measurement corresponding to each of the plurality of frequency shifted signals 240; and

[0059] An output signal is calculated based on the plurality of frequency shifted signals 240 and the plurality of corresponding quality measurements.

[0060] According to a possible embodiment, the step of generating an output signal comprises selecting the frequency shifted signal 247 of highest quality from the plurality of frequency shifted signals 240 .

[0061] According to yet another possible embodiment, the step of generating an output signal comprises soft combining a subset of the plurality of frequency shifted signals 240 having highly corresponding signal quality measures.

[0062] Furthermore, the step of calculating a signal quality measurement corresponding to each of the plurality of frequency shifted signals 240 includes calculating a signal-to-noise ratio (SNR) corresponding to the frequency shifted signal 240 .

[0063] The SNR can be calculated using the demodulated reference signal in its frequency shifted signal 240 .

[0064] According to a preferred embodiment, the SNR is calculated using a data-assisted maximum likelihood algorithm.

[0065] Additionally, the step of determining a plurality of frequency offsets includes: receiving status data corresponding to one or more recent signal quality measurements;

[0066] It is optional to identify a first receiver corresponding to a positive frequency shift of the highest magnitude and a second receiver corresponding to a negative frequency shift of the highest magnitude; and

[0067] De-instantiate the first receiver and the second receiver.

[0068] Optionally, the step of determining a plurality of frequency offsets comprises: receiving status data corresponding to one or more recent signal quality measurements;

[0069] Determining that an additional larger positive frequency offset receiver and an additional larger negative frequency offset receiver are necessary; and instantiating the additional larger positive frequency offset receiver and the additional larger negative frequency offset receiver.

[0070] Figure 5a An exemplary process 500 for instantiating and de-instantiating a set of required receivers 220 according to the present disclosure based on predetermined communication traffic conditions is shown. The steps of this process 500 may fall within the scope of the coordinator 230 and may be encoded into machine-readable instructions that are stored in a non-transitory memory and executed on one or more processors associated with the baseband processor 110. For ease of discussion, it may be assumed that the high-speed train 125 is significantly faster than vehicle traffic and may experience a larger maximum (positive and negative) Doppler shift than a UE on a car or a UE integrated with a car.

[0071] In step 505, the processor executes instructions to query one or more transportation plans in order to identify the time of expected high-speed transportation (eg, high-speed train 125). This information may include the expected speed of the high-speed transportation.

[0072] In step 510, the processor calculates the desired Doppler spread to accommodate high speed traffic. In the case of an expected high speed train 125, this may include the expected speed of the high speed train 125, the coverage (in terms of distance) of the cell associated with the eNodeB 105, and the distance between the eNodeB 105 and the closest point of the track that the high speed train 125 will approach. Given these factors, the processor calculates the desired Doppler spread.

[0073] In step 515, the processor calculates the number of receivers 220 required and the frequency offset 235 to be applied to each of the receivers 220. Figure 4b As discussed, closer spacing between frequency offsets 235 may result in more robust performance, but may require more computing resources. With respect to step 515, given the available computing resources and the desired desired Doppler spread, more or fewer receivers 220 may be instantiated. Further, depending on the range of vehicle traffic 135, there may already be a currently operating set of receivers 220 operating at a given set of frequency offsets 235. In this case, additional receivers 220 may not be required, particularly if the desired Doppler spread is expected to be less than or equal to the Doppler spread already processed by the current set of receivers 220.

[0074] In step 520, the processor executes instructions to instantiate the required receiver 220 and provide the required frequency offset 235. Based on the planned activities, it was determined in step 505 that the current and expected required Doppler spread is less than the current capacity of the receiver 220. In this case, the processor may instantiate receivers at both the high and low ends of the frequency offset 235 range. Figure 4b In the example of FIG. 1 , the receiver 220 may no longer need to process the offset f c +2f n and f c -2f n , in this case, these receivers 220 are de-instantiated.

[0075] Figure 5a The exemplary process 500 is applicable to situations where the expected traffic conditions are known. However, there may be an advantage that the baseband processor 110 can dynamically respond to changes in traffic conditions and handle unexpected changes in high-speed traffic.

[0076] Figure 5b An exemplary process 550 is shown whereby the baseband processor 110 dynamically responds to changes in traffic conditions and handles unexpected changes in high-speed traffic. The process 550 may be a separate process that runs separately from the process 500 or may be a sub-process within the process 500.

[0077] In process 550, the processor executes instructions to determine trends and patterns of high magnitude Doppler shifts. To this end, the selector / combiner 245 (or coordinator 230) may store information about the use of the receivers 220, including the SNR estimates 344 corresponding to each of the UEs and each of the receivers 220, and select the receivers 220 for relaying the PUCCH data and PUSCH data of their respective UEs to the upper layer protocol layer 245. Doing so enables the processor to identify (a) short-term Doppler spread patterns for immediate action and (b) long-term trend information to identify in advance when high Doppler spread is needed as a "look-ahead" function.

[0078] The following describes an example of the former case (a). Figure 4b , if the SNR estimate data (and thus the receiver selection) indicates that there is a probability curve 425 a / b The corresponding capture range has very few (or no) instances of Doppler compensation, which may correspond to a situation with very little high-speed traffic. In this case, process 550 may proceed to step 570 via logic condition 560, where the processor executes instructions to instantiate the frequency offset f c +2f n and f c -2f n The corresponding receiver 220.

[0079] Another example of the previous case (a) is described below. Figure 4b , the SNR estimate data (and thus the receiver selection) may indicate an instance of a rapid sweep from the maximum positive frequency (probability curve 425a) to the zero offset (probability curve 405) to the maximum negative frequency (probability curve 425b), whereby the duration at the extreme end of the Doppler offset (425a / b) is not much longer than the duration in the corresponding probability curve for less offset (e.g., 415a / b). This may indicate that coverage at the higher end of the desired Doppler compensation is insufficient, and that additional receivers 220 with greater frequency offsets may be required, in other words, adding one or more pairs of receivers at the far end of the Doppler compensation. In this case, the process 550 may proceed to step 560 via logic condition 560, where the processor executes instructions to instantiate the receiver 220 with the frequency offset f c +3f n and f c -3f n A corresponding additional receiver 220 is provided.

[0080] With respect to the next case (b), the processor may store longer term historical data related to the SNR estimates, the number of receivers 220, the selected receiver 220, and the timestamp. In this scenario, the processor may execute instructions to determine patterns that occur on a daily (e.g., train passing, traffic jams, etc.) and weekly basis. This may include determining patterns that may indicate an upcoming need for additional or fewer receivers. Such "look-ahead" algorithms and their implementation are known and fall within the scope of the present disclosure.

Claims

1. A wireless telecommunication base station (105) for high mobility scenarios, characterized in that include: a coordinator module (230) configured to determine a plurality of frequency offsets relative to a carrier frequency; a plurality of receivers (220), wherein each of the plurality of receivers is configured to receive a different frequency offset (235) within the plurality of frequency offsets, shift an input signal at a carrier frequency by its corresponding different frequency offset (235) to generate a frequency shifted signal (240), and determine a signal quality measurement corresponding to its frequency shifted signal (240); as well as a selector / combiner module (245) configured to receive the frequency shifted signal (240) and the signal quality measurement from each of the plurality of receivers (220), generate a highest quality frequency shifted signal (247), and send the highest quality frequency shifted signal (247) to an upper protocol layer (250); The selector / combiner module (245) is configured to generate the highest quality frequency shifted signal (247) by selecting the frequency shifted signal (240) corresponding to the highest signal quality measurement value; and / or The selector / combiner module (245) is configured to generate the highest quality frequency shifted signal (247) by soft combining a subset of the plurality of frequency shifted signals (240) having highly corresponding signal quality measurements.

2. The wireless telecommunication base station (105) according to claim 1, characterized in that Each of the plurality of receivers (220) is configured to determine the signal quality measurement value by calculating a signal-to-noise ratio (SNR) corresponding to its frequency-shifted signal (240).

3. The wireless telecommunication base station (105) according to claim 1, characterized in that Each of the plurality of receivers (220) is configured to calculate a signal-to-noise ratio (SNR) using a demodulation reference signal DM-RS within its frequency-shifted signal (240).

4. The wireless telecommunication base station (105) according to claim 3, characterized in that , each of the multiple receivers (220) is configured to calculate the SNR using a data-assisted maximum likelihood algorithm.

5. The wireless telecommunication base station (105) according to claim 1, characterized in that , each of the plurality of receivers (220) is configured to implement a portion of a PHY layer processing of an LTE protocol stack.

6. The wireless telecommunication base station (105) of claim 1, characterized in that , the coordinator module (230) is configured to receive status data from the selector / combiner module (245) and to de-instantiate the first receiver (220) and the second receiver (220).

7. A wireless telecommunication base station (105) according to any one of claims 1 to 6, characterised in that , the coordinator module (230) is configured to receive status data from the selector / combiner module (245), determine that an additional large positive frequency offset receiver and an additional large negative frequency offset receiver are needed, and instantiate the additional large positive frequency offset receiver and the additional large negative frequency offset receiver.

8. A wireless telecommunication method for high mobility scenarios, characterized in that , including at least the following steps: determining a plurality of frequency offsets; assigning each frequency offset of the plurality of frequency offsets to a receiver of a plurality of receivers (220); frequency shifting an input signal by each of the plurality of frequency offsets to generate a plurality of frequency shifted signals (240); calculating a signal quality measurement corresponding to each frequency shifted signal in the plurality of frequency shifted signals (240); and calculating an output signal based on the plurality of frequency shifted signals (240) and the plurality of corresponding quality measurements; The step of generating an output signal comprises selecting a frequency shifted signal (247) of highest quality from the plurality of frequency shifted signals (240); and / or The step of generating an output signal includes soft combining a subset of the plurality of frequency shifted signals (240) having high signal quality measures.

9. The method according to claim 8, characterized in that , the step of calculating the signal quality measurement value corresponding to each frequency shifted signal of the plurality of frequency shifted signals (240) comprises calculating a signal-to-noise ratio (SNR) corresponding to the frequency shifted signal (240).

10. The method according to claim 8, characterized in that The step of calculating the signal quality measurement value corresponding to each frequency shift signal in the plurality of frequency shift signals (240) includes calculating a signal-to-noise ratio (SNR) using a demodulated reference signal in the frequency shift signal (240).

11. The method according to claim 10, characterized in that , the step of calculating the signal quality measurement value corresponding to each frequency shifted signal in the plurality of frequency shifted signals (240) includes calculating the SNR using a data-assisted maximum likelihood algorithm.

12. The method according to claim 8, characterized in that , the step of determining multiple frequency offsets includes: receiving status data corresponding to one or more recent signal quality measurements; It is optional to identify a first receiver corresponding to the highest magnitude positive frequency shift and a second receiver corresponding to the highest magnitude negative frequency shift; and De-instantiate the first receiver and the second receiver.

13. The method according to claim 8, characterized in that , the step of determining multiple frequency offsets includes: receiving status data corresponding to one or more recent signal quality measurements; determining a need for an additional large positive frequency offset receiver and an additional large negative frequency offset receiver; And instantiate the additional large positive frequency offset receiver and the additional large negative frequency offset receiver.

Images