Radio synchronization

Abstract

A radio receiver device is configured to determine whether a received digital radio signal contains a predetermined cyclic preamble having a period. An input section is configured to sample the received digital radio signal and generate a plurality of samples, which are stored in a buffer. A first autocorrelator is configured to correlate a first and a second subset of the plurality of samples to generate a first correlation measure, the second subset being stored in the buffer an even multiple of half the preamble period earlier than the first subset. A second autocorrelator is configured to correlate a first and a third subset of the plurality of samples to generate a second correlation measure, the third subset being stored in the buffer an odd multiple of half the preamble period earlier than the first subset. A processing section is configured to compute a difference between the correlation measures and determine that the radio signal contains the predetermined cyclic preamble when the difference is greater than a threshold.

Description

Technical Field

[0001] This invention relates to a radio receiver, and more particularly, but not exclusively, to a... A low-power long-range receiver that provides an improved ability to detect the synchronization portion of the received data signal. Background Technology

[0002] Many modern electronic devices include one or more radio transceivers, receivers, or transmitters for wireless communication with other devices. Numerous wireless communication protocols known in the art for performing such communication exist, including… blue and

[0003] According to at least some wireless communication protocols, radio receivers typically look for a specific preamble or 'synchronization word' (i.e., a portion of the frequency at which the received data signal is monitored by the receiver). This synchronization word is usually a signal pattern known a priori to the receiver and helps the receiver perform frequency synchronization, automatic gain control (AGC) training, and symbol timing estimation.

[0004] To detect the preamble, some conventional radio receivers typically use an autocorrelator that correlates samples of the incoming signal with the expected preamble signal pattern. When the autocorrelator finds a flat region, the radio receiver determines that a preamble region has been detected. This is usually achieved by using a multi-peak detector and comparing the intensity of the received peaks to a threshold to determine whether a preamble has been detected. Once the preamble is detected, the receiver's 'synchronization engine' moves to the next stage, which may involve frame synchronization.

[0005] However, the applicant has learned that interference from other sources, such as other transmitters, could cause significant problems in the synchronization process. Specifically, if the interfering transmitter (e.g., another...) If the encoder (transmitter) is transmitting a strong, constant carrier signal, the autocorrelator may incorrectly determine that a valid preamble has been detected, triggering the synchronization engine to proceed to the next stage. These false alarms can cause the receiver to be 'locked' while attempting to perform synchronization on an invalid signal. During this time, the receiver is effectively 'blind' to the signal it should be locking, as it is busy searching for a match in the expected access address field (which does not actually exist). Typically, the receiver will remain locked until a timeout occurs or some other check fails.

[0006] The 'RF_PHY' specification includes a test case that requires continuously transmitting random data in the channel of interest using such an interference source. In this test, the data is pattern-mapped in the same way as the synchronization word, which is pattern-mapped as a sequence of ten repeating '01's. Therefore, fragments close to the actual synchronization word may have a high probability of appearing in the interference source data.

[0007] The applicant understands that the aforementioned problems are particularly problematic when communicating at the receiver's sensitivity limits, such as when operating under relatively low signal-to-noise ratio (SNR) conditions. According to Radio receivers operating in 'long range' (which may be interchangeably referred to herein as 'BLE long range' or simply 'BLE LR') mode are expected to operate under extremely low SNR conditions. To achieve this, the receiver can use a relatively low threshold for peak comparison to ensure detection of the preamble; however, this comes at the cost of more false alarms, as the receiver is more susceptible to noise and interference, i.e., detecting false alarms caused by a constant carrier.

[0008] The applicant understands that avoiding these issues would be advantageous. Summary of the Invention

[0009] From a first perspective, embodiments of the present invention provide a radio receiver apparatus configured to receive digital radio signals and determine whether the digital radio signals contain a predetermined cyclic preamble having a period, the radio receiver apparatus comprising:

[0010] The input section is configured to sample the received digital radio signals and generate multiple samples therefrom;

[0011] A buffer configured to store the plurality of samples;

[0012] A first autocorrelator is configured to correlate a first subset of the plurality of samples with a second subset of the plurality of samples to generate a first correlation metric, wherein the second subset of samples is stored in the buffer earlier than the first subset of samples by a first delay, the first delay being an even multiple of half a preamble period.

[0013] A second autocorrelator is configured to correlate a first subset of the plurality of samples with a third subset of the plurality of samples to generate a second correlation metric, wherein the third subset of samples is stored in the buffer a second delay earlier than the first subset of samples, the second delay being an odd multiple of half the preamble period; and

[0014] The processing section is configured to calculate the difference between the first and second correlation metrics and determine that the radio signal contains the predetermined cyclic preamble when the difference is greater than a threshold.

[0015] This first aspect extends to a method for determining whether a received digital radio signal contains a predetermined cyclic preamble having a period, the method comprising:

[0016] Receive digital radio signals;

[0017] The received digital radio signal is sampled to generate multiple samples from it;

[0018] Store the multiple samples;

[0019] Autocorrelate a first subset of the plurality of samples with a second subset of the plurality of samples to generate a first correlation metric, wherein the second subset of samples is stored earlier than the first subset of samples by a first delay, the first delay being an even multiple of half a preamble period;

[0020] The first subset of the plurality of samples is autocorrelated with the third subset of the plurality of samples to generate a second correlation metric, wherein the third subset of samples is stored earlier than the first subset of samples by a second delay, the second delay being an odd multiple of half the preamble period;

[0021] Calculate the difference between the first and second relevant metrics; and

[0022] When the difference is greater than a threshold, it is determined that the radio signal contains the predetermined cyclic preamble.

[0023] The first aspect also extends to a non-transitory computer-readable medium comprising instructions that, when executed by a processor, cause the processor to perform a method for determining whether a received digital radio signal contains a predetermined cyclic preamble having a period, the method comprising:

[0024] Receive digital radio signals;

[0025] The received digital radio signal is sampled to generate multiple samples from it;

[0026] Store the multiple samples;

[0027] Autocorrelate a first subset of the plurality of samples with a second subset of the plurality of samples to generate a first correlation metric, wherein the second subset of samples is stored earlier than the first subset of samples by a first delay, the first delay being an even multiple of half a preamble period;

[0028] The first subset of the plurality of samples is autocorrelated with the third subset of the plurality of samples to generate a second correlation metric, wherein the third subset of samples is stored earlier than the first subset of samples by a second delay, the second delay being an odd multiple of half the preamble period;

[0029] Calculate the difference between the first and second relevant metrics; and

[0030] When the difference is greater than a threshold, it is determined that the radio signal contains the predetermined cyclic preamble.

[0031] The first aspect also extends to a computer software product comprising instructions that, when executed by a processor, cause the processor to perform a method for determining whether a received digital radio signal contains a predetermined cyclic preamble having a period, the method comprising:

[0032] Receive digital radio signals;

[0033] The received digital radio signal is sampled to generate multiple samples from it;

[0034] Store the multiple samples;

[0035] Autocorrelate a first subset of the plurality of samples with a second subset of the plurality of samples to generate a first correlation metric, wherein the second subset of samples is stored earlier than the first subset of samples by a first delay, the first delay being an even multiple of half a preamble period;

[0036] The first subset of the plurality of samples is autocorrelated with the third subset of the plurality of samples to generate a second correlation metric, wherein the third subset of samples is stored earlier than the first subset of samples by a second delay, the second delay being an odd multiple of half the preamble period;

[0037] Calculate the difference between the first and second relevant metrics; and

[0038] When the difference is greater than a threshold, it is determined that the radio signal contains the predetermined cyclic preamble.

[0039] Therefore, it should be understood that embodiments of the present invention provide an improved arrangement in which the cyclic nature of the preamble can be utilized. A first autocorrelator provides 'constructive' correlation (refer to the terminology for constructive interference) because, if a preamble is present, correlating a subset of samples with versions of the same samples delayed by a cyclic period (or an integer multiple of the period) should provide high correlation. A second autocorrelator provides 'cancellous' correlation (refer to the terminology for canceling interference) because, if a preamble is present, correlating a subset of samples with versions of the same samples delayed by half a period (or any integer and half a period) should provide low correlation. The larger the difference between these two correlation metrics (i.e., the first minus the second), the greater the confidence in determining that the received signal contains a preamble. This arrangement can significantly improve the receiver's co-channel interference suppression capability.

[0040] While the difference between the outputs of the first and second autocorrelators can be significant for many different preamble patterns, embodiments of the invention work particularly well when the cyclic preamble is symmetric about its midpoint (i.e., it is 'palindromic') and the number of 0s is the same as the number of 1s. In other words, this arrangement is particularly advantageous when the first half of the preamble is the inverse of the second half. For example, the BLE LR specification outlines the use of the sequence [00111100] as the synchronization word. This byte, if split in half, has the first half

[0011] , i.e., the logical inverse of the second half

[1100] . Thus, when this BLE LR synchronization word is correlated with its own version delayed by a period (i.e., an even multiple of half a period), a high correlation metric, ideally 100%, will be achieved; but when correlated with its own version delayed by half a period (or an odd multiple of half a period), a low correlation metric, ideally 0%, will be achieved. Therefore, in some embodiments, the cyclic preamble is [00111100].

[0041] In some embodiments, the radio receiver device is configured to operate according to the Bluetooth Low Energy protocol. Preferably, the radio receiver device is configured to operate according to the long-range mode of the Bluetooth Low Energy protocol. It should be understood that a typical BLE LR packet includes a preamble field, an encoded access address field, an encoded indicator field, a Protocol Data Unit (PDU) field, a Cyclic Redundancy Check (CRC) field, and a Termination Mark (TERM) field. The preamble field of the BLE LR packet is 80 μs long and consists of ten repeating synchronization word sequences [00111100] used to achieve initial frame synchronization under low sensitivity conditions.

[0042] Those skilled in the art will understand that the radio receiver may include several additional components typical of a radio receiver chain, such as filters, digital signal processors, microcontrollers, etc. In some embodiments, the radio receiver further includes a demodulator configured to demodulate the received digital radio signal. The demodulator may include a peak detector. In one set of such embodiments, the radio receiver is configured to enable the demodulator only when the processing portion determines that the digital radio signal contains a cyclic preamble. Thus, the radio receiver can be arranged to demodulate signals determined to contain a preamble. Advantageously, the demodulator is less likely to be 'locked in' due to false alarms about preamble detection, because the radio receiver is less likely to produce such false alarms than conventional receivers known in the art itself. When not enabled, the demodulator can be disconnected or operate in a low-power mode, thereby saving power. Similarly, processing overhead can be reduced by processing only those signals that the receiver is certain contain a preamble.

[0043] The demodulator can be configured to perform symbol timing recovery and frequency offset correction. The demodulator can also be additionally or alternatively configured to perform frame synchronization. The demodulator may include a Viterbi decoder.

[0044] The difference between the first and second correlation metrics can be used directly. However, in some embodiments, the difference is filtered using a low-pass filter before determining whether the digital radio signal includes a preamble. This filter can advantageously suppress noise such that the time average of the difference must exceed a threshold for the device to determine the presence of a preamble. Such filters can also be used to suppress shorter '0101' sequences of pattern maps from other interfering transmitters of the types previously outlined.

[0045] As outlined above, the processing section calculates the difference between the first and second correlation metrics and determines that the radio signal contains a predetermined cyclic preamble when the difference is greater than a threshold. In some embodiments, the threshold may be a predetermined value. The threshold may be fixed, or it may be configurable, for example, during device initialization or during the design or testing phase. For example, the threshold may be set based on simulation or experimental data by weighing receiver sensitivity against sensitivity to false alarms.

[0046] Those skilled in the art will understand that, according to some protocols (such as BLE), the radio receiver device can be configured to perform resynchronization, wherein it is intended to abandon its current synchronization and instead support another—usually stronger—channel. In some embodiments, the radio receiver device is configured to increase a threshold once synchronization has been performed. In one set of such embodiments, the threshold is increased to a value dependent on the difference calculated by a processing section for the received radio signal for which synchronization has been performed.

[0047] The correlation metric can simply be the output of the corresponding autocorrelator. However, the applicant understands that in some cases, it may be advantageous to normalize the power of the received radio signal to account for variations in the transmit power and / or attenuation of the received signal. In some embodiments, the processing section is configured to:

[0048] Determine the corresponding power for each of the first, second, and third subsets;

[0049] Determine the corresponding power of the output of each of the first and second autocorrelators;

[0050] Divide the power of the output of the first autocorrelator by the product of the corresponding powers of the first and second subsets to generate a first ratio; and

[0051] The power of the output of the second autocorrelator is divided by the product of the respective powers of the first and third subsets to generate a second ratio. According to this embodiment, the first and second correlation metrics can be derived from the first and second ratios, respectively. In a particular set of embodiments, the first and second correlation metrics are the first and second ratios, respectively.

[0052] The principle outlined above utilizes the period and half-cycle delays of the first and second autocorrelators to ensure constructive and destructive correlations, respectively, in the presence of a preamble. However, the applicant understands that for other modes, different autocorrelator delays, particularly for the second autocorrelator, can achieve the same effect. For example, if the cyclic preamble is [00110011], constructive correlation is achieved by correlating the sample with its own version delayed by the entire preamble period, as described above, while destructive correlation is achieved by correlating the sample with its own version delayed by a quarter period (or an odd multiple of a quarter period). By appropriately selecting the delays for each autocorrelation process (i.e., utilizing constructive and destructive correlations), other cyclic preamble modes or preamble modes with symmetry about their middle can also be used with the principles of this invention.

[0053] Therefore, from a second perspective, embodiments of the present invention provide a radio receiver apparatus configured to receive digital radio signals and determine whether the digital radio signals contain a predetermined cyclic preamble having a period, the radio receiver apparatus comprising:

[0054] The input section is configured to sample the received digital radio signals and generate multiple samples therefrom;

[0055] A buffer configured to store the plurality of samples;

[0056] A first autocorrelation unit is configured to correlate a first subset of the plurality of samples with a second subset of the plurality of samples to generate a first correlation metric, wherein the second subset of samples is stored in the buffer at a first delay earlier than the first subset of samples, and the first delay is selected such that when the current preamble is present, the first autocorrelation unit correlates a first copy of the preamble with a second copy of the preamble.

[0057] A second autocorrelator is configured to correlate a first subset of the plurality of samples with a third subset of the plurality of samples to generate a second correlation metric, wherein the third sample subset is earlier than the first sample subset by a second delay stored in the buffer, and the second delay is selected such that, when the current preamble is present, the second autocorrelator correlates a first copy of the preamble with an inverted copy of the preamble; and

[0058] The processing section is configured to calculate the difference between the first and second correlation metrics and determine that the radio signal contains the predetermined cyclic preamble when the difference is greater than a threshold.

[0059] This second aspect extends to a method for determining whether a received digital radio signal contains a predetermined cyclic preamble with a period, the method comprising:

[0060] Receive digital radio signals;

[0061] The received digital radio signal is sampled to generate multiple samples from it;

[0062] Store the multiple samples;

[0063] Autocorrelate a first subset of the plurality of samples with a second subset of the plurality of samples to generate a first correlation metric, wherein the second subset of samples is stored earlier than the first subset of samples by a first delay, the first delay being selected such that, when the current preamble exists, the autocorrelation between the first subset and the second subset includes autocorrelating a first copy of the preamble with a second copy of the preamble;

[0064] The first subset of the plurality of samples is autocorrelated with a third subset of the plurality of samples to generate a second correlation metric, wherein the third subset of samples is stored earlier than the first subset of samples with a second delay different from the first delay, and the second delay is selected such that when the current preamble exists, the autocorrelation between the first subset and the third subset includes autocorrelation between a first copy of the preamble and an inverted copy of the preamble.

[0065] Calculate the difference between the first and second relevant metrics; and

[0066] When the difference is greater than a threshold, it is determined that the radio signal contains the predetermined cyclic preamble.

[0067] The second aspect also extends to a non-transitory computer-readable medium comprising instructions that, when executed by a processor, cause the processor to perform a method for determining whether a received digital radio signal contains a predetermined cyclic preamble having a period, the method comprising:

[0068] Receive digital radio signals;

[0069] The received digital radio signal is sampled to generate multiple samples from it;

[0070] Store the multiple samples;

[0071] Autocorrelate a first subset of the plurality of samples with a second subset of the plurality of samples to generate a first correlation metric, wherein the second subset of samples is stored earlier than the first subset of samples by a first delay, the first delay being selected such that, when the current preamble exists, the autocorrelation between the first subset and the second subset includes autocorrelating a first copy of the preamble with a second copy of the preamble;

[0072] The first subset of the plurality of samples is autocorrelated with a third subset of the plurality of samples to generate a second correlation metric, wherein the third subset of samples is stored earlier than the first subset of samples with a second delay different from the first delay, and the second delay is selected such that when the current preamble exists, the autocorrelation between the first subset and the third subset includes autocorrelation between a first copy of the preamble and an inverted copy of the preamble.

[0073] Calculate the difference between the first and second relevant metrics; and

[0074] When the difference is greater than a threshold, it is determined that the radio signal contains the predetermined cyclic preamble.

[0075] The second aspect also extends to a computer software product comprising instructions that, when executed by a processor, cause the processor to perform a method for determining whether a received digital radio signal contains a predetermined cyclic preamble having a period, the method comprising:

[0076] Receive digital radio signals;

[0077] The received digital radio signal is sampled to generate multiple samples from it;

[0078] Store the multiple samples;

[0079] Autocorrelate a first subset of the plurality of samples with a second subset of the plurality of samples to generate a first correlation metric, wherein the second subset of samples is stored earlier than the first subset of samples by a first delay, the first delay being selected such that, when the current preamble exists, the autocorrelation between the first subset and the second subset includes autocorrelating a first copy of the preamble with a second copy of the preamble;

[0080] The first subset of the plurality of samples is autocorrelated with a third subset of the plurality of samples to generate a second correlation metric, wherein the third subset of samples is stored earlier than the first subset of samples with a second delay different from the first delay, and the second delay is selected such that when the current preamble exists, the autocorrelation between the first subset and the third subset includes autocorrelation between a first copy of the preamble and an inverted copy of the preamble.

[0081] Calculate the difference between the first and second relevant metrics; and

[0082] When the difference is greater than a threshold, it is determined that the radio signal contains the predetermined cyclic preamble. Attached Figure Description

[0083] Certain embodiments of the invention will now be described with reference to the accompanying drawings, by way of only non-limiting examples, in which:

[0084] Figure 1 This is a block diagram illustrating the format of a BLE LR data packet;

[0085] Figure 2 This is a graph showing the correlator response under good SNR conditions;

[0086] Figure 3 This is a graph showing the correlator response under poor SNR conditions;

[0087] Figure 4 It is a graph of the correlator response in the presence of severe co-channel interference sources;

[0088] Figure 5 This is a block diagram of a radio receiver device according to an embodiment of the present invention;

[0089] Figure 6 yes Figure 5 A block diagram of the preamble recognition module in the device.

[0090] Figure 7 yes Figure 6 The core block diagram;

[0091] Figure 8 yes Figure 6 Block diagram of the filter;

[0092] Figure 9 yes Figure 5 A block diagram of the peak classifier in the device;

[0093] Figure 10 This is a graph showing a comparison of packet error rate (PER) when constant carrier is disabled and enabled, with and without using embodiments of the present invention. Detailed Implementation

[0094] Figure 1 This is a block diagram illustrating the format of BLE LR packet 2. BLE LR packet 2 includes: an 80μs synchronization word (SW) preamble field 4, a 256μs encoded access address (AA) field 6, a 16μs encoded indicator (CI) field 8, a 24μs first terminator (TERM1) field 10, an N*16*Pμs protocol data unit (PDU) field 12, a 48*Pμs cyclic redundancy check (CRC) field 14, and a 6*Pμs second terminator (TERM2) field 16.

[0095] AA field 6, CI field 8 and TERM1 field 10 together form the first FEC block 18.

[0096] PDU field 12, CRC field 14 and TERM2 field 16 together form the second FEC block 20.

[0097] The SW field 4 of BLE LR data packet 2 is an 80μs long preamble portion, consisting of ten repeating synchronization word sequences [00111100] used to achieve initial frame synchronization under low sensitivity conditions.

[0098] Figure 2 This is a graph showing the correlator response under good SNR conditions. Figure 2 This shows what the output of a conventional decoder in a prior art radio receiver looks like under good (-90dBm) SNR conditions.

[0099] The highlighted area 21 in SW section 22 shows the location of the actual SW field 4. Frame synchronization (FRS) is performed at the end of this sequence, where the frame synchronization process searches for a match. The word to be matched is called the frame synchronization word (FW).

[0100] from Figure 2 It is readily apparent that the peak in SW region 22 (i.e., the time period corresponding to SW preamble field 4) is much stronger than the peak immediately following it, corresponding to the noise response; and also stronger than the subsequent peak, due to the mismatch of multiple [00111100] sequences in the extended pattern. Therefore, if the peak requirement is set high enough, there is almost no risk of premature synchronization with the noise lock (i.e., false alarms) and thus failure to properly synchronize to the beginning of the actual data packet 2.

[0101] However, on the contrary, Figure 3 A graph of the correlator response under poor SNR conditions (-105 dBm) is provided. Here, the distortion is now severe enough that the peak amplitude in SW region 22 is reduced to the peak amplitude caused by noise. In order to still be able to detect packet 2, the peak requirement must be reduced. However, this increases the possibility of early synchronization noise (i.e., false alarms). At this inflection point, significant distortion has occurred, making it impossible to trust that the SW peak is higher than the noise peak or the peak generated in the FW portion of packet 2.

[0102] Figure 4 This is a graph showing the correlator response in the presence of severe co-channel interference. Figure 4 This illustrates the situation when there are numerous S8 interference sources in the channel. Even if the interference source does not continuously transmit multiple [00111100] patterns, it will still produce decoder peaks that are much stronger than the actual SW under poor SNR conditions. If a single synchronization threshold is to be used, a trade-off must be made between selectivity and sensitivity.

[0103] Figure 5 This is a block diagram of a radio receiver device 100 according to an embodiment of the present invention. The radio receiver device 100 includes a preamble recognition module 102, which processes all calculations related to the autocorrelation used to identify SW 4 in the previously described BLE LR data packet 2. The preamble recognition module 102 includes a core 104 and a filter 106, which will be explained in more detail later.

[0104] The device 100 also includes: a dual correlator (DBC) 108; a peak classifier 110; two data path processing blocks 112a and 112b; a block selector 114; and a frame synchronization checker 116.

[0105] The DBC 108 contains logic for performing various functions including automatic gain control (AGC) training and symbolic timing estimation, as well as finite state machine (FSM) logic.

[0106] Peak classifier 110 is a decision-maker that evaluates the peak chain received from DBC 108 and determines whether it is a strong or weak peak chain. When a strong peak is observed, peak classifier 110 initiates or restarts the processing blocks 112a, 112b of the designated data path used to receive the data. When a weak peak is observed, peak classifier 110 informs all blocks to decay their thresholds, as explained in more detail later. Block selector 114 is configured to determine which blocks 112a, 112b will respond to the strong peak event.

[0107] Frame synchronization checker 116 processes all frame synchronization signals from data path processing blocks 112a and 112b. Once a frame synchronization is received from one of these blocks 112a and 112b, it keeps that block active but deactivates it and resets the other block. If multiple frame synchronization signals arrive simultaneously, it may select the one with the lowest index.

[0108] Typically, delayed resynchronization can cause problems with frame synchronization and / or frequency offset estimation. This usually sets a lower limit to the leniency of the threshold setting.

[0109] Figure 6 yes Figure 5 A block diagram of the preamble recognition module 102 in the device is shown, illustrating a core 104 and a filter 106 for smoothing the output from the core 104. The preamble recognition module 102 is responsible for detecting the start of data packet 2, i.e., detecting SW 4. In some conventional receivers known in the art, the DBC 108 would handle this function, rather than having a dedicated module as provided in the embodiments of the present invention.

[0110] Therefore, the main task of the preamble identification module 102 is to identify the preamble period of the BLE LR data packet and reduce the possibility of incorrect synchronization noise or constant carrier signal before valid data and resynchronize bit patterns that occur in the FRS area.

[0111] Figure 7 yes Figure 6 A block diagram of core 104. Core 104 includes two autocorrelators 202 and 204, which receive samples of the input IQ signal. The latest 32 samples y of the IQ signal. i-1 y i-2 ...y i-32 The samples are stored in a 32-bit shift register 206. This arrangement is configured such that the 32 samples correspond to the entire period of the preamble portion (i.e., SW 4). Of course, it should be understood that different sampling rates can be used, and shift registers of different sizes can be used accordingly.

[0112] The first autocorrelator 202 receives samples from 32 samples prior to y. i-32 And the current sample of the IQ signal, which follows the conjugate function 208. Therefore, the first autocorrelator 202 correlates the IQ signal with itself one period ago. Using the BLE LR preamble mode of [00111100], it can be expected that this autocorrelator 202 will output a high correlation value if the preamble is present. The correlated output is stored in shift register 210, which stores the latest 96 outputs from the first autocorrelator 202 (i.e., y t-32 y t(Latest 96 calculations) u0 to u 95 .

[0113] The second autocorrelator 204 receives samples from 16 samples prior, i.e., y i-16 And the current sample of the IQ signal, which follows the conjugate function 208. Therefore, the second autocorrelator 204 correlates the IQ signal with itself half a cycle ago. Using the BLE LR preamble mode of [00111100], it can be expected that if the preamble is present, this autocorrelator 202 will output a low correlation value. The correlated output is stored in another shift register 212, which stores the latest 96 outputs from the second autocorrelator 204 (i.e., y t-16 y t (Latest 96 calculations) v0 to v 95 .

[0114] Another correlator 214 is configured to correlate samples with their own conjugate (from conjugate function 208) to determine the power of the received IQ signal. The latest 128 power samples p0 to p... 127 It is stored in another shift register 216.

[0115] Therefore, the two autocorrelation results (one in y) t With y t-32 Between, another in y t With y t-16 The values ​​(between) are stored in the corresponding shift registers 210 and 212. Each of these shift registers 210 and 212 stores the latest 96 values, that is, the values ​​corresponding to the last three cycles.

[0116] After data is obtained from registers 210 and 212, the elements of the registers are summed using corresponding summing blocks 218 and 220, and the power of the sum is calculated by corresponding power calculation blocks 222 and 224. Power calculation blocks 222 and 224 multiply their respective inputs from the corresponding summing blocks 218 and 220 by their conjugates to obtain the power.

[0117] The calculated power from power calculation blocks 222 and 224 is respectively transferred to the corresponding calculation ratio units (CRUs) 226 and 228.

[0118] Calculate the partial sums 230a-c of the power samples in the corresponding shift register 216. First, 230a calculates the partial sum of the most recent 96 power samples. Second, 230b calculates the partial sum of the power of the samples 16 samples before and 95 samples before (i.e., corresponding to half a cycle before and 95 samples before that). Third, 230c calculates the partial sum of the power of the samples 32 samples before and 95 samples before (i.e., corresponding to a full cycle before and 95 samples before that).

[0119] The first CRU 226 divides the power from the first power calculation block 222 (i.e., the power corresponding to the autocorrelation of the IQ signal with the sample one period ago) by the product of the first part and 230a (i.e., the sum of the power of the 0th to 95th most recent samples) and the third part and 230c (i.e., the sum of the power of the 32nd to 127th most recent samples) of the output put.

[0120] The second CRU 228 divides the power from the second power calculation block 224 (i.e., the power corresponding to the autocorrelation of the IQ signal with the sample half a cycle ago) by the product of the output put of the second part and 230a (i.e., the sum of the power of the 16th to 111th most recent samples) and the third part and 230c (i.e., the sum of the power of the 32nd to 127th most recent samples).

[0121] Finally, the output of the second CRU 228 is subtracted from the output of the first CRU 226 using summation block 232 to generate the output signal x.

[0122] Figure 8 yes Figure 6 A block diagram of filter 106 is provided. Filter 106 is arranged to receive an output signal x from core 104 and generate a filtered output signal y. Filter 106 is also configured to receive a 'clear' signal, which is used to reset filter 106. If the clear signal is set, the filter output y is set to be equal to the input signal x from core 104.

[0123] Filter 106 follows a specific filter algorithm, utilizing the damping factor α of a moving average filter applied to the correlator output. If a clear signal is confirmed, the filter output y is set equal to the input signal x from core 104, as described above. However, if no clear signal is set, another condition is checked.

[0124] If the damping factor α is zero, then the filter output y is set to the maximum of itself and the input signal x from core 104, or in other words, if x is higher than y, then y is set to the value of the input signal x, i.e., y = max(y, x).

[0125] Otherwise, if the damping factor α is not zero, the filter output y is set to the sum of (1-α)*y and α*x, that is, y = (1-α)y + αx.

[0126] Figure 9 yes Figure 5A block diagram of peak classifier 110 in device 100. Peak classifier 110 is a decision maker and is configured to evaluate the quality of the peak chain detected by DBC 108 (which is supplied to input B) with the assistance of the output y from preamble recognition module 102 (which is supplied to input A). Peak classifier 110 processes two types of peaks.

[0127] First, if both A (i.e., the output y from the preamble recognition module 102) and B (i.e., the peak intensity of DBC 108) are greater than the corresponding thresholds C and D, then the peak classifier 110 considers this event as a strong peak detection. In this case, the peak classifier 110 starts (or restarts) the first data processing block 112a and passes the new value of the peak intensity and the output y from the preamble recognition module 102 to the block to update the thresholds C and D.

[0128] Secondly, if B is not greater than threshold D or A is not greater than threshold C, but they are still greater than the configured lower limit, the peak classifier 110 considers this event as a weak peak detection. In this case, the peak classifier 110 sends a trigger signal to all blocks to correspondingly attenuate the threshold of the blocks. The lower limit can be set via a configurable register (not shown).

[0129] Figure 10 This is a simulated graph comparing the packet error rate (PER) when constant carrier is disabled and enabled in both cases, using and not using embodiments of the present invention. Specifically, Figure 10 The comparison PER curves of a conventional BLE LR receiver with the constant carrier blocker disabled and enabled are shown by curves 500 and 502, and the comparison PER curves of a BLE LR receiver (i.e., with a preamble identification module) operating according to an embodiment of the present invention are shown by curves 504 and 506.

[0130] Specifically, curve 500 shows a conventional receiver with constant carrier disabled; curve 502 shows a conventional receiver with constant carrier enabled; Figure 504 shows a receiver with constant carrier disabled according to an embodiment of the present invention; and curve 506 shows a receiver with constant carrier enabled according to an embodiment of the present invention.

[0131] As can be seen, the performance of a conventional BLE receiver degrades significantly when a constant carrier is enabled, as evidenced by the much higher PER (Performance Ratio) of curve 502 compared to curve 500. In contrast, with a constant carrier enabled, the performance of the BLE LR receiver of this invention is closer to that of a conventional receiver without a constant carrier, as demonstrated by comparing curves 506 and 500. This indicates a significant improvement in the performance of a conventional receiver with a constant carrier, as shown by comparing curves 506 and 502.

[0132] Even with the constant carrier disabled, the BLE LR receiver of the present invention provides improved performance compared to conventional receivers, as shown by comparing curves 504 and 500 respectively.

[0133] from Figure 10 As can be seen, compared with conventional receivers, the BLE LR receiver of the present invention has significantly greater robustness to constant carrier conditions.

[0134] While specific embodiments of the invention have been described in detail, those skilled in the art will understand that the detailed embodiments do not limit the scope of the claimed invention.

Claims

1. A radio receiver apparatus configured to receive digital radio signals and determine whether the digital radio signals contain a predetermined cyclic preamble having a period, the radio receiver apparatus comprising: The input section is configured to sample the received digital radio signals and generate multiple samples therefrom; A buffer configured to store the plurality of samples; A first autocorrelator is configured to correlate a first subset of the plurality of samples with a second subset of the plurality of samples to generate a first correlation metric, wherein the second subset of samples is stored in the buffer earlier than the first subset of samples by a first delay, the first delay being an even multiple of half a preamble period. A second autocorrelator is configured to correlate a first subset of the plurality of samples with a third subset of the plurality of samples to generate a second correlation metric, wherein the third subset of samples is stored in the buffer at a second delay earlier than the first subset of samples. as well as The processing section is configured to calculate the difference between the first and second correlation metrics and determine that the digital radio signal contains the predetermined cyclic preamble when the difference is greater than a threshold. Its features are: The second delay is equal to an odd multiple of half the preamble period; and The radio receiver device includes a low-pass filter configured to filter the difference before the processing section determines whether the digital radio signal includes the predetermined cyclic preamble.

2. The radio receiver apparatus according to claim 1, wherein the first half of the predetermined cyclic preamble is the inverted value of the second half of the predetermined cyclic preamble.

3. The radio receiver apparatus according to claim 1 or 2, wherein the predetermined cyclic preamble is [00111100].

4. The radio receiver device according to claim 1 or 2, configured to operate according to the Bluetooth Low Energy protocol.

5. The radio receiver apparatus of claim 1 or 2, further comprising a demodulator, the apparatus being configured to activate the demodulator only when the processing portion determines that the digital radio signal includes the predetermined cyclic preamble.

6. The radio receiver device according to claim 1 or 2, wherein the threshold is increased once synchronization has been performed.

7. The radio receiver apparatus of claim 6, wherein the threshold is increased to depend on the value of the difference calculated by the processing section for the received digital radio signal that has been synchronized.

8. The radio receiver apparatus according to claim 1 or 2, wherein the processing section is configured to: Determine the corresponding power for each of the first, second, and third subsets; Determine the corresponding power of the output of each of the first and second autocorrelators; The power of the output of the first autocorrelator is divided by the product of the respective powers of the first and second subsets to generate a first ratio; as well as The power of the output of the second autocorrelator is divided by the product of the corresponding powers of the first and third subsets to generate a second ratio.

9. The radio receiver apparatus of claim 8, wherein the first and second correlation metrics are derived from the first and second ratios, respectively.

10. The radio receiver apparatus of claim 9, wherein the first and second correlation metrics are the first and second ratios, respectively.

11. A method for determining whether a received digital radio signal contains a predetermined cyclic preamble having a period, the method comprising: Receive digital radio signals; The received digital radio signal is sampled to generate multiple samples from it; Store the multiple samples; Autocorrelate a first subset of the plurality of samples with a second subset of the plurality of samples to generate a first correlation metric, wherein the second subset of samples is stored earlier than the first subset of samples by a first delay, the first delay being an even multiple of half a preamble period; The first subset of the plurality of samples is autocorrelated with the third subset of the plurality of samples to generate a second correlation metric, wherein the third subset of samples is stored with a second delay earlier than the first subset of samples. Calculate the difference between the first and second relevant metrics; as well as When the difference is greater than the threshold, it is determined that the radio signal contains the predetermined cyclic preamble. Its features are: The second delay is equal to an odd multiple of half the preamble period; and The difference is low-pass filtered before determining whether the digital radio signal includes the predetermined cyclic preamble.

12. The method of claim 11, wherein the first half of the predetermined cyclic preamble is the inverted value of the second half of the predetermined cyclic preamble.

13. The method according to claim 11 or 12, wherein the predetermined cyclic preamble is [00111100].

14. The method of claim 11 or 12, further comprising operating in accordance with the Bluetooth Low Energy protocol.

15. The method of claim 11 or 12, wherein the demodulator is activated only when it is determined that the digital radio signal includes the predetermined cyclic preamble.

16. The method of claim 11 or 12, further comprising increasing the threshold once synchronization has been performed.

17. The method of claim 16, further comprising increasing the threshold to a value dependent on the difference calculated for the received digital radio signal for which synchronization has been performed.

18. The method according to claim 11 or 12, further comprising: Determine the corresponding power for each of the first, second, and third subsets; Determine the corresponding power of the output of each of the first and second autocorrelation steps; The power of the output of the first autocorrelation step is divided by the product of the corresponding powers of the first and second subsets to generate a first ratio; as well as The power of the output of the second autocorrelation step is divided by the product of the corresponding powers of the first and third subsets to generate a second ratio.

19. The method of claim 18, wherein the first and second correlation metrics are derived from the first and second ratios, respectively.

20. The method of claim 19, wherein the first and second correlation metrics are the first and second ratios, respectively.

21. A non-transitory computer-readable medium comprising instructions that, when executed by a processor, cause the processor to perform the method according to claim 11 or 12.

Images