Method for acquisition of a radio signal component among a plurality of radio signal components, related radio navigation system, related radio transmitter, and related radio receiver - Patents.com

Abstract

The present invention relates to a method, associated device and related system for acquisition in a radio receiver of at least one radio signal component of a plurality of radio signal components, each radio signal component being transmitted by a respective radio transmitter of a plurality of radio transmitters, each radio signal component comprising a data signal carrier for carrying a data signal component of said radio signal component, said data signal component comprising a spreading code, said spreading code comprising a predetermined number L of chips, each chip of said spreading code having a duration, said predetermined number of chips being such that a measure calculated as a ratio between the product of a radio receiver front-end bandwidth with the spreading code length and with the chip duration and the total number of radio signal components of a radio navigation system does not exceed 25.

Description

[Technical field]

[0001] Embodiments of the present invention relate to a method for providing a multiple access scheme supporting capture of radio signal components out of a plurality of radio signal components, an associated radio navigation system, an associated radio transmitter and an associated radio receiver of such a radio navigation system. [Background technology]

[0002] The determination of the position of a receiver supported by a Global Navigation Satellite System, further called GNSS, requires the reception and processing of several GNSS signals. Current GNSS systems are considered such that each satellite transmits several GNSS signals that may have the same carrier frequency, such as the GPS C / A and P signals, both transmitted at 1575.42 MHz for the GPS system, or the Galileo E1-B and E1-C signals, both transmitted at 1575.42 MHz for the Galileo system, or may have different carrier frequencies but still be within the same spectral region, such as the GPS L5 signals or the Galileo E5a-I / E5a-Q / E5b-I / E5a-Q signals, all transmitted at 1191.795 MHz, which are spectrally close to the 1575.42 MHz carrier frequency of the GPS C / A and P signals and the Galileo E1-B and E1-C signals previously introduced. From here on, the term radio signal component will be applied to define each of these signals transmitted by a GNSS satellite.

[0003] [Reference 1]: "Understanding GPS Principles and Application", Kaplan Artech House Publish, ISBN0-89006-793-7

[0004] Typically, at least four GNSS radio signal components transmitted by four different GNSS satellites are required to determine the receiver's position (i.e., X, Y, and Z coordinates) and time. In the following description, F will represent the number of radio signal components transmitted per satellite. Furthermore, in the following description, for simplicity, it is considered that V satellites are in view of the radio receiver and that each satellite transmits one radio signal component F=1. Thus, there is a total number of P=Q×F=Q radio signal components provided by a radio navigation system such as a GNSS and comprising Q satellites. v = V × F = V different radio signal components are received at the radio receiver. For example, the negligible number Q of satellites is 24 for the GPS system and 30 for the Galileo system. Furthermore, the radio receiver may have a limited number of satellites, e.g. due to receiver complexity limits. v satellites (W≦P vIt is considered that a receiver may process only a subset of W radio signal components from V visible satellites transmitted by at least four different GNSS satellites (P = V ≤ P = Q). Therefore, W needs to be at least four to guarantee the receiver's position and time estimate. A conventional scheme, described in more detail in [Reference 1], for determining the receiver's position and time therefore consists in estimating a pseudo-range (PR) between the receiver and each of at least four different GNSS satellites based on processing each radio signal component transmitted by at least four different GNSS satellites. The pseudo-range is defined as the estimated distance between the transmission source (i.e., a GNSS satellite) and the radio receiver. In addition to the real distance, the PR includes contributions from the receiver clock offset, contributions from measurement noise (including interference such as multipath and other local effects), and correction residuals of the satellite clock and orbit, or of atmospheric effects such as ionospheric delays. In the following description, when one wireless signal component is processed to estimate the PR between the receiver and the GNSS satellite transmitting this unique wireless signal component, this unique wireless signal component will be defined as the wireless signal component of interest. The presence and reception of several wireless signal components coexisting in the same spectral region and exhibiting some spectral overlap will inevitably result in degradation of the wireless receiver performance. Therefore, when processing one wireless signal component of interest out of W required by the receiver, P v Transmitted by visible satellites (P V The other wireless signal components (W−1=V−1) are called interfering wireless signal components. Thus, the term Mutual Access Interference (MAI) is typically used to describe the effect of mutually received interfering wireless signal components when processing one wireless signal component of interest. When the receiver sequentially processes each of the W wireless signal components of interest to estimate the corresponding pseudoranges, the roles of the wireless signal component of interest and the interfering wireless signal components will change order.

[0005] The typical structure of each radio signal component of conventional GNSS comprises a spreading code, also called spreading sequence or Pseudo-Random Noise (PRN) sequence, made up of L elementary binary symbols, also called chips. Typical spreading codes transmitted in GNSS are periodic, and the corresponding periodicity, i.e. the number of chips L in the spreading code, is called the spreading code length, but when expressed in seconds, can also be called the spreading code period. Thus, a typical GNSS signal is built up of a chip stream obtained by "infinitely" concatenating the same spreading code. Depending on the type of radio signal component transmitted by the GNSS satellite, the chip duration T c typically varies between about 1 microsecond (i.e., corresponding to a distance of 300 m) and one tenth of a microsecond (i.e., corresponding to a distance of 30 m). The chip rate f c is the chip duration T c, which is defined as the reciprocal of . Each spreading code can, but does not have to, be modulated with symbols. When symbol modulation is applied, the corresponding radio signal components are called radio signal components for data transmission. In the absence of symbol modulation, i.e. when the spreading codes are not modulated, the corresponding symbol-less or dataless radio signal components are called pilot radio signal components. The symbols can, but do not have to, be mapped to binary data by applying so-called coding techniques, such as convolutional coding or LDPC coding (which is the most encountered type of coding technique in GNSS radio signal components for data transmission). Coding techniques are particularly suitable for improving demodulation performance. The spreading codes can themselves consist of a first layer of spreading codes, called primary codes, or alternatively, the spreading codes can be obtained by combining a primary code layer and a secondary code layer, with each primary code being modulated with symbols of the secondary code. Finally, the pulse waveform, also called pulse shape, of each chip of the spreading code can be in the form of a simple Binary Phase Shift Keying (BPSK) waveform, called BPSK(N), where N is the normalized chip rate. The actual chip rate of the BPSK(N) waveform is f cis equal to N×f0, where f0 is equal to 1.023 MCps. Compared to the BPSK(N) waveform, another important type of waveform applied by GNSS radio signal components is the Binary Offset Carrier, called BOC(M,N), where N again represents the normalized chip rate and M represents the normalized subcarrier rate. It is pointed out that other more complex waveforms may also be encountered, such as the AltBOC waveform applied to the Galileo radio signal components, which is used to modulate and multiplex the four E5aI, E5aQ, E5bI, and E5bQ Galileo radio signal components. From here, in the framework of the present description, the use of BPSK or BOC waveform modulation is considered sufficient to provide the necessary background. Nevertheless, and without loss of generality, the present background can therefore be extended for such other more complex waveforms. The previous description provides the main characteristics and elements of the most encountered radio signal components transmitted by satellite navigation systems.

[0006] Figure 1 shows in its lower part a navigation signal structure of a radio signal component, consisting of a periodic binary sequence or spreading code containing L chips, modulated with symbols. Figure 1 also shows in its upper part a method for generating such a radio signal component, where a BPSK(N) waveform is applied to each chip. In the following, f s represents the symbol rate. s The maximum nominal symbol rate f s,max is typically determined by the inverse of the spreading code period, also called the spreading code duration, and f s,max =f c For example, the duration of the spreading code of the Galileo E1-B radio signal component is equal to 4 ms and the symbol rate is equal to 250 sps. Nevertheless, it is also possible to concatenate K different spreading codes within one symbol, in which case f s =f cFor example, the GPS C / A signal concatenates K=20 spreading codes, called Gold codes, each with a 1 ms spreading code period, within a symbol duration of 20 ms.

[0007] Data rate f d The coding process used to generate symbols from the binary data generated in is also represented in Fig. 1. The coding rate (CR) represents the ratio between the data rate and the symbol rate (the coding rate is 1 / 2 for example, typical coding rates applied by GPS or Galileo radio signal components for data transmission). In the absence of coding, one binary data is directly represented by one binary symbol, and the logic levels of [0,1] are mapped to the symbol levels of [+1,-1] according to the phase shift keying principle. In the absence of symbol modulation, the signal structure in the baseband of the pilot radio signal component is simply a repetition of the primary codes concatenated with each other. As mentioned before, the symbol generator can be replaced by a secondary code generator based on a predefined sequence corresponding to the secondary code.

[0008] A typical signal generation chain, applicable for a typical GNSS signal and implemented on board a radio navigation satellite, is depicted in FIG. 2 and comprises: - An on-board Frequency and Time Generation Unit (FTGU), containing one or more high-stability clocks (typically atomic clocks) and responsible for providing a highly stable timing or frequency source to the other payload units. - An on-board Navigation Signal Generation Unit (NSGU) responsible for generating the navigation signals transmitted by the satellite. As already pointed out, GNSS satellites usually transmit one or more radio signal components simultaneously, which implies that the NSGU has the capability to generate these different radio signal components. For example, the satellites of the US Global Position System (GPS) use carr = 1575.742 MHz. Similarly, satellites in the European Galileo system generate and transmit both GPS C / A and P signals on the same carrier frequency f carr = 1575.742 MHz. Thus, the NSGUs generate and transmit E1-B and E1-C radio signal components at baseband (BB) or intermediate frequency (f IF,Tx ) and up-converting them to an RF carrier frequency (e.g., 1575.742 MHz). Up-conversion to the appropriate RF carrier frequency is accomplished by applying a complex exponential function exp(2πj(f carr -f a This is typically accomplished by multiplying f by t, where f a is equal to 0 if the radio signal components are initially generated at baseband in the digital domain, and f a is the frequency at which the radio signal component reaches the intermediate frequency f IF,Tx If it is first generated by f IF,Tx Furthermore, the digital-to-analog interface (comprising a Digital-to-Analogue Converter (DAC)) may also occur at different sections of the signal generation chain. Nevertheless, and despite this variation in actual implementation, all of the aforementioned functions are encountered in a typical navigation signal generation unit. a block responsible for amplifying the navigation radio signal components and, in some cases, filtering them to avoid undesired out-of-band emissions. - Navigation antenna, responsible for radiating the generated navigation radio signal components towards the ground. It is customary to consider the apparent center of phase (ACoP) as the geometric reference point for navigation signal transmission.

[0009] The main functional blocks of a conventional GNSS receiver and the associated processing steps are outlined. Figure 3, taken from [Reference 1], depicts a typical architecture of a GNSS receiver. - The signal received by the GNSS antenna is fed into a preamplifier stage, whose purpose is to increase the received signal power to a level (voltage) compatible with the next section of the receiver front end. The amplifier stage can comprise a single or several amplifiers mounted in series, the first amplifier, called the Low Noise Amplifier (LNA), is usually characterized by a small noise figure (NF). In some receiver architectures, the LNA can also be integrated into the antenna. - The signal in RF (e.g., 1575.42MHz for GPS C / A signal) is transmitted at intermediate frequency f IF,Tx (which does not need to be related to IF,RxThe signal is first downconverted to a baseband signal. This downconversion is usually performed in an analog stage (before the analog-to-digital converter ADC), but can also be performed in a digital stage if the sampling frequency of the ADC is large enough (according to the Nyquist criterion). To perform the downconversion, the signal at RF is multiplied or mixed with a cosine or curve function at the [RF-IF] frequency (e.g. 1575.42 MHz-IF). Before or after the actual mixing, the signal bandwidth is limited (tuned to a dedicated signal or band) by a filter stage to reduce the effect of interference, reject the image of the mixing process, and avoid spurious signals. It is outlined that some radio receiver architectures propose to directly downconvert the received and amplified signal, not to an intermediate frequency, but at baseband. In this case, the term direct downconversion is applied to describe this type of receiver architecture. - Then, once downconverted at IF, the signal is sampled and digitized by an ADC. It is common to call the part before the ADC the "analog front end" and the part after the ADC the "digital front end". - To optimize the quantization within the dynamic range of the ADC, an automatic gain control (AGC) monitors the power level of the samples and provides information to a downconverter block, e.g. as shown in Figure 3, to multiply the received signal at RF with a variable gain. - Finally, the output of the ADC is fed into a blanker whose purpose is to set the samples containing large interfering signals at the highest points of the received signal to, for example, 0. In Fig. 3, the proposed blanker is implemented in the digital front end, but other receiver implementations are also proposed to implement the blanker in the analog front end. - The digital samples are then injected into W digital receiver channels, where W represents the number of processed radio signal components determined to be necessary by the radio receiver manufacturer to estimate the corresponding pseudoranges needed to estimate the receiver's position fix, as well as for a given target position and timing estimation performance (usually the more channels the better the position accuracy). Thus, a compromise is applied between positioning and timing estimation performance and receiver complexity (which depends directly on the number of W digital receiver channels). Each digital receiver channel transmits one of the remaining carrier frequencies f IF,Rx The aim is to process the IF signal by cancelling the IF signal and feeding the different correlator channels at baseband, which are necessary not only for signal acquisition but also for code and carrier estimation and, finally, for navigation data demodulation. Cancellation is done by applying a complex exponential function exp(2πjf IF,Rx ×t) It is noted that this stage of downconversion is not necessary when the RF signal is directly downconverted to baseband.

[0010] [Reference 2]: "On the Detection Probability of Parallel Code Phase Search Algorithm in GPS Receivers?", BCGeiger, 21st Annual IEEE International Symposium on Personal,Indoor and Mobile Radio Communications

[0011] [Reference 3]: "Resource and Performance Comparisons for Different Acguisition Methods that can be Applied to a VHDL-based GPS Receiver in Standalone and Assisted Cases", J.Leclere, IEEE2010

[0012] Having described the overall characteristics of the GNSS radio signal components as well as the overall functional blocks of the GNSS satellite transmitter and the GNSS receiver, it is possible to present the acquisition process of the radio signal components transmitted by the GNSS satellites, which represents one of the most important and demanding processing steps for a GNSS radio receiver potentially integrated in a user device. The acquisition of the GNSS radio signal components, also called signal acquisition, is in fact the very first step of the receiver processing chain, which is necessary not only to estimate the pseudoranges between the transmitter and the radio receiver used for the position determination of the radio receiver, but also to retrieve the data modulated onto the radio signal components, if applicable.

[0013] Signal acquisition aims to estimate with a coarse accuracy the distance and Doppler offset between the transmitting GNSS satellite and the radio receiver. As mentioned before, the radio signal components are modulated with a spreading code made up of (binary) symbols called chips. The coarse distance estimate is therefore reduced to determine which chip, or more precisely which fraction of a chip, of the total spreading code the radio receiver belongs to on an "imaginary line" starting at the transmitting satellite and ending at the radio receiver, in other words to determine the coarse spreading code delay or shift that needs to be applied to the undelayed version of the spreading code (called spreading code replica) known at the radio receiver to match the actual received radio signal components including the delayed spreading code. In the following, the term code hypothesis will also be applied to denote the aforementioned spreading code delays that are to be coarsely estimated. The Doppler offset, which also needs to be roughly estimated, is the result of the relative movement between the GNSS satellites and the radio receiver, resulting in a shift or offset of the actual or observed received carrier frequency to its negligible value (e.g. equal to 1575.42 MHz for GPS C / A or Galileo E1-B or E1-C radio signal components). In the following, the term Doppler will often be used instead of Doppler offset for simplicity. When considered necessary, a distinction will be made regarding the use of Doppler or Doppler offset, depending on the context.

[0014] In the following, the actual coarse distance expressed in chips is considered to belong to the code uncertainty region, which includes many possible code hypotheses that will have to be tested by the radio receiver during the acquisition process. In the absence of a priori information, the code uncertainty region covers the entire period of the spreading code, i.e. the spreading code length. Similarly, the correct Doppler is considered to belong to the Doppler uncertainty region, which includes many possible Doppler hypotheses that will have to be tested by the radio receiver during the acquisition process. In the absence of such a priori information, the acquisition is operated with a "cold start", which represents the most challenging acquisition scenario. Signal acquisition may also use some a priori information to narrow down the uncertainty region of both the spreading code delay and the Doppler offset. The corresponding a priori information in spreading code delay and Doppler can be estimated from the latest position fix if this is not too old or if it is provided to the receiver / UD by an external source, in which case the term assisted acquisition is usually intended. Therefore, during the acquisition process, candidate code and Doppler hypotheses in the uncertainty region are tested to determine the correct values ​​that are closest to the actual spreading code delay and Doppler offset.

[0015] The granularity of the code hypotheses, also called the code bin width, is defined as the minimum distance (usually expressed in chips) between two candidate code hypotheses. Similarly, the granularity of the Doppler hypotheses, also called the Doppler bin width, is defined as the minimum distance (usually expressed in Hz) between two candidate Doppler hypotheses. Both the code bin width and the Doppler bin width are intrinsic to the acquisition process and directly affect the performance as well as the complexity of the acquisition search process. Furthermore, depending on the acquisition algorithm and the hardware setup available for this search, the different tests of spreading code and Doppler hypotheses are performed simultaneously (in this case parallel acquisition is intended) or sequentially (in this case sequential acquisition is intended).

[0016] The complexity of the hardware required for signal acquisition typically depends on the number of physical memories (i.e. "cells"), the memory usage with the number of charge / discharge cycles per unit time, and finally the number of operations (multiplications and additions). The two latter contributions will affect the power consumption of the user device hosting the radio receiver, and therefore its autonomy due to the embedded battery, while the first contribution exacerbates the cost of the chipset.

[0017] Therefore, if the hardware setup is not configured to test all code and Doppler hypotheses simultaneously, different chunks of hypotheses have to be tested successively. This process slows down the acquisition time, defined as the time required to roughly estimate both the correct code and Doppler hypotheses, which represents the most representative figure of merit for acquisition performance. This highlights the trade-off between hardware complexity and acquisition performance.

[0018] To obtain a rough estimate of the code delay and Doppler offset of the radio signal component of interest, Hyp,Code code delay hypotheses, and N Hyp,Dopp A set of N Doppler offset hypotheses is tested. Hyp,Code (or N Hyp,Dopp ) represents the number of code delay (or Doppler) hypotheses to be tested over the code (or Doppler) uncertainty region.

[0019] When no a priori information about the code delay between the satellite and the receiver is available (as in the case of a "cold start" acquisition scenario), the code delay uncertainty region typically corresponds to one spreading code period (expressed in units of seconds) and is expressed as N Hyp,Codeis proportional to the spreading code length L (for GNSS radio signal components with periodic spreading codes), which is equal to L=1023 chips, corresponding to a spreading code period of 1 ms for GPS C / A signals, and equal to L=4092 chips, corresponding to a spreading code period of 4 ms for Galileo E1-B and E1-C signals. Similarly, in the absence of a priori information on the observed Doppler of the satellite-user link, Hyp,Dopp The Doppler hypotheses must be examined over a Doppler uncertainty region that can cover up to ±10 KHz (see [Reference 3]). This span primarily takes into account the Doppler dynamics of the satellite-user link, which ranges from ±3.5 KHz to ±4.5 KHz depending on the altitude of the MEO satellite orbit, as well as the satellite and user clock (in)stabilities, which are known to be dominated by the user clock stability. [Reference 3] shows that for medium to low grade receiver clocks (±3.5pp), the contribution of the receiver clock stability to the (observed) Doppler uncertainty region is equal to ±5.25 KHz.

[0020] When deductions are available, the actual code uncertainty region and Doppler uncertainty region to be examined are usually narrower, which means that for the same sensitivity, the code hypotheses N Hyp,Code and / or Doppler Hypothesis N Hyp,Dopp or conversely, keeping the same number of code or Doppler hypotheses in the case of improved acquisition performance, for example, by reducing the spreading code and / or Doppler search granularity (i.e., bin width).

[0021] For each code and Doppler hypothesis pair, a replica is first generated at the receiver side by shifting the spreading code corresponding to the satellite transmitting the radio signal component of interest with the code delay to be tested, thus resulting in a so-called spreading code replica. This spreading code replica is then multiplied by a carrier modulated with the Doppler hypothesis to be tested in order to generate the replica. The received signal is then integrated over a so-called coherent integration time Tint The received signal is correlated with this replica generated for the code and Doppler hypothesis to be tested by calculating the integral of the product between the replica and the received signal over the entire signal, yielding a complex correlation value. Finally, the corresponding complex correlation value is typically modulus-squared to construct an acquisition detector (equivalent to a power estimator). One common detection strategy consists in comparing the detector output with a threshold that is usually set to the desired Probability of False Alarm (PFA) caused by unavoidable thermal noise. Another strategy consists in comparing the ratio between the maximum and the second-largest detector output, calculated over the entire uncertainty region of the code and Doppler hypotheses, with another threshold that is also set according to some specified PFA level (see [Reference 2]). Such an alternative strategy is particularly suitable for parallel acquisition methods (i.e., when several detector outputs corresponding to several code and Doppler test hypotheses are simultaneously available). A typical trade-off that is primarily encountered when designing acquisition algorithms is to limit the PFA level to have a higher detection threshold but still satisfy the Probability of Detection (PD), obtained when the code and Doppler hypotheses closely match the actual code and Doppler of the radio signal components of interest contained in the received signal; in other words, by minimizing the Probability of Mis-Detection (PMD).

[0022] To improve acquisition performance, it is possible to add the squared absolute values ​​of successive correlation values ​​between the received signal and replicas, generated with the same spreading code delay and Doppler offset hypotheses to be tested. This process is defined as noncoherent summation. The number of noncoherent summations is N NC It is called the product N NC ×T int is called the dwell time. Due to the incoherent summation, squared losses must be taken into account. To avoid such squared losses, the coherent correlation time T intIt is possible to increase the BER, but this will result in increased Doppler shift losses in the correlation process. The selected configuration of the acquisition scheme (spreading code delay and Doppler bin width, coherent correlation time, number of non-coherent sums, PFA and PMD rates, ...) is set by the receiver manufacturer to match the specified acquisition performance (i.e. acquisition time) of the supported location application, and also to take into account the complexity and power consumption of the radio receiver or user device hardware.

[0023] The determination of the number of code and Doppler hypotheses within each uncertainty region, in other words the granularity of the code and Doppler hypotheses, i.e. the code and Doppler bin width, is usually determined by the so-called code and Doppler offset losses, which will affect the size of the detector. In practice, the code (or Doppler) bin width is equal to the code (or Doppler) uncertainty region divided by the number of code (or Doppler) hypotheses. The search space granularity solves a trade-off between the complexity (the finer the code and Doppler bin width, the larger the number of hypotheses to be tested and the higher the computational load) and the code and Doppler offset losses (the losses are the highest when the code and Doppler hypotheses are the most distant from the actual code and Doppler hypotheses). Thus, in order to limit the code and Doppler offset losses during the correlation process to a few dB (typically 1 to 3 dB), a typical value of the code hypotheses is T (applicable for BPSK waveforms). c / 2 or T c / 4, and typical values ​​for the Doppler bin width are 50 to 100 Hz. As an illustration, for a GPS CA signal modulated with the BSPK(1) waveform and having a spreading code length of 1023 chips, the Doppler bin width must be at least N Hyp,Code = 2046 code hypotheses can be tested. It should be noted that these values ​​are given only as an indication of magnitude, and the previous quantitative settings will depend on the actual deductions, the specifications of the complexity of the radio receiver, as well as the actual characteristics of the radio signal components, such as pulse waveforms (i.e. BPSK, BOC), etc.

[0024] FIG. 4 illustrates an acquisition architecture, having the functional blocks previously described with respect to a conventional wireless receiver, and based on the digital receiver channel shown in FIG.

[0025] FIG. 5 represents the “ideal” Auto-Correlation Function (ACF) for an unfiltered BPSK(1) pulse waveform used to detect the presence of the radio signal component of interest for the proposed code and Doppler hypotheses. This ideal ACF is obtained by correlating “only” the received signal containing the radio signal component of interest (i.e., no other interfering radio signal components or additive noise) with the spreading code replica corresponding to the satellite transmitting the radio signal component of interest for different values ​​of code delay to be tested. Furthermore, and for simplicity of expression, it is considered that the Doppler hypothesis to be tested matches the actual Doppler exactly in such a way that the ACF is real. In this illustrative example, two hypotheses are tested per chip, T c This results in a code bin width equal to T c We also highlight the worst case alignment between the actual position of the correlation function and the code hypothesis grid, which corresponds to a code delay equal to half of T c It is evaluated for a code delay of / 4.

[0026] The methods for sharing the same resources, such as a spectrum region, by different sources, i.e. different radio signal components transmitted by GNSS satellites, are called multiple access schemes. Two main multiple access schemes are currently used to avoid MAI between different radio signal components received simultaneously from different GNSS satellites:

[0027] The first multiple access scheme is Code Division Multiple Access, further called CDMA, where the GNSS radio signal components transmitted by Q different GNSS satellites at the same carrier frequency are modulated with a different spreading code for each GNSS satellite. If we again consider that the radio receiver sees V satellites transmitting radio signal components one by one (F=1), the number of received radio signal components is P v = V. The radio signal component of interest is therefore P v The received signal, which includes the sum of the received radio signal components, possible multipaths and additive noise, is correlated at the radio receiver with the replica, the generation process of which was previously described in the context of the acquisition phase, to obtain the other (P V-1) from interfering radio signal components. To avoid interference between different received GNSS radio signal components, orthogonal spreading codes are required to guarantee sufficient cross-correlation isolation. Cross-correlation isolation is a measure of the mismatch between the replica corresponding to the radio signal component of interest and the received and interfering radio signal components. Cross-correlation isolation is calculated based on a Cross-Correlation Function (CCF), taking into account not only the spreading code orthogonality, but also the pulse shape of both aforementioned signal types, and finally the interfering Doppler residual as the difference between the Doppler offset hypothesis and the actual observed Doppler offset of the interfering radio signal component. Cross-correlation isolation is usually expressed in dB applied to one of the maximum CCF values ​​calculated over all possible code and Doppler hypotheses, but it can also be calculated based on other statistics of the distribution of CCF values. As mentioned before, typical spreading code lengths are L = 1023 chips for GPS C / A signals and L = 4092 chips for Galileo E1-C signals. If a longer spreading code length L is beneficial in terms of code cross-correlation isolation, this still forces an increase in the number of code hypotheses to be tested during the signal acquisition phase, since the code uncertainty region is proportional to the spreading code length in the "cold start" acquisition mode. Furthermore, in the absence of information about the V satellites that are actually visible out of the total Q satellites of the GNSS system, the radio receiver will have to agnostic consider Q different spreading codes in order to generate the spreading code replica later. For example, methods based on some a priori information about the very coarse location of the radio receiver on the Earth's surface, W acq This allows the actual number of different spreading codes that need to be fed into the different acquisition channels, further referred to as Q, to be reduced to less than Q. A typical number of acquisition channels (i.e., different spreading codes) is W, which varies between 8 and 12 for the very small number of 24 GPS satellites. acq However, W acq W may be equal to the total number of satellites Q (e.g., 24) for wireless receivers that benefit from more computational resources.acq does not have to be equal to the number of W wireless signal components of interest that are subsequently processed to estimate the pseudoranges, and is typically set at receiver initialization such that W≦W acq It is emphasized that the reason is that the modules including the acquisition engine and the tracking loop for PR estimation are separate in a typical receiver architecture. acq Considering the different spreading code replicas that will be generated for each acquisition channel, as well as, on the other hand, the number of tested code hypotheses per spreading code replica, introduces complexity to the GNSS receiver since more memory is required to process the received signal and to belong to the code uncertainty region that is proportional to the spreading code length L. When further considering the impact of CDMA schemes on the receiver complexity, it can also be stated that the required bandwidth of the receiver front end that guarantees a minimum performance can be reduced to the main lobe width of the spectral representation of the radio signal components, which corresponds to twice the chip rate (2 × N × f0) for the BPSK(N) waveform, e.g. 2.046 MHz for the GPS C / A signal. For BOC(M,N) modulated (CDMA) signals, this minimum bandwidth increases to (2 × (M+N) × f0), e.g. 4.092 MHz for the Galileo (BOC(1,1)) E1-C signal.

[0028] The second multiple access scheme is frequency division multiple access, further called FDMA, where radio signal components transmitted by different GNSS satellites are transmitted on different carrier frequencies, each carrier frequency depending on each GNSS satellite. Interference between different GNSS radio signal components is avoided by strictly separating the different transmitted GNSS radio signal components in the frequency domain. To reduce MAI, the separation of the carrier frequencies of the transmitted radio signal components is designed such that there is no overlap between the spectral occupancy of signals of the same type, and thus each FDMA signal occupies an FDMA frequency channel that has no elements in common with the frequency channels of other FDMA signals of the same type. Since the spectral occupancy of each radio signal component transmitted by a GNSS satellite is typically expressed in MHz, this means that the spectral separation between the different radio signal components transmitted by a GNSS satellite is also expressed in MHz. A direct consequence is that to guarantee a sufficiently low MAI between the Q respective radio signal components, a rather wide bandwidth, including several MHz, is required for the transmission and reception of the different GNSS radio signal components, which also introduces complexity in the GNSS radio receiver, since the analog front end needs to support the acquisition of radio signals for this wide bandwidth. If the FDMA scheme also allows to efficiently reduce the MAI thanks to the frequency separation, it is still necessary to modulate the corresponding carrier with a spreading code, as in the case of CDMA, to guarantee a sufficient performance for the PR estimation between the transmitting satellite and the receiver. Thus, chip rates on the order of magnitude of MCps are still required for FDMA, which also accounts for the order of magnitude in MHz for the spectral occupation of the FDMA channel. One example of an operational FDMA scheme is implemented in the Russian GLONASS GNSS for the transmission of L1OCd radio signal component type, employing 20 carrier frequencies spaced apart by 562.5 KHz between 1609.3125 MHz and 1598.0625 MHz in the "L1 sub-band", and 20 other carrier frequencies spaced apart by 437.5 KHz between 1251.6875 MHz and 1242.9375 MHz in the "L2 sub-band".Each of the GLONASS L1OCd radio signal components employs a BPSK(0.5) waveform, and therefore f. c = 0.5MCps chip rate and includes a spreading code of 511 chips. One major and recognized drawback of traditional FDMA schemes is that they force GNSS receivers to implement an analog receiver bandwidth proportional to the number of FDMA carrier frequencies or channels. In the case of GLONASS, the total receiver bandwidth can therefore cover between 10MHz and 20MHz (depending on the receiver grade as well as the number of W processed signals required to achieve the target positioning performance). Figure 6 shows one example of an FDMA spectrum allocation for Q = 12 radio signal components and their 12 different FDMA carrier frequencies and corresponding channels. In this illustrative example (different from the GLONASS signal structure), each FDMA channel corresponding to a different signal component modulated with BPSK(1) occupies a 4MHz bandwidth (with a main lobe having a 2MHz bandwidth and first left and right secondary lobes having 1MHz bandwidth each).

[0029] It is proposed here to present the main effect of MAI on acquisition performance. (P V In the presence of F = 1, the effective correlation function therefore has replicas corresponding to the wireless signal components of interest generated for the code and Doppler hypotheses to be tested, with their respective code delays and observed Doppler offsets (P V 5, each of which is obtained as the correlation between the (P V-1) so-called cross-correlation functions (CCFs). The presence of additional CCF values ​​will therefore affect both the probability of false alarms and false detections. When the code and Doppler hypotheses to be tested do not correspond to the real ones, the magnitude of the acquisition detector will be raised due to the presence of undesired and additional CCF values. A detection threshold derived from the PFA, which assumes only thermal noise, will therefore result in a substantial increase in the PFA. Similarly, when the code and Doppler hypotheses to be tested correspond to the real ones, the CCF values ​​can be negatively added to the main peak of the autocorrelation function when the phase of the interfering radio signal component is opposite to the phase of the radio signal component of interest. This will reduce the acquisition detector magnitude and therefore increase the PMD for an acquisition detection threshold set when only thermal noise is considered.

[0030] The previous discussion therefore highlights that the signal characteristics of the multiple access schemes adopted in current GNSS result in a non-optimal solution for a trade-off between, firstly, on the receiver side, a small impact on the receiver hardware such as low computational complexity, small memory resources and front-end bandwidth of the receiver hardware, secondly, on the transmitter side, a small transmitter bandwidth and finally, in terms of MAI performance, a sufficient cross-correlation performance, i.e. cross-correlation isolation between the replica corresponding to the signal component of interest and the interfering signal component.

[0031] An overview of the main advantages and disadvantages between the existing CDMA and FDMA multiple access schemes applied in current GNSS represents evidence of this non-optimality. On the one hand, the receiver bandwidth capturing the CDMA GPS C / A radio signal component can be as low as 2.046 MHz, but the spreading code length is very large, 1023 chips. On the other hand, the spreading code length of the GLONASS signal component is smaller, 511 chips, but the receiver bandwidth processing the FDMA GLONASS L1OCd signal is as wide as 12.276 MHz, much larger than the receiver bandwidth used to process the CDMA GPS C / A radio signal component. Due to the MAI, the CDMA code length and the frequency offset of the FDMA signal cannot be reduced.

[0032] WO2019048341A1 additionally discloses a method and associated device for signal acquisition of a generalized binary offset carrier G-BOC modulated signal including K basic BOC(Mk,Nk) modulated signals. The method comprises: sc,kThe method includes extracting lower and upper sideband signals for each of the K basic BOC(Mk,Nk) modulated signals by up-converting and down-converting the respective basic BOC(Mk,Nk) modulated signals and by low-pass filtering the respective up-converted and down-converted basic BOC(Mk,Nk) modulated signals during use. The method includes correlating the extracted lower and upper sideband signals, respectively, with replicas modulated with a basic pulse shape included in generating the BOC(Mk,Nk) modulated signals, such as a binary phase shift keying BPSK pulse shape. The replicas have a chip rate, e.g., fc,k=Nk×f0, and a spreading sequence according to the respective basic BOC(Mk,Nk) modulated signal. The chip rate can be expressed as chips per second, abbreviated as Cps. The spreading sequence can be a binary noise-like sequence. The replicas are generated based on a code delay hypothesis and a Doppler frequency hypothesis. The method includes multiplying each correlation result by a complex or complex conjugate multiplication. The method further includes generating a first detector output by adding corresponding K complex or K complex conjugate multiplications and / or generating a second detector output by adding NNC successive detector outputs, each of which is a sum of K complex or K complex conjugate multiplications. [Prior art documents] [Patent documents]

[0033] [Patent Document 1] WO2019048341A1 [Non-patent literature]

[0034] [Non-Patent Document 1] "Understanding GPS Principles and Application", Kaplan Artech House Publishing, ISBN0-89006-793-7 [Non-Patent Document 2] "On the Detection Probability of Parallel Code Phase Search Algorithm in GPS Receivers?", BCGeiger, 21st Annual IEEE International Symposium on Personal,Indoor and Mobile Radio Communications [Non-Patent Document 3] "Resource and Performance Comparisons for Different Acguisition Methods that can be Applied to a VHDL-based GPS Receiver in Standalone and Assisted Cases", J.Leclere, IEEE2010 [Non-Patent Document 4] "Correlation and Randomness Properties of the Spreading Coding Families for the Current and Future GNSSs", F.Soualle, Signal Workshop at DLR, Oberpfaffenhofen, December 2009 [Non-Patent Document 5] "Impact of Galileo Spreading Code Selection and Data Rate onto Navigation Signal Interference", F. Soualle, ION GPS / GNSS 2003, September 9-12, 2003, Portland, Oregon [Non-Patent Document 6] "Quantization", Allen Gersho, IEEE Communications Society Magazine, pp. 16-28, September 1977, doi:10.1109 / MCOM.1977.1089500 [Non-Patent Document 7] "Section1.2.3: Quantization, in Chapter 1: JPEG 2000 Core Coding System (Part 1)", In Schelkens, Peter; Skodras, Athanassios; Ebrahimi, Touradj (eds.), The JPEG 2000 Suite, John Wiley&Sons, pp. 22-24, ISBN 978-0-470-72147-6 [Non-Patent Document 8] "Efficient Scalar Quantization of Exponential and Laplacian Random Variables", Gary J. Sullivan, IEEE Transactions on Information Theory, Vol. IT-42, No. 5, pp. 1365-1374, September 1996, doi:10.1109 / 18.532878 [Non-Patent Document 9] “Resource Efficient Parallel Acquisition Architecture for Modernized GNSS Signals”, J.Leclere, ECOLE POLYTECHNIQUE FEDERALE DE LAUSANNE, THESE NO6190(2014) [Non-Patent Document 10] "Performance Analysis of Modernized GNSS Signals Acquisition", M.Foucras.Thesis, 2015.Ecole Nationale d'Aviation Civile(ENAC) Summary of the Invention

[0035] It is an object of an embodiment of the present invention to provide a method for acquisition of at least one radio signal component of interest among a plurality of radio signal components, an associated radio navigation system, an associated radio transmitter and a radio receiver for such a radio navigation system of the above-mentioned known type, wherein the above-mentioned shortcomings or drawbacks of the known solutions are mitigated or overcome.

[0036] In particular, it is an objective of embodiments of the present invention to reduce the required bandwidth at both the transmitter and receiver sides, and further to reduce the required power consumption as well as the computational complexity and required memory of the wireless receiver, without negatively affecting mutual access interference.

[0037] In practice, the objective is to reduce the frequency offset Δf for 1≦p≦P. carr,p A data signal with carrier frequency f car,p =f central +Δf carr,p To obtain a data signal carrier having a center carrier frequency f central Frequency offset Δf for 1≦p≦P carr,p to the data signal carrier of said radio signal component by each said radio transmitter prior to transmission of each radio signal component, said frequency offset corresponding to said radio signal component among a plurality of signal components, said frequency offset value being smaller than the chip rate of a spreading code, such that the distance between spectral spikes, taking into account the possibly symbol modulation of the spectral representation of the different received radio signal components, is small, and therefore the overall bandwidth required to transmit the radio signal component per transmitter as well as the plurality of radio signal components combined, across all transmitters of a radio navigation system, is small. P , and the frequency offset Δf carr,p is the radio signal component S p This is regarding.

[0038] Additionally, the number of chips L, also called the spreading code length, is determined by the spreading code length L in the reference numerator and the chip duration (T c ) and the total number P of radio signal components of the radio navigation system in a reference denominator does not exceed 25, in combination with the use of a data signal component including a short spreading code having a predetermined number L of chips of said data signal component, and oscil,Rx a Doppler hypothesis to be tested that also takes into account the Doppler hypothesis (D TRX1 ,...,D TRXP ), the occupied bandwidth per radio signal component of the plurality of radio signal components at the time of transmission can be small, and therefore the total bandwidth required to transmit said plurality of radio signal components is smaller compared to known systems. Applying the previous measure usually leads to short spreading codes, for example comprising a few tens of chips, which on the one hand makes it possible to reduce the number of code hypotheses to be tested, and on the other hand makes it possible to reduce the required receiver front-end bandwidth, both of which make it possible to simplify the receiver hardware complexity as well as the power consumption. Having a smaller bandwidth at the transmission side is also advantageous, since it reduces the hardware and implementation complexity. Having a small spectrum occupation for all transmitters of a radio navigation system also represents an advantage in frequency access and procurement, i.e. costs, in the regulatory framework.

[0039] Thereafter, a radio receiver R for the acquisition of at least one radio signal component, i.e. the radio signal component of interest, is xThe at least one wireless signal component of interest is transmitted from a plurality of wireless transmitters T X1 ...T XQ a plurality of radio signal components S1..S transmitted by respective radio transmitters of P , wherein said plurality of radio signal components S1..S are selected as a basis for estimating the code delay and Doppler offset of the received signal. P Each of the radio signal components includes a data signal carrier for carrying a data signal component, the data signal component including a spreading code including a predetermined number L of chips, each chip of the spreading code being of duration T c where the duration is the chip rate R c The reciprocal of the radio receiver R x is the stability D oscil,Rx and the plurality of wireless transmitters T X1 ...T XQ The speed V for each radio transmitter TRX1 ...V TRXQ and a plurality of radio signal components S1..S P Each of the radio signal components is transmitted to the radio receiver R for the corresponding radio transmitter. x and the observed Doppler D as a function of the data signal carrier frequency of the data signal carrier corresponding to each of the radio signal components. TRX1 ...D TRXP A plurality of radio signal components S1..S P Based on the received radio signal components among the X1 ...T XQ the plurality of radio signal components S1..S Pa data signal component modulated onto the data carrier of the corresponding radio signal component by comparing a detection threshold to a capture detector, the capture detector being based on at least one cross-correlation value obtained for a tested code and Doppler hypothesis of a set of code and Doppler hypotheses to be tested, the correlation value being generated by correlating a received signal including the plurality of received radio signal components with a local replica of the data signal component including a spreading code corresponding to the signal component of interest by the radio receiver Rx over a coherent integration time, the correlation value comprising a combination of a cross-correlation function value corresponding to an interfering signal component and one auto-correlation function value corresponding to the signal component of interest.

[0040] a Doppler hypothesis to be tested for at least one wireless signal component of interest; TRXp (not including the radio signal component of interest D TRX1 ...D TRXPThe difference between the observed Doppler for each further interfering signal component, not including the Doppler component, and extended with a frequency offset corresponding to said interfering radio signal component, will have a direct impact on the cross-correlation value calculated between the local replica and each of the interfering radio signal components. Taking into account the reproduction of the cross-correlation values ​​for some specific intervals of the Doppler difference depending also on the spreading code length, it has been practically shown that there is an overall trend between the Doppler magnitude and the reduction of the cross-correlation value. Thus, by including a frequency offset between the different transmitted radio signal components, which is a frequency offset that can also be interpreted as an artificial Doppler, the overall Doppler span of the possible Doppler differences between the tested Doppler for the received radio signal component of interest and each of the interfering radio signal components extended with a frequency offset corresponding to said interfering radio signal component will be expanded, making it possible to reduce the cross-correlation value and therefore improving the cross-correlation isolation. In this way, as a result of the reduction in the required bandwidth as a result of the low offset frequency applied to the data signal carrier of each radio signal component, in combination with the low predetermined number of chips of the spreading code, an optimal radio signal component acquisition is obtained, which makes it possible to estimate the distance and Doppler between the transmitting GNSS satellite and the radio receiver with a rough accuracy without affecting the complexity and power consumption of the receiver. Another advantage of the small spectrum occupation for the entire transmitted signal is therefore that the bandwidth of the receiver front end is also small, making it possible to reduce the hardware cost and complexity of the analog section of the receiver. In conclusion, the smaller bandwidth, accompanied by this low predetermined number of chips of the spreading code, makes it possible to reduce the complexity of the radio receiver, which affects the power usage and the memory required.

[0041] An embodiment of the present invention relates to a method for acquisition of at least one radio signal component of interest, the method comprising: filtering the received signal comprising the plurality of received radio signal components with the spreading code and further with the frequency offset (Δf carr,p , 1≦p≦P) with a local replica of the vector p.

[0042] An alternative embodiment of the present invention relates to a method for acquisition of at least one radio signal component of interest, the method comprising: determining from a received signal comprising the plurality of received radio signal components the respective frequency offsets (Δ fcarr,p , 1≦p≦P) by the radio receiver prior to the step of correlating, and carr,p , 1≦p≦P) has been cancelled out, with a local replica of the data signal component including the spreading code corresponding to the signal component of interest.

[0043] A further embodiment of the invention is such that the spreading code is modulated with symbols, each symbol having a coherent integration time used to calculate the cross-correlation of the symbol duration T s This implies that the duration T can be limited by s The present invention relates to a method for the acquisition of at least one radio signal component, characterized in that it comprises:

[0044] One advantage is that it enables the provision of symbols having radio signal components which can be used to improve the performance of the position and timing estimation of a radio receiver without significantly degrading acquisition performance, further providing the advantages previously described and in particular reduced receiver hardware complexity together with a reduced spectral footprint on transmission.

[0045] In order not to degrade the cross-correlation properties obtained with short spreading codes, which are also a function of Doppler difference, it is important to ensure that the symbol duration is greater than the coherent integration time used to calculate the cross-correlation function. If this constraint is not met, satisfactory cross-correlation properties would be destroyed, again resulting in unacceptable MAI levels.

[0046] A further characterizing embodiment of the present invention is a radio transmitter further comprising: said frequency offset applying means further applying a central carrier frequency (f central ) with respect to the frequency offset (Δf carr,p , 1≦p≦P) to the data signal carriers of the radio signal components.

[0047] Thus, by applying this first alternative for implementing a frequency offset by multiplying an appropriate analog complex or real offset carrier according to said frequency offset, taking into account the possibly symbolic modulation of the spectral representation of the different received radio signal components, the distance between the spectral spikes is small and therefore the overall bandwidth required to transmit a radio signal component per transmitter as well as combined, across all transmitters of a radio navigation system, is small.

[0048] Another characterizing embodiment of the present invention is a radio transmitter, comprising: a frequency offset application means for applying a central carrier frequency f by multiplying a quantized complex or real offset carrier according to said frequency offset; central Frequency offset Δf carr,p , 1≦p≦P applies to the data signal carriers of the radio signal components.

[0049] Thus, by applying this further alternative for implementing a frequency offset by multiplying an appropriate quantized complex or real offset carrier according to said frequency offset, taking into account the possibly symbolic modulation of the spectral representation of the different received radio signal components, the distance between the spectral spikes is small and therefore the overall bandwidth required to transmit a radio signal component per transmitter as well as combined, across all transmitters of a radio navigation system, is small.

[0050] Furthermore, a quantized version of the real or complex carrier offset can be implemented in a simpler manner for satellite navigation generation chains where the digital stage applies a high sampling frequency. Reducing the number of quantization levels of the generated signal waveform makes it possible to work with a higher sampling frequency and to apply, for example, a direct up-conversion at the final RF carrier using a suitable digital-to-analog converter. One extreme case for a digital or quantized implementation of the carrier offset is the sign(sin(2πΔf carr,p ), or when considering the complex binary carrier offset, [sign(cos(2πΔf carr,p )+j×sign(sin(2πΔf carr,p )], where sign(x) is a function that provides the sign of the argument x.

[0051] Another alternative embodiment of the invention is a radio transmitter further comprising a frequency offset application means for applying a central carrier frequency f by modulating a secondary NRZ code of complex or real symbols according to said frequency offset onto a spreading code. central Frequency offset Δf carr,p , 1≦p≦P to the data signal carriers of the radio signal components.

[0052] Thus, by applying this further alternative for implementing a frequency offset by modulating a secondary NRZ code of complex or real symbols according to said frequency offset into a spreading code, taking into account the possibly symbolic modulation of the spectral representation of the different received radio signal components, the distance between the spectral spikes is small and therefore the overall bandwidth required to transmit a radio signal component per transmitter as well as combined, as said multiple radio signal components across all transmitters of the radio navigation system, is small.

[0053] One advantage of modulating a secondary NRZ code of complex or real symbols onto a spreading sequence is that the offset is no longer considered as a carrier, but as a modified spreading code. The result is that the corresponding offset can be implemented using a similar scheme used to modulate the symbols or secondary codes onto the spreading code that takes the role of the primary code.

[0054] Moreover, a further advantage of this alternative implementation of the frequency offset comes from the magnitude of the frequency offset, which is smaller than the chip rate and allows several chips per NRZ code to be included, which corresponds to a differentiator, when compared to existing schemes such as FDMA, where the frequency offset is larger than the chip rate and several NRZ codes are included within each chip.

[0055] Another alternative embodiment of the invention provides a method for applying a frequency offset to a central carrier frequency f by modulating a longer spreading code obtained as a combination of the secondary NRZ code of complex or real symbols according to the frequency offset modulated by the spreading code. central Frequency offset Δf carr,p , 1≦p≦P to the data signal carriers of the radio signal.

[0056] Thus, by applying this further alternative for implementing a frequency offset by modulating a longer spreading code, obtained as a combination of an NRZ code of complex or real symbols according to said frequency offset, with a spreading code, taking into account the possibly symbolic modulation of the spectral representation of the different received radio signal components, the distance between the spectral spikes is smaller and therefore the overall bandwidth required to transmit a radio signal component per transmitter as well as combined, across all transmitters of the radio navigation system, is smaller.

[0057] One advantage of modulating the longer spreading code as a combination of the original spreading code and an NRZ code on complex or real symbols is that the complete and longer spreading code can be stored in a memory on board the satellite payload, and the symbols of the longer spreading code are read out cyclically from the memory.

[0058] Yet another alternative embodiment of the invention relates to a radio receiver arranged, using processing means, to generate said local replica by multiplying said spreading code by a complex or real offset carrier depending on said frequency offset.

[0059] Thus, by generating a local replica with a frequency offset obtained by multiplying the spreading code with a complex or real offset carrier, the correlation can be performed in a simpler manner already in the digital domain or stages, as long as the sampling frequency allows this operation.

[0060] Yet another alternative embodiment of the invention relates to a radio receiver configured, using processing means, to generate said local replica by multiplying said spreading code by a quantized complex or real offset carrier depending on said frequency offset.

[0061] Thus, by multiplying the quantized complex or real offset carrier, it is possible to reduce the number of quantization levels, simplifying the digital implementation. One extreme case for a digital implementation is when considering a real binary carrier offset, sign(sin(2πΔf carr,p ), or when considering the complex binary carrier offset, [sign(cos(2πΔf carr,p )-j×sign(sin(2πΔf carr,p )], where sign(x) is a function that provides the sign of the argument x. Such a two-level carrier format is particularly suitable for loose digital implementations.

[0062] Yet another alternative embodiment of the invention relates to a radio receiver configured to generate said local replica by modulating, using processing means, a secondary NRZ code of complex or real symbols that depends on said frequency offset onto said spreading code.

[0063] In this way, by multiplying the received signal by a local replica with a frequency offset obtained by modulating a secondary NRZ code of complex or real symbols with said spreading code, correlation can be performed in a simpler manner already within the digital stage, as long as the sampling frequency allows this operation.

[0064] Yet another alternative embodiment of the invention relates to a radio receiver configured to generate the local replica by generating, using processing means, a longer spreading code obtained by combining with the spreading code a secondary NRZ code of complex or real symbols depending on the frequency offset.

[0065] Thus, by multiplying the received signal by a local replica with a frequency offset obtained by combining a secondary NRZ code of the complex or real symbol with the spreading code to generate a longer spreading code, the correlation can be performed in a simpler manner already in the digital stage, as long as the sampling frequency allows this operation. One advantage of modulating a longer spreading code as a combination of the original spreading code of the complex or real symbol with an NRZ code is that the complete and longer spreading code can be stored in a memory in the radio receiver, and the symbols of the longer spreading code are read out cyclically from the memory.

[0066] Yet another alternative embodiment of the invention relates to a radio receiver configured to cancel said frequency offset from said received signal by using processing means (22) to multiply said received signal by a complex or real offset carrier which is dependent on said frequency offset.

[0067] In this way, it is possible to implement standard techniques applied for FDMA schemes to multiply the received signal with a complex or real offset carrier to cancel the frequency offset, and then to correlate the output signal obtained from said multiplication with a spreading code replica comprising only the spreading code. Furthermore, the proposed alternative embodiment allows to apply a frequency offset smaller than the chip rate, making it possible to reduce the receiver bandwidth.

[0068] Yet another alternative embodiment of the invention relates to a radio receiver configured, using processing means (22), to cancel the frequency offset from the received signal by multiplying the received signal by a quantized complex or real offset carrier which is responsive to the frequency offset.

[0069] In this way, multiplying the captured data signal by a quantized complex or real offset carrier, by multiplying the radio signal, simplifies the implementation at the receiver side and, in particular, by applying frequency cancellation in the digital stage for receivers where the sampling frequency is sufficient.

[0070] The invention will be further explained with the aid of the following description and the accompanying figures. [Brief description of the drawings]

[0071] [Figure 1] 1 illustrates a navigation signal structure for a radio signal component consisting of a periodic binary sequence having L chips and modulated with symbols, together with a method for generating such a radio signal component. [Diagram 2] 1 is a diagram illustrating a general signal generation chain applicable to typical GNSS radio signal components and implemented on board a radio navigation satellite. [Diagram 3] FIG. 1 shows the main functional blocks of a radio receiver used as part of a GNSS together with the main processing steps. [Figure 4] FIG. 1 illustrates a capture architecture having the functional blocks required to generate a capture detector from a received signal that includes a wireless signal component. [Diagram 5] 1 illustrates an autocorrelation function for an unfiltered BPSK(1) waveform and used to detect the presence of a wireless signal component of interest. [Figure 6] 1 illustrates an example FDMA spectrum allocation including Q=12 wireless signal components, each having a 4 MHz bandwidth FDMA channel. [Figure 7] 1 is a diagram illustrating a system for acquisition of a radio signal component of interest from a plurality of radio signals at a radio receiver, the radio navigation system comprising a plurality of radio transmitters and a radio receiver. [Figure 8]FIG. 2 shows functional elements of a radio transmitter TXq and a radio receiver RX according to an embodiment of the present invention. [Figure 9a] This figure shows the distribution of observed Doppler offsets for GNSS radio navigation satellites, such as the Galileo satellites, and for a ground user latitude of 0°, due to the relative motion of the satellite and the user. [Figure 9b] This figure shows the distribution of observed Doppler offsets for GNSS radio navigation satellites, such as the Galileo satellites, and for a ground user latitude of 20°, due to the relative motion of the satellite and the user. [Figure 9c] This figure shows the distribution of observed Doppler offsets due to the relative motion of the satellites and the user for GNSS radio navigation satellites such as the Galileo satellites, and for a ground user latitude of 51°. [Figure 9d] This figure shows the distribution of observed Doppler offsets due to the relative motion of the satellites and the user for GNSS radio navigation satellites such as the Galileo satellites, and for a ground user latitude of 82°. [Figure 9e] This diagram shows the distribution of observed Doppler offsets for GNSS radio navigation satellites such as Galileo, and across all latitudes, due to the relative motion of the satellite and the user. [Figure 10] 3 is a diagram representing the span of the distribution of the interfering (or desired) Doppler residual ΔfDopp,test,p (or ΔfDopp,test,0), calculated as ΔfDopp,test,p=fDopp,p-fDopp,test,0, where fDopp,p (or fDopp,0) represents the actual and observed Doppler offset for the pth (or 0th) interfering radio signal component, while fDopp,test,0 represents the Doppler offset hypothesis to be tested for the radio signal component of interest. [Figure 11a]The values ​​of the autocorrelation or cross-correlation functions are shown in the upper part and the corresponding distributions in the lower part, respectively, showing the auto- and cross-correlation functions when applying 0 Hz Doppler residual between the received radio signal components and the replica. [Figure 11b] The values ​​of the autocorrelation or cross-correlation functions are shown in the upper part and the corresponding distributions in the lower part, respectively, showing the auto- and cross-correlation functions when applying 0 Hz Doppler residual between the received radio signal components and the replica. [Figure 11c] The values ​​of the autocorrelation or cross-correlation functions are shown in the upper part and the corresponding distributions in the lower part, respectively, showing the auto- and cross-correlation functions when applying a 3500 Hz Doppler residual between the received radio signal components and the replica. [Figure 11d] The values ​​of the autocorrelation or cross-correlation functions are shown in the upper part and the corresponding distributions in the lower part, respectively, showing the auto- and cross-correlation functions when applying a 3500 Hz Doppler residual between the received radio signal components and the replica. [Figure 12a] 1 shows the variation of the maximum ACF value expressed in dB as a function of the desired Doppler residual (i.e., the Doppler coarse estimation error) when applying a coherent integration time of 1 ms, as well as the percentiles of the ACF values, representing the 99%, 95% and 67% percentiles of the ACF distribution, as the desired Doppler residual varies between -8KHz and 8KHz. [Figure 12b] 1 shows the variation of the maximum ACF value expressed in dB as a function of the desired Doppler residual (i.e., Doppler coarse estimation error) when the desired Doppler residual varies between −8 KHz and 8 KHz, when applying a coherent integration time of 5 ms, as well as the percentiles of the ACF values, representing the 99%, 95% and 67% percentiles of the ACF distribution. [Figure 12c]1 shows the variation of the maximum ACF value expressed in dB as a function of the desired Doppler residual (i.e., the Doppler coarse estimation error) when the desired Doppler residual varies between -8KHz and 8KHz, when applying a coherent integration time of 1 ms, as well as the percentiles of the CCF values, representing the 99%, 95% and 67% percentiles of the ACF distribution. [Figure 12d] 13 shows the variation of the maximum ACF value expressed in dB as a function of the desired Doppler residual (i.e., the Doppler coarse estimation error) when the desired Doppler residual varies between -8KHz and 8KHz, when applying a coherent integration time of 5ms, as well as the percentiles of the CCF values, representing the 99%, 95% and 67% percentiles of the ACF distribution. [Figure 13a] Figure 14 shows the cumulative CCF distribution together with the corresponding percentiles when the CCF is calculated for an interferometric Doppler residual ΔfDopp,test,p equal to 0 Hz and obtained for a family of Q=24 spreading codes similar to the Gold codes applied by GPS C / A signals. [Figure 13b] Figure 13 shows the cumulative CCF distribution along with the corresponding percentiles when the CCF is calculated for an interferometric Doppler residual ΔfDopp,test,p equal to 2500 Hz for the same code family. [Figure 13c] The 99%, 95%, and 67% percentiles are provided over a Doppler span ranging from 0 Hz to 10 KHz, along with the corresponding maximum value variation. [Figure 14a] 13 shows the cumulative CCF distribution along with the corresponding percentiles when the CCF is calculated for an interference Doppler residual ΔfDopp,test,p equal to 0 Hz and obtained for a family of Q=24 spreading codes with a shorter spreading code length L=341 chips. [Figure 14b]Figure 13 shows the cumulative CCF distribution along with the corresponding percentiles when the CCF is calculated for an interferometric Doppler residual ΔfDopp,test,p equal to 10460 Hz for the same code family. [Figure 14c] The 99%, 95%, and 67% percentiles are provided over a Doppler span ranging from 0 Hz to 10 KHz, along with the corresponding maximum value variation. [Figure 15a] 13 shows the cumulative CCF distribution along with the corresponding percentiles when the CCF is calculated for an interference Doppler residual ΔfDopp,test,p equal to 0 Hz and obtained for a family of Q=24 spreading codes with a shorter spreading code length L=31 chips. [Figure 15b] Figure 13 shows the cumulative CCF distribution along with the corresponding percentiles when the CCF is calculated for an interferometric Doppler residual ΔfDopp,test,p equal to 2500 Hz for the same code family. [Figure 15c] The 99%, 95%, and 67% percentiles are provided over a Doppler span ranging from 0 Hz to 10 KHz, along with the corresponding maximum value variation. [Figure 16] This is a diagram of the spectral representation, or PSD, of a GNSS signal generated with a spreading code having a very long or infinite length (L → ∞) modulated with a BPSK(1) waveform. [Figure 17a] 1 is a diagram of the spectral representation, or PSD, at two different scales of a GNSS pilot radio signal component modulated with a BPSK(1) waveform and having a spreading code length of 1023 chips. [Figure 17b] 1 is a diagram of the spectral representation, or PSD, at two different scales of a GNSS pilot radio signal component modulated with a BPSK(1) waveform and having a spreading code length of 1023 chips. [Figure 18a]Spectral representation, or PSD, at two different scales of a GNSS radio signal component for data transmission modulated with a BPSK(1) waveform and with a spreading code length of 1023 chips, where a symbol duration of Ts=20 ms is modulated onto the spreading code, resulting in a symbol rate of fs=50 sps. [Figure 18b] Spectral representation, or PSD, at two different scales of a GNSS radio signal component for data transmission modulated with a BPSK(1) waveform and with a spreading code length of 1023 chips, where a symbol duration of Ts=20 ms is modulated onto the spreading code, resulting in a symbol rate of fs=50 sps. [Figure 19a] 1 shows the PSD of Pv GNSS pilot radio signal components transmitted at the same carrier frequency fcentral and having a spreading code length L=31 chips, received at a ground-based GNSS receiver. [Figure 19b] 1 shows the PSD of Pv GNSS pilot radio signal components transmitted at the same carrier frequency fcentral and having a spreading code length L=31 chips, received at a ground-based GNSS receiver. [Figure 20a] FIG. 13 shows the PSD of Pv GNSS radio signal components for a data transmission with a symbol rate fs=50 sps, transmitted at the same carrier frequency fcentral, having a spreading code length L=31 chips, and received at a ground-based GNSS receiver. [Figure 20b] FIG. 13 shows the PSD of Pv GNSS radio signal components for a data transmission with a symbol rate fs=50 sps, transmitted at the same carrier frequency fcentral, having a spreading code length L=31 chips, and received at a ground-based GNSS receiver. [Figure 21] FIG. 13 shows graphical elements for demonstrating the relationship between different Doppler-dependent percentiles of the CCF (shown in the lower part) as a function of the interference Doppler residual (shown in the upper part) and the spectral representation (i.e., PSD) of the radio signal components of the (received and) interference signal and of the replica signal. [Figure 22] FIG. 13 shows a graphical element to explain the principle of allocation of two CDDMA carriers based on the percentile variation of the distribution of CCFs calculated with a coherent integration time of 1 ms and considering radio signal components generated with two spreading codes having the same length of 31 chips, according to the (extended) interference Doppler residual corresponding to the satellite-radio receiver link dynamic for MEO satellites such as Galileo, and also taking into account the frequency offset. [Diagram 23] The figure further shows graphical elements to explain the principle of allocation of 4 CDDMA channels and associated carriers based on the percentile variations of the distribution of CCFs calculated with a coherent integration time of 1 ms and considering radio signal components generated with 4 different spreading codes having the same length of 31 chips, according to the (extended) interference Doppler residual corresponding to the satellite-radio receiver link dynamic for MEO satellites such as Galileo satellites and also taking into account the frequency offset. [Figure 24a] This figure represents the PSD for the two main multiple access schemes and the FDMA scheme proposed for four FDMA channels with 4 MHz occupancy. [Figure 24b] The figure shows the PSD for the two main multiple access schemes and the CDDMA scheme proposed for four CDDMA channels. [Figure 24c] The figure shows the PSD for the two main multiple access schemes. The CDDMA scheme is proposed for four CDDMA channels, but the smaller overall spectral occupancy is due to the interleaving of spikes due to the spectral occupancy of the wireless signal components when observed on a smaller scale. [Diagram 25]25A and 25B show distributions of percentiles of CCF obtained for spreading code length L1=31 chips (upper part of FIG. 25) or for spreading code length L2=33 chips (lower part of FIG. 25). These distributions make it possible to prove the allocation of CDDMA channels for wireless signal components with different spreading code lengths. [Figure 26a] FIG. 1 shows a graphical element for calculating the cross-correlation value between two radio signal components having different spreading code lengths of 31 and 33 chips, respectively, for fc=1MCps, when calculated over a coherent integration time of 1 ms, corresponding to 1023 chips, representing two chip streams corresponding to the radio signal components and constructed as a concatenation of their respective elementary spreading codes over a duration corresponding to the coherent integration time. [Figure 26b] FIG. 1 shows a graphical element for calculating the cross-correlation value between two radio signal components having different spreading code lengths of 31 and 33 chips, respectively, for fc=1MCps, when calculated over a coherent integration time of 1 ms corresponding to 1023 chips; and FIG. 2 shows the Doppler dependent percentiles of the distribution of CCFs calculated between Q=24 concatenated basic spreading codes (randomly generated) having lengths of 31 and 33 chips. [Figure 26c] Figure 1 shows a graphical element for calculating the cross-correlation value between two radio signal components with different spreading code lengths of 31 and 33 chips, respectively, for fc=1MCps, when calculated over a coherent integration time of 1 ms, corresponding to 1023 chips; Figure 2 shows the Doppler dependent percentiles of the distribution of CCFs calculated between Q=24 spreading codes (also randomly generated, i.e. not Gold codes), with a length of 1023 chips. [Figure 27]Figure 1 shows a graphical element to explain the principle of allocation of P=34 CDDMA carriers based on the variation of percentiles of the distribution of the autocorrelation function calculated with a coherent integration time of 1 ms and considering radio signal components generated with the same spreading code having a length of 31 chips and taking into account the frequency offset, according to the (extended) interference Doppler residual corresponding to the satellite-radio receiver link dynamic for MEO satellites such as Galileo. The spreading code is an m-sequence generated with the polynomial "o45". [Figure 28] FIG. 13 shows graphical elements to explain the principle of allocation of P=17 CDDMA carriers based on the Doppler variation of percentiles of the distribution of the cross-correlation function calculated with a coherent integration time of 1 ms and considering radio signal components generated with two different spreading codes both having a length of 31 chips, according to the (extended) interference Doppler residual corresponding to the satellite-radio receiver link dynamic for MEO satellites such as Galileo, and also taking into account the frequency offset. [Figure 29] The diagram shows an implementation for modulating symmetric NRZ onto a spreading code with L=31 chips and applying a frequency offset of 33KHz when the falling edge of NRZ falls in the middle of the 15th chip of the spreading code. [Diagram 30] 30 shows an alternative implementation for application of a 33 KHz frequency offset by modulating an asymmetric NRZ onto a spreading code having a length L=31 chips and when the falling / rising edges of the asymmetric NRZ correspond to the edges of the spreading code. Figure 30 can also support understanding of the implementation of the 33 KHz frequency offset when considering the modulation of a secondary NRZ code onto the spreading code. [Diagram 31] The diagram shows an implementation for modulating symmetric NRZ onto a spreading code with length L=31 chips and for applying a frequency offset of 33 / 2KHz when the falling edge of the NRZ coincides with the edge of one chip of the spreading code. [Diagram 32] FIG. 13 illustrates an implementation for modulating asymmetric NRZ onto a spreading code with length L=31 chips and for applying a frequency offset of 24.75 KHz when the falling / rising edges of the asymmetric NRZ correspond to the edges of the spreading code. [Figure 33a] 13 shows percentile Doppler variation for ACF and CCF distributions when applying three different implementation options of a frequency offset of 0 KHz (ACF). [Figure 33b] 13 shows percentile Doppler variation for ACF and CCF distributions when applying three different implementation options of a frequency offset of 8.25 KHz (CCF). [Figure 34a] 1 shows percentile Doppler variation for the distribution of CCF when applying three different implementation options of a frequency offset of 16.5KHZ (CCF). [Figure 34b] 13 shows percentile Doppler variation for the distribution of CCF when applying three different implementation options of a frequency offset of 24.75 KHz (CCF). [Figure 35a] 13 shows percentile Doppler variation for the distribution of CCFs when applying three different implementation options of a frequency offset of 33KHz (CCF). [Figure 36a] FIG. 13 illustrates an implementation of a high-level model for an FFT-based acquisition architecture for a spreading code with length L=1023 chips. [Figure 36b] FIG. 13 illustrates an adapted implementation of a high-level model for an FFT-based acquisition architecture for another spreading code having a shorter length L'=31 chips. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0072] Modes for carrying out the present invention The present invention will be described with respect to particular embodiments and with reference to certain drawings but the invention is not limited thereto but only by the claims. The drawings described are only schematic and are non-limiting. In the drawings, the size of some of the elements may be exaggerated and not drawn to scale for illustrative purposes. The dimensions and relative dimensions do not necessarily correspond to actual reductions to the practice of the invention.

[0073] Moreover, the terms first, second, third, and the like in the description and claims are used to distinguish between similar elements and are not necessarily intended to describe a sequential or chronological order. The terms are interchangeable under appropriate circumstances, and embodiments of the invention are capable of operating in sequences other than those described or illustrated herein.

[0074] Moreover, the terms top, bottom, over, under, and the like in this description and claims are used for purposes of description and not necessarily to describe relative positions. The terms so used are interchangeable under appropriate circumstances, and the embodiments of the invention described herein are capable of operation in orientations other than those described or illustrated herein.

[0075] The term "comprising" used in the claims should not be interpreted as being limited to the means listed thereafter, nor excluding other elements or steps. It does not have to be interpreted as specifying the presence of the stated features, integers, steps or components as mentioned, but does not exclude the presence or addition of one or more other features, integers, steps or components, or groups thereof. Thus, the scope of the expression "a device comprising means A and B" should not be limited to a device consisting only of components A and B. This means that in the context of the present invention, the relevant components of the device are only A and B.

[0076] In the following paragraphs, with reference to the drawing in Fig. 7, an implementation of a radio navigation system for acquisition of at least radio signal components of interest in a radio receiver is described, the radio navigation system comprising a plurality of radio transmitters for transmitting the plurality of radio signal components and a radio receiver according to an embodiment of the invention. In the further paragraphs, all relationships between the mentioned elements are defined.

[0077] Thereafter, a plurality of wireless transmitters TX1...TX2 according to an embodiment of the present invention as presented in FIG. Q Radio transmitter and radio receiver R x All relevant functional means of the present invention are described, followed by a description of all interrelationships of these functional means.

[0078] In the following paragraphs a practical implementation of a radio navigation system for acquisition of at least a radio signal component of interest out of a plurality of radio signal components in a radio receiver according to an embodiment of the present invention is described.

[0079] The radio navigation system consists of several radio transmitters TX1...TX Q each radio transmitter is connected to at least one radio receiver R of the radio navigation system; x The radio receiver R is configured to transmit at least one radio signal component to the x based on at least a certain number of the radio signal components received at the radio receiver R x can determine, inter alia, the location of the radio receiver based on information derived from the received radio signal components.

[0080] Such a radio receiver may be a navigation device, which is a device comprising a processor coupled to a memory and interface means such as a display and a keyboard, or a GNSS receiver integrated into a user device, such as a personal mobile device, such as a smartphone. Such a user device may then be configured to install many different types of applications, the execution of each such application being intended to perform a different type of task, such as navigation. A radio navigation system according to an embodiment of the invention may be a satellite radio navigation system, such as a Global Navigation Satellite System (GNSS), or a single positioning beacon, such as a pseudolite, or a network of positioning beacons, or a terrestrial system, such as a wireless communication network, that requests synchronization with a user terminal or device. An embodiment of the invention is further applicable to wireless communication systems, where MAI also represents a degradation effect to be overcome. Details of this enhanced framework of application are provided later in this document.

[0081] A first essential element of a radio navigation system comprises a plurality of radio transmitters TX1...TX Q Of the wireless transmitters, TX q and the radio transmitter transmits at least one radio signal component to the radio receiver R, in particular for navigation and synchronization purposes. x The device is configured to send

[0082] wireless transmitter tx q First, the frequency offset application means (11) is provided, and the frequency offset application means (11) applies a frequency offset Δf carr,p , data signal carrier frequency (f car,p =f central +Δfcarr,p), the center carrier frequency f central Frequency offset Δf carr,p, 1≦p≦P to each wireless signal component to be transmitted before transmission. Each wireless signal component is denoted by S p where p can have a value between 1 and P, and the same index p is used to carry the data signal components D p The data signal components comprise a spreading code, the spreading code having a predetermined number of chips L, each chip of the spreading code having a chip rate f c Duration T equal to the reciprocal of c has.

[0083] A frequency offset corresponds to a radio signal component of the plurality of radio signal components, each radio signal component having its own dedicated frequency offset. The frequency offset applying means is configured to apply a frequency offset value that is less than the chip rate of the spreading code. The intended spreading code is a product B×L×T c Δf ΔP / P contains a predetermined number of chips L such that P / P does not exceed 25, where B represents the bandwidth of the receiver front end in Hz. Each data signal carrier is assigned a frequency offset, and the frequency offset Δf carr,p , 1 ≤ p ≤ P is the frequency offset Δf carr,p , data signal carrier frequency (f car,p =f central +Δf carr,p ) to obtain a center carrier frequency f central This applies to.

[0084] Such a frequency offset application means 11 may in particular comprise a microprocessor for processing the radio signal components to be transmitted, and the processing means may further comprise a memory device coupled to said microprocessor for storing electronic information such as computer instructions, results of the signal processing including final and intermediate results, and further information.

[0085] wireless transmitter tx q comprises additionally signal modulation means (12) further configured to modulate a data signal component including a spreading code, optionally modulated with a symbol, onto said data signal carrier of said radio signal component having a dedicated frequency offset.

[0086] Such a frequency offset application means (11) may, for example, apply a complex or real offset carrier corresponding to the frequency offset to the central carrier frequency f central Frequency offset (Δf carr,p , 1≦p≦P) to the data signal carriers of the radio signal components.

[0087] Such a frequency offset application means (11) may, for example, apply a central carrier frequency f by multiplying a quantized complex or real offset carrier according to the frequency offset. central Frequency offset (Δf carr,p , 1≦p≦P) to the data signal carriers of the radio signal components.

[0088] Such a frequency offset application means (11) modulates a secondary NRZ code of complex or real symbols corresponding to the frequency offset into a spreading code, thereby applying a central carrier frequency f central Frequency offset (Δf carr,p , 1≦p≦P) to the data signal carriers of the radio signal components.

[0089] Such a frequency offset application means (11) applies a central carrier frequency f by modulating a longer spreading code, obtained for example as a combination of the secondary NRZ code of complex or real symbols according to the frequency offset modulated by the spreading code. central Frequency offset (Δf carr,p , 1≦p≦P) to the data signal carriers of the radio signal components.

[0090] This radio transmitter TX q further comprises transmitting means (13) configured to transmit said at least one radio signal component, each radio signal component carrying a data signal component modulated onto a data signal carrier of said corresponding radio signal component, such data signal component comprising a spreading code, the spreading code comprising a predetermined number L of chips, each chip of said spreading code having a duration T c Such a spreading code is optionally modulated with a symbol.

[0091] The data signal carrier may apply a waveform to modulate the spreading code or the primary code, for example with Binary Phase Shift Keying (BPSK) in the case of GPS C / A signals, or with Binary Offset Carrier (BOC) in the case of Galileo E1-B / -C or GPS L1C.

[0092] Radio transmitter TX1...TX x Each of the q Note that it has the same functional structure as

[0093] Such a radio navigation system according to an embodiment of the invention comprises a plurality of radio signal components S1..S P a radio receiver R configured to capture at least one radio signal component of interest among x said plurality of radio signal components S1..S P The wireless signal components of X1 ...T XQ The signal is transmitted by a respective wireless transmitter among the plurality of wireless transmitters.

[0094] The plurality of radio signal components S1..S PEach such radio signal component of includes a data signal carrier for carrying a data signal component of the corresponding radio signal component, the data signal component including a spreading code, the spreading code including a predetermined number L of chips, each chip of the spreading code being of duration T c has.

[0095] The wireless receiver (R x ) includes the plurality of wireless transmitters (TX1...TX Q ) for each radio transmitter (VT RX1 ...VT RXQ ).

[0096] Radio receiver R X may be any kind of device incorporating a GNSS receiver.

[0097] Such a radio receiver R x comprises first a receiving means (21) arranged to receive said plurality of radio signal components, each of said radio signal components being coupled to said radio receiver (R) for said corresponding radio transmitter. x ) and an observed Doppler offset D as a function of the speed of the data signal carrier of the data signal carrier corresponding to each of the radio signal components. TRX1 ...D TRXP has.

[0098] Such a radio receiver R x is the receiver front-end bandwidth B, and stability D oscil,Rx The present invention has an oscillator or clock having a

[0099] Radio receiver R x detects the data signal components modulated onto the data signal carrier of the corresponding radio signal component by comparing a detection threshold with a capture detector for the plurality of radio signal components S1 ..S Pand a processing means (22) configured to acquire from each of the at least one radio signal component of interest of a set of code and Doppler hypotheses to be tested, the acquisition detector being based on at least one correlation value obtained for a tested code and Doppler hypothesis of a set of code and Doppler hypotheses to be tested, the correlation value being generated by correlating a received signal comprising the plurality of received radio signal components with a local replica of the data signal component comprising a spreading code corresponding to the signal component of interest by the radio receiver (Rx) over a coherent integration time using the processing means (22), the correlation value comprising a combination of cross-correlation function values ​​corresponding to interfering radio signal components and one auto-correlation function value corresponding to the signal component of interest.

[0100] According to one embodiment of the present invention, a radio receiver R x The processing means (22) of the carr,p , 1≦p≦P), of a received signal comprising the plurality of received wireless signal components.

[0101] According to an alternative embodiment of the invention, a radio receiver R x The processing means (22) extracts from a received signal comprising a plurality of received radio signal components the respective frequency offsets (Δf carr,p , 1≦p≦P) prior to the step of correlation, and the processing means (22) is further configured to correlate the received signal comprising the plurality of received radio signal components, from which the respective frequency offsets (Δf carr,p , 1≦p≦P) are cancelled and the local replica of the data signal component includes a spreading code corresponding to the signal component of interest.

[0102] wireless transmitter T X1The frequency offset application means (11) is coupled with an output terminal to an input terminal of a signal modulation means (12), which in turn is coupled with an output terminal to an input terminal of a transmission means (13), which in turn is coupled with a radio transmitter T Xq The output terminal is O1.

[0103] The radio receiver Rx has an input terminal I1 which is at the same time an input terminal of receiving means (21) which in turn is coupled with an output terminal to an input terminal of processing means (22).

[0104] For the purposes of describing the embodiments of the present invention, it is assumed that the user equipment required to estimate the user position and time is based solely on a radio receiver processing radio signal components transmitted by satellites belonging to the same radio navigation system, such as GNSS, and that the user equipment therefore does not comprise any additional sensors, such as a barometer, accelerometer, or gyrometer, as part of an inertial measurement unit (IMU). This simplifying assumption does not affect the overall applicability of the embodiments of the present invention, which is nevertheless extendable to other user equipment with additional sensors on top of the radio receiver. Furthermore, the following conventions shall all be applied in line with the description of the embodiments: - Q represents the number of radio transmitters, such as satellites, included in a radio navigation system, such as GNSS. - F represents the number of radio signal components transmitted by each radio transmitter of the radio navigation system. P represents the total number of radio signal components transmitted by the Q radio transmitters, where P=F×Q. - V represents the number of radio transmitters visible to a radio receiver. - P v represents the number of received radio signal components, and P v =F×V. - W represents the maximum number of received and processed radio signal components for pseudo-range estimation that can be implemented in a radio receiver and considered sufficient to reach a target or a specified position and timing estimation performance. - W acq represents the maximum number of received and captured radio signal components implemented in a radio receiver, for example to take into account limited radio receiver complexity.

[0105] Based on these assumptions and conventions, at least four Lines-of-Sight (LoS) are required to determine the user position and time: three for the three X, Y, and Z user coordinates, and an additional one for the user clock offset (time). Because obstacles typically block visibility to satellites, it is necessary to have more visible satellites so that an unobstructed horizon guarantees that at least four LoS are substantially available for an obstructed horizon (e.g., with a user mask angle from 5° to 15° in altitude). Furthermore, position accuracy has been shown to be inversely proportional to the number of LoS (i.e., visible satellites) processed, assuming a uniform spatial distribution of the LoS. Finally, and contrary to the Regional Satellite Navigation Systems (RNSS), all major GNSSs provide a Global Positioning Service, i.e. the ability to position anywhere on Earth with a given accuracy. Combining the aforementioned factors, the number Q of radio navigation satellites, flying in medium orbit at altitudes of 19E3 to 23E3 km, typically amounts to Q=24 to 30 satellites to meet both availability and location performance requirements.

[0106] In the following embodiments of the present invention, the GPS C / A radio signal component will be frequently referred to as a representative radio signal component applying the CDMA scheme. A quantitative performance comparison, such as an evaluation of cross-correlation isolation or receiver complexity, will also be provided between the performance achieved with the scheme proposed in the embodiments of the present invention and the performance obtained with the GPS C / A signal. For a similar evaluation, the GLONASS L1COd radio signal component will also be proposed as a representative radio signal component applying the FDMA scheme. To support the aforementioned comparison, a similar number of radio transmitters, i.e. satellites, as one of the GPS or GLONASS constellations having approximately the same altitude, will also be considered for the scheme described in the embodiments of the present invention to propose a first dimension and allocation of frequency offsets. Nevertheless, it has already been outlined that the embodiments of the present invention are not limited to this operating point used for comparison, and that the embodiments of the present invention are applicable for other dimension parameters, such as pulse shape, number and speed of radio transmitters, or MAI level at which a specified cross-correlation isolation should be guaranteed.

[0107] In the following, the radio receiver R is used to support the acquisition of the signal components of interest. x It is proposed to explain in more detail the mathematical background of the correlation process applied in the method of the present invention. More specifically, the received signal (s(t)) including the plurality of received radio signal components is divided into a local replica of the data signal component including the spreading code corresponding to the signal component of interest and a local replica of the data signal component including the spreading code corresponding to the signal component of interest by the radio receiver (R x ) coherent integration time T int to yield a correlation value, said correlation value including a combination of a cross-correlation function value corresponding to an interfering wireless signal component and one auto-correlation function value corresponding to a wireless signal component of interest.

[0108] Thus, two main types of correlation functions will be identified. The first correlation function is the autocorrelation function (ACF) obtained when correlating the radio signal component of interest contained in the received signal s(t) with a replica matching the corresponding radio signal component of interest. The second correlation function is the cross-correlation function (CCF) obtained when correlating each of the interfering radio signal components also contained in the received signal s(t) that are different from the radio signal component of interest and are unrelated to the radio signal component of interest. These CCF values ​​corresponding to all interfering radio signal components contained in the received signal will constitute an acquisition degradation factor and will represent the MAI effect during the acquisition process.

[0109] To support these formal descriptions, the following equations express the p-th transmitted radio signal component u at the output of an antenna mounted on the radio transmitter platform, i.e., the satellite's substrate: P Provide a model for (t).

number

[0110] For each P transmitted navigation radio signal components with index p, -

number

number

number

number

[0111] Reusing the convention introduced earlier, if a radio receiver is in view of V satellites each transmitting F radio signal components, and one of these V visible satellites transmits a signal component of interest, the correlation output will include (V-1) x F + (F-1) values ​​of the cross-correlation function corresponding to (V-1) x F + (F-1) different interfering radio signal components, and one value of the autocorrelation function corresponding to the radio signal component of interest. For ease of explanation, in the following it is considered that only one signal component is transmitted by each satellite (F = 1). Nevertheless, the following evaluation and proof of the proposed scheme can be extended to the case where F is 2 or more.

[0112] For these simplified assumptions, the received signal s(t) is v = F × V = V (≦ P) transmitted radio signal components u p The following model is proposed for the received signal s(t), using complex notation and obtained after direct downconversion to baseband in the radio receiver:

number

[0113] P with index p v For each received radio signal component, -

number

number

number

[0114] The previous equation represents a two-stage downconversion (i.e., RF to IF (f IF,Rx), and then IF(f IF,Rx It is finally pointed out that the received signal obtained after ) to baseband can also be modeled.

[0115] To capture the radio signal components of interest, and in particular to roughly estimate the code delay and Doppler offset, the received signal s(t) is passed through a matched filter. This filter is used to estimate the Doppler offset hypothesis f Dopp,test,0 and the delay τ corresponding to the frequency modulation test,0 After application of , the received signal is correlated with the replica corresponding to the radio signal component of interest. According to convention and for simplicity, it is considered that the index p of the signal component of interest is equal to p=0. The formula for the correlation function is:

number

number

number

number

number

[0116] The correlation operation (4) is applied to the digital signal, i.e. after the ADC module previously described. For simplicity, the different delays τ p and the estimated delay τ test,0 is the sampling interval T, which corresponds to the code bin width in the case of capture. samp (For example, T samp =T c / 2 or T c / 4). Furthermore, and for ease of explanation, the sampling rate f samp is equal to the chip rate, or a multiple of the chip rate, and f samp =U×f c These simplifications are usually applied in the technical literature. Nevertheless, the proof and the main conclusions of the proposed scheme do not change when these simplifying assumptions are waived.

[0117] [Reference 4]: "Correlation and Randomness Properties of the Spreading Coding Families for the Current and Future GNSSs", F.Soualle, Signal Workshop at DLR, Oberpfaffenhofen, December 2009

[0118] By correlating the received signal s(t), two main types of correlation functions are therefore to be distinguished:

[0119] The first correlation function, called the autocorrelation function (ACT), indicates how well the received radio signal component transmitted by the desired satellite and its corresponding replica corresponding to the radio signal component of interest match together for a given tested code delay and Doppler offset.

number

number

number

number

[0120] moreover,

number

number

[0121] For all m, no symbols are modulated onto the radio signal components.

number

[0122] In equation (5), the term

number

number

[0123] Dependence of the received radio signal component of interest on the received power

number

number

number

[0124] The second correlation type is called the cross-correlation function (CCF) and is the function that is used to compare the other (V-1)=(P V - indicates how much a received interfering radio signal component transmitted by one of the satellites (1) mismatches a replica corresponding to the radio signal component of interest. The formula is:

number

number

number

[0125] In equation (9), the term

number

number

[0126] Similarly, the received power dependence of the received interference radio signal component

number

number

number

[0127] It is noted that the formula of the cross-correlation function derived previously applies for an interfering radio signal component having the same pulse shape p(t) as the radio signal component of interest, which is a simplifying assumption for the explanation of the embodiment of the present invention. Nevertheless, a similar statement can be derived when extending the cross-correlation function to an (interfering) radio signal component having a different shape from the radio signal component of interest. In this case, the formula of the (normalized) cross-correlation function (equation (10)) should be derived when the autocorrelation pulse shape function (ACPWF) introduced when deriving the ACF function formula (see equations (5) and (6)) is replaced by the cross-correlation pulse shape function (CCPWF). When the shape p(t) of the radio signal component of interest is different from the shape q(t) of the interfering radio signal component, the formula of the CCPWF becomes:

number

[0128] The formula for the cross-correlation function applying the CCPWF is therefore:

number

[0129] In the following description of embodiments of the present invention, the term cross-correlation function will be applied in a broad sense, i.e., when the pulse waveform q(t) of an interfering radio signal component is either identical (p(t)=q(t)) or different (p(t)≠q(t)) from the waveform of the radio signal component of interest.

[0130] The cross-correlation function is f Dopp,0 The observed Doppler offset of the signal component of interest, also called D TRx0 as well as the wireless receiver (R x ) according to the speed

number

[0131] As far as the first contribution covering link dynamics is concerned, the relative displacement of a satellite of index p with respect to the GNSS receiver or user device will cause a time-varying shift of the received carrier frequency with respect to the carrier frequency of the transmitted signal. The general expression for the observed Doppler offset due to the satellite-receiver link dynamics and experienced at the terrestrial radio receiver is given by the following equation:

number

number

number

[0132] It is also noted that Doppler due to link dynamics does not solely affect the carrier by shifting it, but will also affect other characteristics of the received signal, such as a time warping effect on the spreading code. In effect, Doppler may expand or compress the apparent chip duration when compared to its nominal duration at the time of transmission. Nevertheless, such effects are of minor importance to acquisition performance (and in particular to the correlation losses used for acquisition) when compared to the Doppler effect on the carrier.

[0133] The assignment of frequency offsets to the radio signal components is deeply coupled to the nature of the cross-correlation function depending on the spreading code and in particular its length L, the type (i.e. family) of the applied pulse waveforms p(t) and q(t), as well as the type (i.e. family) of the physical Doppler observed at the radio receiver, as shown in equation (10). Thus, one first step is to determine the span of the observed Doppler (offset) for the transmitter-receiver link. In the following example, the determination of the Doppler span is proposed for GNSS satellites located at Medium Earth Orbit (MEO) altitude. Nevertheless, the same method for estimating the Doppler span is applicable for satellites at other altitudes (such as Geostationary Earth Orbit (GEO) or Low Earth Orbit (LEO)), or possibly other platform types such as HAPS, drones, and UAVs, or possibly transmitters stationary with respect to the receiver.

[0134] Figures 9a, 9b, 9c, 9d and 9e aim to determine, thanks to equation (Eq. 13), the span and distribution of the observed Doppler offset, i.e., that perceived at the radio receiver and due only to the relative motion of the satellite and the user when considering terrestrial and fixed users and GNSS satellites such as Galileo satellites with altitude ~23E3km and inclination 56°. Figures 9a, 9b, 9c and 9d represent the Doppler span and distribution of the Doppler offset of users at latitudes 0°, 20°, 51° and 82°, respectively. Figure 9e below represents the distribution aggregated over all latitudes. From these results, it is possible to determine the actual received and observed Doppler offsets of the satellites, f Dopp,qIt can be verified that for the Galileo system, the Doppler span can vary between -3.5 KHz and +3.5 KHz. The same magnitude can be stated for other GNSS with MEO satellites. For example, for the GPS system, the Doppler span is equal to ±4.9 KHz, from [Reference 3]. Furthermore, it is also shown in [Reference 3] that this Doppler span of the satellite-user link will increase by ±190 Hz for a user moving at 130 km / h, which indicates that the main contribution to the received and observed Doppler comes from the satellite dynamics (here the contribution of the received oscillators is not considered).

[0135] Figure 10 shows the Δf Dopp,test,p =f Dopp,p -f Dopp,test,0 The interferometric Doppler residual Δf calculated as Dopp,test,p represents the span of the distribution of the desired Doppler residual Δf Dopp,test,0 =f Dopp,0 -f Dopp,test,0 Therefore, the following description provided for the coherent Doppler residual can also be made applicable for the desired Doppler residual. Dopp,test,0 is assumed to be uniformly distributed over the span of the physical Doppler offset, which is the case when considering a cold acquisition scenario where there is no a priori knowledge of the actual Doppler offset of the satellite transmitting the radio signal component of interest (all that is available a priori is the maximum Doppler span, e.g. ±3.5 KHz). The interferometric Doppler residual Δf Dopp,test,pThe distribution of is thus obtained by convolving the distribution of the tested Doppler hypotheses applied to the replica with the distribution of the actual and observed Doppler offsets, shown in Fig. 9e, when averaged over all latitudes. Fig. 10 shows that the span of the interferometric Doppler residual extends over ±7KHz (the sum of the spans of the tested Dopplers applied to the replica and to the observed Doppler offsets of the received and interfering radio signal components). Fig. 10 also shows that the distribution of the interferometric Doppler residual can be approximated by a triangular function with a reference of ±7KHz. In the following, the span of the interferometric Doppler residual is called D. For the proposed illustrative example, D=+7KHz-(-7KHz)=14KHz.

[0136] The corresponding span is determined by considering only the Doppler effect of the satellite-to-user link (satellite clock stability D osc,sat Receiver oscillator or clock stability D osc,Rx If the additional contribution of the receiver clock stability is taken into account, the Doppler hypothesis f Dopp,test,0 The range or span of the proposed example is ±(3.5KHZ+D osc,Rx ) and, in the absence of further deduction, also takes into account the uniform distribution of the Doppler hypotheses to be tested. By applying the same method, the interferometric Doppler residual Δf Dopp,test,p The distribution of the Doppler residuals is obtained by convolving the distribution of the observed Doppler offsets shown in Fig. 9e with the modified distribution of the Doppler hypothesis to be tested, which applies to the replica and has a larger span. The distribution of the interferometric Doppler residuals is also calculated according to the criterion D = +(7KHz + 2D osc,Rx )-(-7KHz+2D osc,Rx )=14KHz+4D osc,Rx It can be shown that the D can be approximated by a triangular function by considering again [Reference 3]. For a low-grade oscillator with 3 ppm, D osc,Rx= 5.25KHz and therefore the basis of the triangular model approximating the distribution of the interferometric Doppler residual is shown to be equal to D = 35KHz. The previous calculation of the span is applicable to other receiver oscillator grades as well as other advanced platforms of radio navigation systems.

[0137] Introducing the definitions of the auto- and cross-correlation functions, together with the definitions of the Doppler and the corresponding span of a typical MEO satellite-radio receiver link, it is now proposed to calculate the variation of the ACF and CCF values ​​for realistic values ​​of Doppler offset for signal components such as the GPS C / A signal components, as well as to express the variation of the percentiles of the CCF according to the desired or estimated Doppler residual values.

[0138] In each of Fig. 11a, Fig. 11b, Fig. 11c and Fig. 11d, the values ​​of the autocorrelation or cross-correlation function are represented in the upper part and the corresponding distribution is represented in the lower part. In this example, spreading codes belonging to the Gold code family currently implemented for GPS C / A radio signal components have been applied. Each code contains 1023 chips and is modulated with a BPSK(1) pulse waveform. For both ACFs (or CCFs), the code delay between the spreading code replica and the received spreading code applied to the radio signal component of interest (or to the interfering radio signal component) is determined by the code bin width (e.g. T c The ACF covers the entire spreading code length L with a granularity corresponding to Q × (Q-1) / 2). For the ACF calculation, Q = 24 spreading codes, i.e., the Gold codes of the GPS C / A radio signal components, have been considered. For the CCF calculation, Q × (Q-1) / 2 = 276 possible pairs of different GPS C / A Gold codes have been considered. In addition, and for ease of understanding, the normalized correlation function (see equation (8) for the ACF and equation (10) for the CCF) is considered. Finally, the coherent integration time T int (chip rate f cFor the determination of the ACF and CCF values, a sampling frequency of twice the chip rate (U=2) was applied, corresponding to two samples or code hypotheses per chip. Figure 11a shows the results for the desired Doppler residual at 0 Hz (Δf Dopp,test,0 = 0). It can be observed that the maximum ACF peak of 1 is obtained for a delay of 0. For code delays different from 0, the ACF values, also called side-peak ACF, degrade quickly and are less than ∼0.05 (for ACF peak of 1). The lower part of Fig. 11a therefore shows the variation of the ACF for all Q = 24 Gold codes when applying a delay of 0 dB(20 × log 10 (|r a Figure 11b shows the corresponding distribution of ACF values ​​expressed in terms of the ACF-values ​​in units of 10 Hz, expressed as |f |) for the Q-th delay. The ACF peak 1 for 0 delay is recognizable with a single value amounting to 0 dB, while the other side-peak ACFs are reduced to -24 dB, -30 dB or -60 dB, since the side-peak ACF of Gold codes is so-called ternary and two samples per chip are considered to calculate the ACF. Figure 11b shows the CCF values ​​between two different spreading codes, i.e. Gold codes, one for the spreading code replica corresponding to the radio signal component of interest and the other for one of the Q interfering radio signal components, and also for the interference Doppler residual (Δf Dopp,test,p=0 ) for the case where there is no match between the two different spreading codes, so there is no more peak at 1. Nevertheless, the same ternary nature can be recognized in both the upper and lower figures due to the inherent properties of the CCF of the Gold code, observed in the case of 0 Hz interference Doppler residuals. This "ternary CCF" can also be recognized in the lower plot of Fig. 11b, which shows the CCF distribution. Fig. 11c and Fig. 11d show that in the case of ACF, the desired Doppler residual Δf Dopp,test,0 = 3.5KHz, and in the case of CCF, the interference Doppler residual Δf Dopp,test,p= 3.5KHz is applied. It can be observed that in the presence of Doppler residuals, the main ACF peak 1 disappears completely as a result of the phase rotation due to the exponential term in the coherent integration time observed in equation (8). When equation (10) is taken into account, a similar effect applies to the CCF in the presence of non-zero coherent Doppler residuals, with a strong reduction in the CCF value. Moreover, the ACF and CCF values ​​have the same magnitude, which can also be confirmed by the corresponding distributions in the lower parts of both Figures 11c and 11d.

[0139] 12a and 12b show the maximum ACF value expressed in dB when this Doppler varies between -8 KHz and 8 KHz (corresponding to a range of 16 KHz, slightly wider than the span of the desired Doppler residual equal to 14 KHz), as well as the Doppler variation of the 99%, 95% and 67% percentiles of the ACF distribution as a function of the desired Doppler residual. int = 1 ms, and T corresponding to a spreading code period of X = 5 of 1 ms (shown in FIG. 12b). int Two different coherent integration times have been considered, 0 dB = 5 ms. The main ACF peak (0 dB) for 0 Hz Doppler is recognizable in both Fig. 12a and Fig. 12b. Similarly, Fig. 12c and Fig. 12d represent the maximum CCF value expressed in dB when this Doppler varies between -8 KHz and 8 KHz (again corresponding to a (slightly) wider range than the one of the interferometric Doppler residuals equal to 14 KHz), as well as the Doppler variation of the 99%, 95% and 67% percentiles of the CCF distribution as a function of the interferometric Doppler residual. Again, the T (shown in Fig. 12c) int = 1 ms, and T (shown in FIG. 12d) intTwo different coherent integration times have been considered, T = 5 ms. For both ACF and CCF, it is possible to see that the maximum and percentile values ​​are "stable" with respect to Doppler when the coherent integration time is equal to the code period, i.e., 1 ms (Figs. 12a and 12c). Nevertheless, when the coherent integration time is extended to 5 ms, a greater dynamic of the variation of the maximum and percentile values ​​of both ACF and CCF is observed (Figs. 12b and 12d). Finally, T int = 5 ms, a regrowth of the ACF (or CCF) maximum can be observed at multiples of 1 KHz of the desired (or interfering) Doppler residual, which corresponds to the inverse of the spreading code period (1 / 1 ms).

[0140] [Reference 5]: "Impact of Galileo Spreading Code Selection and Data Rate onto Navigation Signal Interference", F. Soualle, ION GPS / GNSS 2003, September 9-12, 2003, Portland, Oregon

[0141] The previous results show that the maximum value of the CCF and the cross-correlation isolation of the CCF depend on the Doppler of the actual received carrier frequency of the interfering signal relative to the carrier frequency being tested for the radio signal component of interest, or in a more general sense, on the offset Δf Dopp,test,p It is therefore proposed here to investigate how the CCF values ​​and percentiles of the corresponding CCF distributions vary over interference Doppler residual spans higher than D when considering different types and especially spreading code lengths L. One proof for higher or extended interference Doppler residual values ​​from the initial span D is the Doppler f Dopp,test,0 is an additional offset frequency applied to the carrier frequency of the interfering signal component relative to the carrier frequency of the signal component of interest corresponding to this offset frequency. This said offset frequency is the basis of the proposed scheme which will be explained in more detail later.

[0142] Figure 13a shows the coherent Doppler residual Δf equal to 0 Hz. Dopp,test,p Figure 12c shows the distribution and cumulative distribution of the CCF calculated for and obtained for a family of Q = 24 Gold codes similar to those applied to the GPS C / A radio signal components. In the case of Figure 12c, the coherent integration time is T with 2 samples per chip. int =1ms (U=2, f s =2×f c ). In Fig. 13a, the maximum CCF, equal to -23.9 dB, as well as the 99%, 95% and 67% percentiles of the CCF distribution are also represented. Fig. 13b shows the interference Doppler residual Δf for the same family of 24 spreading codes. Dopp,test,p The distribution and cumulative distribution of the CCF for = 2500 Hz are presented here. The corresponding maximum values ​​are also presented, together with the 99%, 95% and 67% percentiles of the CCF distribution. Finally, in Fig. 13c, the variation of the corresponding maximum, 99%, 95% and 67% percentiles is provided over the span of the interferometric Doppler residuals extending from 0 Hz to 10 KHz. Since Fig. 12c showed that the behavior of negative Doppler (-8 KHz to 0 KHz) is identical to the positive Doppler (0 KHz to 8 KHz), only the positive part from 0 KHz to 10 KHz is shown in Fig. 13c. It can be observed that the corresponding percentile levels are very stable. Furthermore, for purposes of illustration, the span of the interference Doppler residual has been extended to 10 KHz, thus intentionally exceeding the maximum "natural" Doppler residual span of 7 KHz (±7 KHz when taking into account the negative and positive parts) that would be obtained if the carrier frequencies of the interference signal components and the signal components of interest were identical.

[0143] 14a, 14b, and 14c show the CCF and cumulative CCF distributions (with percentiles) taking into account the interference Doppler residual when the CCF is calculated for a spreading code with a shorter spreading code length L=341 chips (as compared to the spreading code length L=1023 of the GPS C / A radio signal component). c= 1MCps. The span of the interference Doppler residual is set from -10KHz to +10KHz, but for symmetry reasons, 0KHz to 10KHz is shown. Similarly, Figures 15a, 15b and 15c show the CCF (with percentiles) and cumulative CCF distributions taking into account the interference Doppler residual when the CCF is calculated for a spreading code with a shorter spreading code length L = 31 chips. The spreading code is also a Gold code, but with a length of 31 chips (instead of 1023 chips) and f c = 1MCps. The span of the interferometric Doppler residual is set from -40KHz to +40KHz, but for symmetry reasons, only 0KHz to 40KHz is shown. For all figures, the CCF calculation is based on a set of Q = 24 spreading codes, resulting in 276 combinations of spreading code pairs. From Figures 14c and 15c, the 99%, 95% and 66% percentile curves derived from the CCF distributions show that the maximum value K of the interferometric Doppler residual times the reciprocal of the code period. Harm / (L×T c ) for L = 341 chips, this is equivalent to 0 Hz (K Harm =0), 3000Hz (K Harm =1), 6000Hz (K Harm =2), and 9000Hz (K Harm For L=31 chips, this corresponds to 0 Hz (K Harm =0), 33KHz(K Harm =1). These maxima correspond to the "first-order lobes", whose width is twice the reciprocal of the coherent integration time, 2 / T int Between these first-order lobes is the inverse of the coherent integration time, 1 / T int A "secondary lobe" with a width equal to is also visible. For L=341 chips, the CCF percentiles show a strong reduction of 10 dB between the first and second lobes, while for L=31 chips, this reduction reaches an even stronger isolation of approximately 25 dB relative to the maximum value obtained for the main lobe.

[0144] It is pointed out that these observed characteristics of the CCF distribution as a function of the interferometric Doppler residual (spacing between the main lobes, shape and width of the lobes) are closely related to the nature of the spectral representation of the GNSS radio signal components, and in particular to the nature of the Power Spectral Distribution (PSD). The corresponding GNSS PSD characteristics are now presented for different types of radio signal component structures.

[0145] First, Figure 16 provides a spectral representation, or PSD, of a GNSS radio signal component generated by a spreading code with very long or infinite length (L → ∞) modulated with a BPSK(1) waveform. The analytical expression for this unique normalized PSD is Γ c (f)=(1 / f c )×sinc 2 (f / f c ) as shown in [Reference 5]. Since the spreading code is assumed to have infinite length, the resulting PSD is smooth. It is emphasized that even in the case of infinite or very long spreading code lengths, it is still possible to apply data modulation to the signal.

[0146] 17a and 17b provide a spectral representation, or PSD, of a GNSS radio signal component modulated with a BPSK(1) waveform and having a finite spreading code length, shown at two different frequency scales. In this case, the GNSS radio signal component is T c 17b, the PSD is assumed to be a pilot (i.e., dataless) radio signal component, applying a spreading code comprising L=1023 chips with a chip duration of L=(1 / 1.023E6)s. Due to the code periodicity, the PSD, also called spikes in the following embodiments of the invention, is 1 / (L×T c )=f c The corresponding spectral lines, or spike peaks, are still spaced apart by the inverse of the code period, which is equal to Γ / L=1KHz. c (f) follows the envelope shape given by equation (see also [Reference 5]).

[0147] Figures 18a and 18b provide the spectral representation, i.e. PSD, of the GNSS radio signal components modulated with a BPSK(1) waveform, also shown on two different frequency scales. Contrary to Figure 17a, the GNSS signal is modulated with symbols, resulting in radio signal components for data transmission, with a symbol rate f equal to 50 sps. s T c 18b, we still apply a spreading code containing L=1023 chips with a chip duration of Γ=(1 / 1.023E6)s. To obtain the PSD of such a radio signal component for data transmission, each spectral line or spike of the PSD of a corresponding dataless or pilot radio signal component having the same waveform and spreading code is replaced by the equivalent PSD of the symbol modulation, with appropriate scaling to ensure real power, as shown in FIG. 18b. The analytical expression for the PSD of the symbol modulation is now s (f)=(1 / f s )×sinc 2 (f / f s ).

[0148] Figures 19a and 19b show the same carrier frequency f central FIG. 19b (a zoomed-in view of the spectrum) shows the PSD of P GNSS pilot radio signal components with a spreading code length L = 31 chips, transmitted at 1 kHz and received at a ground-based GNSS receiver, at two different scales. Harm 20a and 20b show the variations caused by the Doppler offset of P observed pilot radio signal components whose spikes fall in the frequency interval of ±3.5 KHz for a frequency of 100 Hz (f × 33 KHz). Similarly, Fig. 20a and Fig. 20b show the variations caused by the Doppler offset of P observed pilot radio signal components whose spikes fall in the frequency interval of ±3.5 KHz for a frequency of 100 Hz (f × 33 KHz). central and is received by a ground GNSS receiver, has a spreading code length L=31 chips, and f s Figure 20b (a zoomed-in view of the spectrum) shows the PSD of P GNSS radio signal components for data transmission, modulated here with a symbol rate of Γ = 1 Ksps, at two different scales.s (f)=(1 / f s )×sinc 2 (f / f s ) PSDs, the variation in the center frequency of the individual spikes is caused by the ±3.5 KHz observational Doppler span. Figures 19b and 20b clearly show the spectral overlap at lower scales of the corresponding individual PSDs due to possible observational Doppler collisions, i.e. when two different satellites are observed at the same Doppler. In order to avoid such collisions resulting in MAI, one basis of the present invention is therefore to implement a synthetic or "artificial" frequency offset, different for each radio transmitter, which will allow interleaving the spikes of the spectral representation between the different radio transmitters.

[0149] Having described the main characteristics of the different types of PSD of GNSS radio signal components, it is now proposed to relate these properties to the properties of the CCF depending on the interference Doppler residual, or more generally, any frequency residual between the observed Doppler frequency of the received interference radio signal component and the tested frequency of the replica corresponding to the radio signal component of interest. In [Reference 5], the coherent integration time T int It has been shown that a correspondence exists between the CCFs between the replica and the interference signal evaluated over a period, and the convolution between the spectral representation of the replica and the spectral representation of the interference signal component. The following derivation provides the main steps to develop this relationship. First, let u(t) (or v(t)) represent the analytical expression for the time domain representation of the replica (or interference) signal component. Let v(t) be the time domain representation of the replica (or interference) signal component, and v(t) be the time domain representation of the replica (or interference) signal component. The coherent integration time T int The correlation function between the replica and the interference signal, calculated over τ and as a function of the code delay τ, can therefore be expressed as:

number

[0150] Furthermore, the Fourier transform (FT) for the product of both signal components and as a function of frequency ν can be expressed as a convolution of the individual Fourier transforms.

number

[0151] In these equations, B represents the bandwidth of both sides of the receiver front-end filter. When υ is 0, it is possible to realize the relationship between the time-domain correlations of the individual signal components and their spectral representation (i.e., PSD), which can also be derived from Gabor theory.

number

[0152] If Doppler affects the interfering signal v(t), this Doppler will be perceived as a shift in the spectral representation V(f) (which depends on the code delay τ through (14)). If the spectral representations of both the signal replica (u(t)) and the interfering signal (v(t)) contain spikes spaced apart by the inverse of the spreading code duration (for simplicity, both signals are assumed to be identical), a frequency offset, e.g., caused by Doppler or any other "synthetic" frequency offset, applied to the spectral representation of the interfering signal such that these spikes "fall" between the spikes of the spectral representation of the signal replica, will, if not eradicated, result in a strong reduction of the cross-correlation effect, in other words, a strong improvement of the cross-correlation isolation, based on the previous analytical formula. The previous high-level mathematical derivation describes the principles that will be exploited by the proposed embodiment of the present invention to reduce the MAI between different received signals.

[0153] Figure 21 further illustrates the previous mathematical proof of the observed behavior on the cross-correlation function as a function of frequency offset, such as Doppler frequency, and shown in Figures 14c and 15c. [Reference 5] has proposed the formula Γ, as shown in the middle left part of Figure 21. w (f)=(T int)×sinc 2 (f × T int ) to obtain the spectral representation of the signal replica by replacing each spectral line of the spectral representation of the equivalent pilot radio signal component with a spectral representation corresponding to a time windowing corresponding to the coherent integration time. int = 1ms and Γ w In the case of the proposed example with (f), it shows a main lobe with a width of 2 KHz and a secondary lobe with a width of 1 KHz. Considering the pilot interference radio signal component, its spectral representation consists of spectral lines, as shown in the left part of Fig. 21. In the case of the proposed example with L = 341 chips and BPSK(1) modulation, the distance between the spectral lines is equal to 3 KHz. In the lower diagram (same diagram as Fig. 14c) which shows the Doppler residual dependent percentiles of the CCF distribution, the previous visual element (1 / (L × T c )=3 KHz apart and main lobes having a width of 2 KHz can be discerned, which may explain the aforementioned correspondence to FIG. 14c.

[0154] The following paragraphs describe the adaptation of the previous sentence to the acquisition of a radio signal component that includes a spreading code modulated with symbols, each symbol being of duration T s where the coherent integration time T used to calculate the cross-correlation int is the symbol duration T s In this case, the P v The received radio signal components have a symbol rate f, possibly with different values ​​between different radio signal components. s For the acquisition of wireless signal components for data transmission, the coherent integration time used to calculate the correlation function is the symbol duration T s =1 / f s (T int ≦T s) cannot exceed T, otherwise symbol transitions occurring within the coherent integration time would strongly reduce the integrated energy. The spectral representation of the replica maps each spectral line or spike of the equivalent pilot signal to a int ≦T s The induced PSD for Γ w (f) is obtained by replacing the spectral representation of the interfering radio signal components for data transmission as explained in the description of Figs. 18a and 18b, and each spectral line or spike of the interfering radio signal components is then represented by the formula Γ s (f)=(1 / f s )×sinc 2 (f / f s ) is substituted for the PSD. Reusing the principle of convolution of both spectral representations to evaluate the percentiles of the CCF distribution according to the interferometric Doppler residuals shows that the main lobe of the Doppler-dependent percentiles of the CCF distribution is still 1 / (L×T c ), but now the main lobes are spaced apart by the product Γ w (f)×Γ s (f) is replaced by T int However, the maximum allowable value T int =T s When taking s (f) 2 can be replaced by using

[0155] The following paragraphs deal in particular with the frequency offset (Δf carr,p , 1≦p≦P) with data signal carrier frequency (f car,p =f central +Δf Carr,p ) to obtain a data signal carrier having a center carrier frequency (f central ) with respect to the frequency offset (Δf carr,p , 1≦p≦P) corresponding to a frequency offset of a plurality of signal components, prior to transmission of each wireless signal component.

[0156] The previous result is the interference Doppler residual Δf of the observed Doppler frequency of the interference signal against the Doppler frequency to be tested in the generated local replica corresponding to the radio signal component of interest. Dopp,test,p It has been highlighted that the cross-correlation isolation is extremely sensitive to the Doppler frequency of the interfering signal. This observation is the basis of the proposed embodiment of the invention. More precisely, by applying a method for increasing the difference between the actual observed Doppler frequency of the interfering signal relative to the Doppler frequency tested for the radio signal component of interest, it is possible to improve the corresponding cross-correlation isolation even further. The proposed solution to benefit from this property therefore consists in applying a preferably small (compared to the chip rate) frequency offset Δf 1 at transmission to a negligible or central carrier frequency of the interfering radio signal component, offset depending on the transmission source, i.e. the satellite (and not only this, but also depending on the signal component identifier, if each transmission source transmits several signal components). carr,p The purpose of applying the central carrier f central Frequency offset Δf to carr,p’ The application of the offset carrier frequency f, also called the data signal carrier frequency, carr,p Furthermore, the magnitude of the offset also depends on the desired cross-correlation isolation obtained for a given spreading code family.

[0157] The following paragraphs describe a frequency offset Δf for each of the P signal components and which is less than the chip rate of the spreading code. carr,p The present invention describes a formal methodology used behind the design of a frequency offset signal component (FATCC) for a digital multi-channel signal having a frequency offset of 10 MHz, the frequency offset value being dependent on a specified cross-correlation isolation applied to each of a cross-correlation function value calculated on a local replica of the data signal component including a spreading code corresponding to the signal component of interest and each of further interfering radio signal components of the plurality of received radio signal components, the cross-correlation function value being dependent on a specified cross-correlation isolation applied to each of a local replica of the data signal component including a spreading code corresponding to the signal component of interest and each of further interfering radio signal components of the plurality of received radio signal components, the cross-correlation function value being dependent on a specified cross-correlation isolation applied to the receiver oscillator stability (D oscil.Rx) and the observed Doppler offset (D TRX1 , ..., D TRXP ) is calculated similarly for the interferometric Doppler residual value, which is calculated as the difference between

[0158] The following equation is based on equation (1) and includes the frequency offset Δf in the argument of the complex exponential function: carr,p The transmitted radio signal component u with index p obtained by applying P Provide an expression for (t) resulting in a complex carrier offset.

number

[0159] Previous equation u P In (t), the possibility of also modulating the symbols to obtain radio signal components for data transmission is given by the term d P (t). For the pilot radio signal component, the term d P (t) should not exist.

[0160] The following equation is therefore P v Each of the received signals has a different frequency offset Δf carr,p We provide an expression for the aggregate received signal s(t) when applying the frequency offset Δf applied to the wireless signal component of interest, which, by convention and for ease of understanding only, has index 0. carr,0 is null, and Δf carr,0 It is proposed to yield =0.

number

[0161] To formally prove the expected (normalized) correlation performance achieved with the proposed scheme, the frequency offset Δf applied to the p-th interfering radio signal component iscarr,p is therefore also introduced into the equation for (see (normalized) CCF(10)) which represents the effect of mutual access inference (MAI) on the acquisition detector output for the wireless signal component of interest.

number

[0162] term

number

number

number

[0163] This formula is the coherent Doppler residual Δf (calculated for an unoffset carrier frequency). Dopp,test,p frequency offset Δf carr,p It is emphasized that by adding (Δf Dopp,test,p +Δf carr,p ), we show the maximum as well as percentile behavior of the CCF distribution as it depends on 2 / Tint The width of the main lobe of ("from zero to zero") is 1 / T int It is possible to estimate, together with the second lobe having a width of

[0164] When a frequency offset is applied, the name (extended) coherent Doppler residual is derived from the central (That is, f carr,p , (see also Eq. 13))

number

number

[0165] To derive this formula, the frequency offset Δf to the center carrier frequency carr,p has negligible effect on the observed Doppler offset, f central +Δf carr,p f calculated for Dopp,p But, f central f calculated for Dopp,p is approximated to, and therefore

number

[0166] To make a conclusion, we represent the maximum and percentile values ​​of the CCF distribution (CCF calculated between the local replica and each of the interfering radio signal components), respectively, as well as the (extended) interferometric Doppler residual Δf Dopp,test,p The function according to will have similar properties to the spectral representation of the corresponding interference signal component, and the main lobe distance will be 1 / (L×T c ) and the main lobe width is 2 / T int and for data modulated signals, 2 / T int and 2×f s Finally, the secondary lobe width is equal to the minimum between 1 / T int and for data modulated signals, 1 / T int and f s is equal to the minimum value between

[0167] FIG. 22 shows the (extended) coherent Doppler residual value Δf Dopp,test,p and all Q×(Q−1) / 2 pairs of spreading codes and all possible code delay shifts (τ P -τ test,0 ) Maximum CCF value CCF max (Δf Dopp,test,p The paper describes the structure of the proposed solution, called the Code Doppler Division Multiple Access (CDDMA) scheme, based on the (extended) interference Doppler residual. The aim is to ensure that the CCF taken on the (extended) interference Doppler residual is equal to the CCF req The goal is to offset the carrier frequency of each of the P GNSS signals in such a way that it is not higher than the minimum and required CCF isolation, called the minimum required CCF isolation, called the minimum required CCF isolation, called the minimum required CCF isolation. This is done by applying different offsets to the carrier frequency of each of the P GNSS signals, over the different combinations of Q × (Q-1) / 2 pairs of spreading codes, all possible code delay shifts (τ P -τ test,0 ) and over the entire span of the (extended) coherent Doppler residual, the CCF is reqThis very strict definition will ensure the best possible isolation, but it will have some negative impacts on the multiple access capacity (i.e. the number of signals P that can be received simultaneously with such required CCF isolation). v , and thus limiting the total number of P of radio signal components). Dopp,test,p It does not consider the maximum CCF over the 99% (Δf Dopp,test,p )), 95% (CCF 95% (Δf Dopp,test,p )), or 66% (CCF 66% (Δf Dopp,test,p It is possible to relax the corresponding condition on CCF isolation by considering percentiles of the CCF distribution, such as the 95% percentile (in the proposed example, the CCF 99% (Δf Dopp,test,p ) CCF req The proportion of CCF values ​​that exceed the requirement is determined by the number of different combinations of Q × (Q-1) / 2 pairs of spreading codes for all possible code delay shifts τ p -τ test,0 It is possible to guarantee that the CCF never exceeds 5% of the total CCF values ​​over the entire span of the (extended) interference Doppler residual. The same principle applies for 99% or 66% or any other percentile XX% of the CCF distribution per Doppler. This relaxation therefore makes it possible to increase the number P of radio signal components. In the following, the method used to determine the frequency offset to be applied to the transmit carrier frequency to each of the P radio signal components is called the (extended) interference Doppler residual CCF max (Δf Dopp,test,p ) span of the target or specified CCF insulation CCF reqHowever, as a target or designated CCF isolation, the (extended) coherent Doppler residual CCF xx% (Δf Dopp,test,p The same method can be applied when considering any other percentile of the CCF distribution over 100 MHz. Moreover, for the maximum value of the CCF distribution and for the radio signal components transmitted by GNSS, when this cross-correlation isolation is defined, it is usual to specify cross-correlation isolation values ​​between -30 dB and -15 dB. Nevertheless, alternatives (more relaxed or stricter values) of the specified cross-correlation isolation can be proposed for other types of radio navigation systems and supported services, such as terrestrial radio navigation systems, having other magnitudes of the number Q of radio transmitters, as well as other specified levels of ranging and positioning performance.

[0168] As an example, the spreading code length L = 31 chips and the chip rate f c = 1MCps, and coherent integration time T int It is proposed to take into account the following design parameters, with Δf = 1 ms: In Figure 22, the transmission carrier frequency of the radio signal component of interest, when downconverted to baseband, is equal to 0 Hz (i.e., the carrier offset Δf carr,0 =0) is considered. The radio signal components of interest are P v As a fraction of the visible received signal, W v 1 received signal can be selected. By way of example, the signal component of interest is selected as the first CDDMA carrier frequency, the exact definition of which is given later. Nevertheless, the same principles and explanations should still be valid for the selection of another CDDMA carrier. At baseband, (P v - The received carrier frequency of 1) other interfering signals is referenced to 0 Hz. Minimum CCF Isolation CCF req The minimum (extended) coherent Doppler residual that guarantees opp,isol As a proposed example, CCF req = -25dB, the maximum CCF value is CCFreq The minimum (extended) coherent Doppler residual, ΔDopp isol = 1.5KHz (corresponding to the peak of the first secondary lobe, see the structure in FIG. 22). isol is also called the frequency isolation margin in Figure 22. isol It is outlined that depends on the shape of the main and secondary lobes of the function showing the maximum of the CCF distribution as a function of the (extended) coherent Doppler residual. As explained before, the width of these main (or secondary) lobes is 2 / T int (or 1 / T int 1), and therefore ΔDopp isol is determined analytically or by applying numerical simulations, and is expressed as 1 / T int It can be further expressed as proportional to T int For =1ms, the numerical simulation shows that CCF req If a =-25 dB isolation needs to be met, then the proportionality factor is α=1.5 (ΔDopp corresponding to the first secondary lobe peak) isol =1.5 / T int ) where α is the higher CCF req Increased insulation is required, CCF req Dependence α with CCF req ) Therefore, ΔDopp ,isol The preliminary setting for ΔDopp isol =α(CCF req )×(1 / T int ) is given by the symbol rate f s For a data modulated CDDMA signal with s Since it cannot exceed ΔDopp ,isol = α × f s It is.

[0169] CCF[dB]≦CCF req To ensure that p,isoi It is therefore necessary that θ must not be smaller than θ. To meet this condition, Δf carr,1 = ΔDopp isol +D / 2 (Equation 22) The value Δf given by carr,1 It is then necessary to offset the carrier frequency of the interfering radio signal component for the second CDDMA channel having index p=1, here with index p=1.

[0170] As a proposed example, for L=31 chips, Δf carr,1 = 1.5KHz + 7KHz = 8.5KHz (not considering the contribution of the receiver oscillator stability).

[0171] Therefore, Δf carr,1 By providing a carrier frequency of the first CDDMA channel corresponding to the first interfering radio signal component having test,0 Over the entire range of (extended) coherent Doppler residuals, the cross-correlation is req To simplify the structure of the CDDMA scheme, FIG. 22 shows a first CDDMA channel for the first interfering wireless signal component, and carr,1 We also provide some graphical elements, in particular the distribution, which has a triangular shape, of the (extended) coherent Doppler residual shifted by .

[0172] For the first CDDMA channel corresponding to the first interfering wireless signal component, a frequency offset Δf carr,1 Having presented a method for setting the frequency offset, it is proposed to generalize the frequency offset for other CDDMA channels corresponding to other interfering wireless signal components. Figure 23 shows the required cross-correlation isolation CCF with the first p=0 CDDMA channels corresponding to the signal component of interest. req To ensure that the (P-1) carrier offsets Δf carr,pThe present invention describes a method for determining the carrier offsets for all possible user locations and therefore P of the P wireless signal components. v It is pointed out that since the allocation method is guaranteed for all possible combinations of p visible radio signal components, the allocation method therefore applies to the full capacity of P radio signal components. By repeating the same procedure, the corresponding offset applied to the carrier frequency of each CDDMA channel of index p is given by: Δf carr,p = p × (ΔDopp isol +D / 2)=p×8.5KHz (Equation 23)

[0173] For the fourth radio signal component (p=3, corresponding to the last P=4 CDDMA channels), the previous equation reduces to Δf carr,p =(P-1) × (ΔDopp isol In FIG. 23, for the hypothetical fifth (p=4) interfering radio signal component, the carrier frequency is 1 / (L×T c ) at 33KHz. The CCF of this second main lobe can reach a maximum value of -8dB, so a CCF of -25dB req The condition for isolation is no longer met. This proves why for the proposed configuration applying different spreading codes with length L=31 chips for the different radio signal components, only four frequency carriers (p=0, 1, 2, and 3) can be allocated to a maximum of four (P=4) CCDMA channels in order to satisfy the requirement of correlation isolation.

[0174] From equation (Eq. 23), the P CDDMA carriers in the P CDDMA channels are CCDMA =(Δ Dopp,isol +D / 2), which is also depicted in FIG.

[0175] In the same Figure 23, it is also shown that each of the four CDDMA channels uses a different code, or PRN (PRN#1 for CCDMA channel #1, PRN#2 for CCDMA channel #2, ...), and all PRNs have the same length L (equal to 31 chips in this illustrative example). This validates the application of four different PRNs, since CCDMA allocation is based on the behavior of the cross-correlation function with the (extended) interference Doppler residual. It will be shown later that a variant of CDDMA proposes to use the same PRN for all CDDMA channels, in which case the dependence of the autocorrelation function with Doppler is applied for allocation.

[0176] Frequency offset Δf carr,pIt is important to emphasize that depends on the maximum span or excursion of the interference residual Doppler D (here there is no expansion with carrier offset as in (equation 23)). When the Doppler contribution is taken into account, D depends on the deduction on the Doppler uncertainty. D is maximum in cold acquisition, but can be strongly reduced by deduction to hundreds of Hz (instead of a few KHz). Such a deduction on the Doppler of the interference radio signal component can be provided, for example, by a local base station (LBS) to the connected user device. In case of such an assisted scenario, D becomes small and it is possible to reduce the Doppler offset between the CDDMA carriers and thus the distance between the CDDMA channels. As mentioned before, the maximum excursion of the interference residual Doppler will depend on the contribution of the radio transmitter-radio receiver link dynamics (different for GEO, MEO, LEO or terrestrial radio emitters) on the one hand, and on the other hand, on the stability of the radio receiver oscillators or clocks. The CDDMA allocation shown in FIG. 22 was proposed for an excursion or span of the interference Doppler residual D equal to 7 KHz, since it does not take into account the receiver oscillator stability. If the receiver oscillator stability is taken into account, D increases to 35 KHz for a low grade oscillator with 3 ppm, as shown in the background art section. For higher grade receiver oscillators, D should be between 7 KHz and 35 KHz. The same method can therefore be applied for this updated value of D, taking into account the receiver oscillator stability as well.

[0177] As mentioned before, the frequency offset is shaped according to the span of the interference Doppler residual D, which justifies the term Doppler in the proposed nomenclature of the solution, “Code Doppler Division Multiple Access”. Here, and since the corresponding signal components are spread with a pseudorandom code, the name of the proposed solution also includes the term Code. Besides the nomenclature proof, the design constraints on both Doppler span and spreading code length L (intrinsic to the CDMA scheme) used to find the optimal combination for the CDDMA signal design also justify the inclusion of both Doppler and Code terms in the CDDMA nomenclature.

[0178] In the following, each of the P carriers will be called a "CDDMA channel". The CDDMA scheme shares some similarities with the Frequency Division Multiple Access (FDMA) scheme in the sense that a frequency offset is applied to ensure spectral isolation to reduce the MAI. Here, and contrary to this traditional FDMA scheme, it is possible to significantly reduce the spectral occupancy for P transmitted signals modulated with BPSK(N) or BOC(M,N) pulse waveforms. Figure 24a shows, for example, that an FDMA scheme with P=4 FDMA channels of 4 MHz each (assuming BPSK(1)) would request an overall spectral occupancy of 16 MHz. In contrast, Figure 24b shows that the CDDMA scheme would fully request a bandwidth slightly higher than 4 MHz. This overall bandwidth occupancy is in fact due to the fact that the increase in bandwidth is ΔDopp isol and for typical values ​​of D, this represents no more than 5 to 10% of the initial bandwidth of 4 MHz, 4 MHz + (P × (ΔDopp isol+D / 2). Figure 24c shows a zoomed-in view of four dataless CDDMA channels. When compared to the PSDs of radio signal components obtained without the application of frequency offsets (see zooms on Figure 19b (pilot radio signal components) and Figure 20b (radio signal components for data transmission)), it can be observed in Figure 24c that the application of frequency offsets ensures interleaving between spikes of different FDMA PSDs, which strongly reduces the spectral overlap and therefore improves the cross-correlation isolation. Thus, the spectral isolation of the CDDMA scheme is guaranteed on a small micro (frequency) scale, while in the case of the FDMA scheme it is guaranteed on a macro (frequency) scale. Furthermore, in the case of the FDMA scheme, in order to limit the overall spectrum occupancy, only the main lobe and possibly the first side lobe (for example as shown in Figure 24a) of the PSD of each radio signal component applying the FDMA scheme is included per FDMA channel. As another example, only the main lobe of the GLONASS L1COd signal is transmitted per FDMA channel (1 MHz wide). A direct consequence of the limited transmission signal bandwidth is a degradation of the ranging performance in terms of noise and multipath robustness. Such an effect is not present in the proposed CDDMA scheme (but also in CDMA schemes), which allow transmission over a wider bandwidth, since they are not subject to strict bandwidth limitations. The strongly reduced spectrum occupancy also represents an important advantage for two main reasons when compared to the one of the FDMA schemes. Firstly, it can facilitate the International Telecommunication Union (ITU) allocations for new GNSS signals especially when considering the constantly increasing demand for spectrum resources in the frequency bands allocated to Radio Navigation Satellite System (RNSS) services. Thus, a reduced spectrum overlap should be achieved with coexisting GNSS signals sharing the same RNSS bands. Another very important benefit of the reduced spectrum occupancy is the smaller receiver front-end (~4 MHz for CDDMA instead of ~12 to 16 MHz for traditional FDMA), which definitely reduces the receiver complexity and cost.

[0179] In the proposed example, and for the configuration depicted in Figure 23, it has been shown that "only" four CDDMA channels can be created since the fifth channel "falls" in the 33 KHz region which also contains the main lobe and therefore would not meet CCF isolation. Different options are now proposed which are also exploited by embodiments of the present invention to enhance the overall capacity for the number of channels, and therefore the number of P = F x Q radio signal components, that can be transmitted by a satellite radio navigation system.

[0180] A first option, or an alternative embodiment of the invention, consists in allocating different spreading code lengths to different spreading codes among different GNSS signals. Such spreading code lengths will exhibit inherent algebraic properties, such as being mutually prime, or will exhibit the largest Least Common Multiple (LCM) value. For illustration, two different code lengths L1 and L2 can be considered. It is proposed to modulate half of the P radio signal components with a spreading code having a first spreading code length L1 (equal to 31 chips in the proposed example), while the other half of the P radio signal components are modulated with a spreading code having a second spreading code length L2 (equal to 33 chips in the proposed example). As 31 and 33 are mutually prime, the LCM is equal to 1023 chips, or 1 ms, the coherent integration time of BPSK(1). In this first option, it is outlined that CDDMA signal components with the same spreading code length L1 (or L2) still apply different spreading codes (PRN#1 for CCDMA channel #1, PRN#2 for CCDMA channel #2, ... for spreading code length L1, and PRN#5 for CCDMA channel #5, PRN#6 for CCDMA channel #6, ... for spreading code length L2) to exploit the nature of CCF as in Fig. 23. Fig. 25 represents the distribution of Doppler-dependent percentiles of the obtained CCF with L1=31 chips (top of Fig. 25) or L2=33 chips (bottom of Fig. 25). The second main lobe is 1 / (L1×T c 1 / (L2×T)=33 KHz for the second category of signal components that are far from the first one but apply a PRN with a spreading code length of L2=33 chips. cIt can be observed that the frequency offsets are far from the 1023-31 (or 31) KHz. In the lower part of Fig. 25, four CDDMA channels for radio signal components applying spreading code length L1 = 31 chips and four other CDDMA channels for radio signal components applying spreading code length L2 = 33 chips are also represented. It is pointed out that the small difference of 33-31 = 2 KHz allows applying the same frequency offsets for channel #1, channel #2, channel #3 and channel #4. These common offsets will therefore be calculated based on the largest (Doppler) frequency value of 33 KHz (to further ensure cross-correlation isolation). Fig. 26a provides a time representation of both spreading code streams over 1 ms as a concatenation of 33 (or 31) PRN codes of length 31 chips (or 33 chips). These longer spreading codes, having a length of 1023 chips and based on a basic spreading code of 31 or 33 chips, will be referred to as concatenated spreading codes in the following. Fig. 26b presents different Doppler-dependent percentiles of the CCF distribution when the CCF is calculated between concatenated spreading codes belonging to both families of spreading codes, i.e. 31 chips and 33 chips, and over a range of (extended) interference residual Dopplers up to 35 KHz (while Fig. 25 presented the Doppler-dependent percentiles of the CCF distribution when the CCF is calculated for concatenated spreading codes belonging to the same family, i.e. 31 chips or 33 chips). For this calculation, 12 basic spreading codes with a length of 31 chips and 12 basic spreading codes with a length of 33 chips were generated randomly (i.e. not based on constructed spreading codes such as Gold codes). The coherent integration time is still T, as in Fig. 25. int= 1 ms. As shown in Fig. 26c, if the maximum CCF value (~-15 dB) over the (extended) interferometric Doppler residual is larger than the value obtained with a longer sequence of 1023 chips (such as the Gold code showing -24 dB), it can be observed that the corresponding 99% and 95% percentile values ​​have the same magnitude as the values ​​corresponding to a longer spreading code with L = 1023 chips. Fig. 26c represents the percentiles of the distribution of CCFs calculated for 276 pairs of spreading codes with randomly selected chip values ​​(no Gold codes as in Fig. 13). Since the probability of reaching the maximum CCF value is relatively small, considering the largest percentiles (e.g. 95%, 99%) of the CCF distribution of acquisition performance is sufficient for most of the GNSS applications, avoiding a very conservative and pessimistic design approach. Therefore, by selecting periodic spreading codes with spreading code lengths close to each other (31 vs. 33 chips) and having the largest LCM to nearly reach the overlap of the main lobes of the CCF distributions, as shown in Figure 25, it is possible to double the number of CDDMA channels (8 channels in the proposed example instead of 4 channels with a single spreading code length). Even if the corresponding spreading codes have a short length, sufficient cross-correlation isolation corresponding to the LCM of the spreading codes is still obtained when both spreading code lengths are relatively prime. The proposed example can be extended to spreading code lengths of 3 or more to further extend the number of CDDMA channels.

[0181] A second option, or alternative embodiment of the invention, consists in allocating the same spreading code to two diametrically opposed satellites. This solution thus allows doubling the number of CDDMA channels. This principle can be extended to allocate the same spreading code to satellites that cannot be simultaneously seen everywhere in the coverage area of ​​the radio navigation service. In fact, in the case of Space Based Augmentation Systems (SBAS) operating Geo-synchronous satellites (GEO), the coverage area is reduced to the footprint of the satellite shaped by the directivity of the transmitting antennas on board the GEO satellite. Thus, in the case of such SBAS or regional navigation systems, it is possible to allocate the same spreading code to different GEO satellites, as long as the footprints of the satellites do not overlap. This second option is also applicable for terrestrial systems with coverage areas limited to geographical zones.

[0182] A third option, or alternative embodiment of the invention, consists in using the same spreading code for all P=F×Q radio signal components occupying P CDDMA channels. As a proposed alternative embodiment, P=Q (F=1) CDDMA channels are occupied, each corresponding to one radio signal component obtained with the same spreading code as the other CDDMA channel. Since the interfering radio signal components are modulated with the same spreading code as that applied to the replica signal corresponding to the radio signal component of interest, an autocorrelation function (ACF) must be applied. In this case, equation (8), showing the dependence of the ACF on the desired Doppler residual, would be used instead of equation (10), showing the dependence of the CCF on the (extended) interfering Doppler residual, but now using the interfering Doppler residual instead of the desired Doppler residual. Introducing a frequency offset for each of the radio signal components modulated with the same spreading code should therefore result in the following equation, which is very similar to equation (19), proving that it is possible to follow a method very similar to that based on CCF.

number

[0183] The ACF equation (24) is 0,● ) (i.e., from another CDDMA channel) corresponds to the spreading code of a replica of the “desired” CDDMA channel (c 0,● ), is identical to the one of the cross-correlation functions (CCFs) (see (19)). Figure 27 presents the variation of percentiles of ACFs as a function of (extended) interference residual Doppler. As an illustrative example, it can be observed that for L = 31, the maximum CCF value does not exceed -15 dB, even for main lobes at ±66 KHz and ±33 KHz from the central main lobe. Moreover, when considering the Doppler-dependent distribution of ACFs, the frequency isolation margin ΔDopp isol can be reduced down to 1.25KHz to ensure that a particular CDDMA channel fits exactly to the inverse of the 33KHz spreading code duration, resulting in CDDMA channel frequencies of ±8.25KHz, ±16.5KHz, ±24.75KHz, ±33KHz, etc. As a comparison, it is recalled that the main lobe of the CCF between two different spreading codes of length L=31 chips only reaches -8dB, thus providing 7dB less isolation, as can be seen in Figure 22 for the same main lobe. The main result is that it is possible to allocate "as many CDDMA channels as needed" when considering this single spreading code option. As a proposed example, up to 17 CDDMA channels spreading over ~140.25KHz can thus be exploited.

[0184] The advantage of applying the same spreading code for all radio signal components is that it offers the possibility to apply a completely parallel search, per code hypothesis, over P different CDDMA channels for the Doppler offsets of the different satellites in view and having the same PRN sequence. Thus, contrary to conventional FFT-based acquisition, which is used to simultaneously test all code delay hypotheses per Doppler offset hypothesis (see later in this description of the embodiment for further details on FFT-based acquisition), applying the same spreading code makes it possible to test all Doppler offset hypotheses per code delay hypothesis using FFT. Thanks to the strong reduction of the spreading code length (31 chips in the proposed example compared to 1023 chips for GPS C / A signals), it is possible to reorder the code delays with respect to the Doppler offsets, since the number of code delay hypotheses is small. Since all satellites apply the same spreading code, the corresponding Doppler search also makes it possible to parallelize the satellite (PRN) acquisition, whereas in the case of conventional CDMA schemes the acquisition search needs to be repeated many times per satellite. This simultaneous acquisition search of all visible satellites in a "single shot" also represents an additional reduction factor for the receiver when applying the same spreading code. Thus, the reduction in the code search complexity per satellite can be reallocated and used in parallel for the acquisition of several satellites. For example, consider a scenario with 17 CDDMA channels, each spaced 8.25KHz apart. By testing ~300 Doppler hypotheses with a 500Hz bin width (8.25KHz x 17 / 0.5KHz = 280.5 ≤ 300), W satellites that will be acquired from P visible lines of sight are acq Corresponding to W acq Here, it is possible to detect the output of 9 detectors. 9=512>300) should therefore be sufficient for the concrete implementation. Furthermore, since only one spreading sequence is implemented, it is possible to choose an m-sequence (see [Reference 1]) for which only a single Linear Feedback Shift Register (LFSR) needs to be generated, compared to two LFSRs for the Gold code generation. This therefore represents another advantage when using a single spreading code for all radio transmitters, since it represents a reduction in hardware complexity when compared to Gold codes, which require two LFSRs.

[0185] It is noted that the three previous options can be combined to further increase the number of available CDDMA channels and reach the target capacity of the Q radio signal components. For example, it is possible to apply the same PRN sequence to the opposite satellites. In the proposed example, this doubles the number of CDDMA channels, resulting in 34 CDDMA channels. The lower part of Figure 27 represents 17 additional CCDMA channels for these opposite satellites.

[0186] Another possible option of the combination of the first and third options is represented in Figure 28. The first two spreading codes PRN#1 and PRN#2 are used in a similar manner to the first option, both having the same spreading code length L. Then, the first spreading code PRN#1 is used for half of the P CDDMA channels, and the second spreading code PRN#2 is used for the other half. To avoid unacceptable cross-correlation isolation, the first sequence PRN#1 is allocated to the odd CDDMA channels (#1, #3, #5, ..), while the second sequence PRN#2 is allocated to the even CDDMA channels (#2, #4, #6, ..). By following this allocation strategy by interleaving the CDDMA channels, it can be verified that none of the CDDMA channels corresponding to PRN#2 falls in the main lobe of the CCF between PRN#1 and PRN#2. Vice versa, it can be verified that none of the CDDMA channels corresponding to PRN#1 fall in the main lobe of the CCF between PRN#2 and PRN#1. This is highlighted, for example, by the black bold frame of CDDMA channel #5. This proposed CDDMA scheme, which aims to ensure that the CDDMA channels of PRN#1 (or PRN#2) do not correspond to the main lobe of the CCF between PRN#1 and PRN#2 (or PRN#2 and PRN#1), can be extended to more sequences PRN#1, PRN#2, ..., PRN#I with the same length L, as long as the conditions on "main lobe avoidance" are met. It is outlined that the proposed solution may also be extended by interleaving PRN sequences of different lengths L, as long as the "main lobe avoidance" constraint is still observed. A wider maximum span D, which also takes into account the clock stability, can be achieved by increasing the distance between the channels using the same PRN (e.g., channels #1, #3, #5 in the case of PRN#1) as a result of CDDMA channel interleaving. Dopp+Clock This is also illustrated in FIG. 28, which shows the basis D for the first (considering only Doppler) Dopp , as well as a wider D for the second (taking into account both Doppler and user clock stability)Dopp+clock Here, the CCF isolation between PRN2 of CDDMA channel #2 and PRN#1 of CDDMA channel #3 is wider than the CCF isolation between PRN2 of CDDMA channel #2 and PRN#1 of CDDMA channel #3, and the CCF isolation between PRN2 of CDDMA channel #3 and PRN#1 of CDDMA channel #4 is wider than the CCF isolation between PRN2 of CDDMA channel #4 and PRN#1 of CDDMA channel #5. Dopp+clock Nevertheless, it can still be verified to be satisfied.

[0187] Further embodiments consist in proposing alternative implementations for the application of a frequency offset at transmission for signal generation or at reception for signal downconversion to baseband, or for replica generation, if the replica includes a frequency offset. It is proposed to concentrate first on those implementations applied at the transmission side, in order to support later a similar demonstration applied at the reception side, with the necessary adaptations. Thus, the following describes the implementations for the central carrier frequency (f central ) with respect to the frequency offset (Δf carr,p , 1≦p≦P) to the data signal carriers of the radio signal components.

[0188] In effect, equation (17) expresses how the application of the frequency offset is achieved by multiplying a stream of spreading codes, possibly modulated with symbols. The following equation is a further form of equation (17) corresponding to a complex carrier offset, which, after exponential expansion, is:

number

[0189] Based on this equation, it is possible to present alternative implementations that fall into two main categories. The first category supports the transmission of complex signals, as presented in equation (25), while the second supports the transmission of real signals. In this latter case, the implementation of the carrier offset consists in replacing the multiplication with the exponential function by a multiplication with a cosine or sine function, generated with the same frequency offset, resulting in a real carrier offset. In the following, and for illustration, a multiplication with a cosine function is proposed as shown in the following equation:

number

[0190] The following paragraphs describe how to obtain the center carrier frequency (f central ) with respect to the frequency offset (Δf carr,p An alternative embodiment for the frequency offset application means (11) configured to apply a frequency offset of p (1≦p≦P) to the data signal carriers of the radio signal components is now described. As an alternative implementation for the transmission of complex signals, it is therefore proposed to replace the cosine and sine functions in (25) by quantized versions of these functions. In the simplest case of this alternative implementation, both the cosine and sine functions are quantized with two levels, which is equivalent to applying the sign(●) operator to each of them, as shown in the equation below, resulting in a 1-bit quantized complex carrier offset.

number

[0191]

number

number

[0192] [Reference 6]: "Quantization", Allen Gersho, IEEE Communications Society Magazine, pp. 16-28, September 1977, doi:10.1109 / MCOM.1977.1089500

[0193] [Reference 7]: "Section 1.2.3: Quantization, in Chapter 1: JPEG 2000 Core Coding System (Part 1)", In Schelkens, Peter; Skodras, Athanassios; Ebrahimi, Touradj (eds.), The JPEG 2000 Suite, John Wiley&Sons, pp. 22-24, ISBN 978-0-470-72147-6

[0194] [Reference 8]: "Efficient Scalar Quantization of Exponential and Laplacian Random Variables", Gary J. Sullivan, IEEE Transactions on Information Theory, Vol. IT-42, No. 5, pp. 1365-1374, September 1996, doi:10.1109 / 18.532878

[0195] A more general formula is to use Quant(●) instead of the sign(●) function to quantize cosine and sine functions. Nq (●) It is outlined that it consists of applying a function. Quant NqTwo variants of the (●) function can therefore be distinguished, mid-tread and mid-riser, reusing a specific nomenclature as in [Reference 6]. Their main distinction is in the provision of the zero value as a possible output of the quantized (cosine or sine) function. The mid-tread quantizer provides a zero-valued quantization level, whereas the mid-riser quantizer does not provide a zero-valued quantization level.

number

[0196] In the previous expression, the operator

number

[0197] Mid-tread and mid-riser are the most encountered forms of quantizers implemented with uniform quantization laws, nevertheless other more sophisticated quantizers using alternative laws can also be proposed, such as dead-zone quantizers or possibly quantizers with non-uniform laws, see [Reference 7] and [Reference 8].

[0198] In the following, the expansion for the alternative implementation is q Consider the application of a sign(●) function corresponding to a mid-riser quantizer using ∑ n = 1). Nevertheless, the proposed solution is not limited to the application of a 1-bit quantizer, resulting in a "two-level" quantized sine wave, but can also be extended for multi-level quantized sine waves.

[0199] The 1-bit quantized cosine and sine functions of the 1-bit complex quantized offset carrier shown in equation (27) each have a periodicity P equal to the inverse of the frequency offset. carr,p =1 / Δf carr,pIt can be observed that each of the NRZ waveforms can be considered as a binary NRZ waveform with a period [+1, -1]. The NRZ waveform modulated onto the spreading code can therefore be identified as a secondary code with two symbols [+1, -1], repeating indefinitely. The spreading code therefore takes on the role of the primary code. The duration of the secondary code symbol is a function of the frequency offset, in particular the parameter σ which represents the ratio between half the equivalent period of the (NRZ) sinusoidal waveform obtained for a frequency offset (which is also equal to the secondary code symbol duration) and the spreading code period.

number

[0200] Different implementation strategies can be proposed, whether η is an integer or not. To present these strategies, let us assume that the spreading code length is 31 chips and the chip rate is f c As an illustrative case where is 1MCps, different carrier offsets Δf corresponding to different CDDMA channels are carr,p Furthermore, to explain and illustrate these strategies, it is proposed to consider the option of transmitting a real signal with a frequency offset as shown in (26), with 1-bit quantization (N q = 1) is applied to the cosine function, resulting in a real 1-bit quantized carrier offset shown in equation (29) below. Nevertheless, the following statement also applies when considering the option of modulating the spreading code by a secondary code constructed from a non-return to zero (NRZ) code with real symbols.

number

[0201] FIG. 29 shows the Δf carr,p The situation where Δf is set carr,p =f c / L=33KHz. In this case, η=0.5. The top part of Fig. 29 shows the chip stream corresponding to the concatenation of spreading codes with L=31 chips. The middle part of Fig. 29 shows the frequency offset Δf carr,p 29. The dashed line represents an unquantized sinusoidal waveform (whose value is zero for the first chip and increases forwards) that corresponds to the offset frequency waveform and is aligned with the spreading code without detuning. The NRZ waveform derived from the 1-bit quantization applied to the previous sinusoidal waveform is also recognizable in the same middle part of FIG. 29. It can be verified that the sinusoidal waveform is substantially imprinted within the NRZ waveform. Finally, the lower part of FIG. 29 represents the product of the chip stream and the product of the NRZ waveform. FIG. 29 substantially confirms that the half period of the offset frequency waveform contains 15.5 chips.

[0202] The first implementation therefore consists in dividing the chip at the central transition of the NRZ waveform into two pieces: a first piece, the polarity of which is not reversed after multiplication with the positive part of the NRZ waveform (corresponding to the first "+1" binary symbol of the secondary code), and a second piece, the polarity of which is reversed after multiplication with the negative part of the NRZ waveform (corresponding to the first "-1" binary symbol of the secondary code). The principle is illustrated in the lower left part of Fig. 29. Moreover, the first part of the spreading code period not placed at a transition will not be reversed after multiplication with the positive part of the NRZ waveform, while the last part of the spreading code period not placed at a transition will be reversed after multiplication with the negative part of the NRZ waveform. This structure is repeated for the multiplication of all spreading code periods with a binary NRZ waveform.

[0203] The second implementation option shown in FIG. 30 consists of applying a modified NRZ waveform made from two NRZ sub-waveforms. For the first NRZ sub-waveform, the plateau with positive polarity is one chip shorter than the plateau with negative polarity. For the second NRZ sub-waveform, the plateau with positive polarity is one chip larger than the plateau with negative polarity. Thus, both NRZ sub-waveforms are concatenated to form a modified NRZ waveform that is repeated throughout the entire radio signal component generation. This structure results in an NRZ waveform with non-uniform or asymmetric plateaus. The advantage of delaying (or delaying) the falling edge of the transition of the first (or second) NRZ sub-waveform is that it fits exactly to the falling (or rising) edge of the transition of the chip divided into two pieces, as shown in FIG. 30. By alternating the two NRZ sub-waveforms, a precise frequency offset will generally be achieved based on this modified binary offset carrier scheme. It is pointed out that the non-uniform NRZ plateau may nevertheless contain some additional harmonics or spurs at the output of the (digital) generation, but these harmonics are expected to be removed in the general block "amplification and filtering" shown in FIG. 2.

[0204] FIG. 31 shows the Δf carr,p The situation where Δf is set carr,p =f c / (2L)=16.5KHz. In this case η=1. It can be observed here that one (quantized) sinusoidal waveform fits exactly to one spreading code period, according to the same formula as in Figure 31. In this case a simple multiplication is sufficient, and the transitions of the NRZ waveform imprinted on the sinusoidal waveform correspond exactly to the chip edges of the spreading code.

[0205] Figure 32 shows the Δf carr,p =(3f c29 shows a further example when η = 2 / 3, where η = 2 / 3. In this case, it is shown that the minimum number of complete (quantized) sinusoidal waveforms that contain an integer number of spreading code periods and guarantee a match of the (quantized) sinusoidal waveform transitions with the chip edges is equal to 3, corresponding to 8 full spreading code periods. A structure similar to that of FIG. 29 is proposed, but here the chips falling into the (quantized) sinusoidal / NRZ waveform transition are cut by 1 / 3 or 2 / 3 of their duration depending on the spreading code period to which they belong, over 8 full spreading code periods. Then, the frequency offset Δf carr,p The multiplication of the initial spreading code stream, which consists of 8 full spreading code periods, with a corresponding (quantized) sine wave / NRZ waveform is obtained by combining the corresponding spreading code with "incomplete chips" on its edges and with a change of polarity depending on the NRZ symbol polarity.

[0206] moreover,

number

[0207] The previous figure and the accompanying discussion can be used to extend the concept to other values ​​of the ratio η.

[0208] 29, 30, 31 and 32 also show that the proposed CCDMA schemes with frequency offsets smaller or much smaller than the chip rate (a few KHz in the proposed example for the present embodiment, compared to the typical chip rate expressed in MCps) allow to include several chips of the spreading code per plateau of the secondary NRZ code symbol, and as a result, allow to implement the frequency offset in a simple manner directly in the digital domain, for example, by using a secondary NRZ code or a longer spreading code. This represents one distinguishing enabler for FDMA schemes, where the carrier offset between FDMA channels is typically larger than the chip rate, thus mandating that per spreading code chip there are several periods of a corresponding sine or cosine wave oscillating at the FDMA carrier offset. This as a result does not allow to apply the proposed offset implementations through NRZ secondary codes or longer spreading sequences.

[0209] Figures 29, 30, 31 and 32 and the associated description show the implementation of a real 1-bit quantized carrier offset (1-bit quantized sine and / or cosine), which can also be interpreted as a modulation of the spreading code by a secondary code, composed of a non-return to zero (NRZ) code of (real) symbols, and in particular when η is an integer. The following paragraphs will discuss the modulation of the central carrier frequency (f central ) with respect to the frequency offset (Δ fcarr,p An alternative frequency offset applying means (11) is now described, further configured to apply a frequency offset of 1≦p≦P to said data signal carriers of said radio signal. Including equation (2) in equation (27) yields the following equation:

number

[0210] The previous equations therefore represent an alternative interpretation of the multiplication with 1-bit quantized sine and cosine, where the generated signal can be obtained by applying a complex spreading code, the real and imaginary components of which are a real spreading code obtained by modulating the original spreading code with a first NRZ secondary code to obtain the real component, and another NRZ secondary code in quadrature with the first NRZ secondary code (as a result of 1-bit quantization of a sine function) to obtain the imaginary component. Therefore, the explanations provided for Figures 29, 30, 31 and 32, which are applicable for a single stream of real spreading codes modulated with an NRZ waveform, can also be used for the real and imaginary parts of equation (30), respectively.

[0211] This paragraph describes the method of modulating a central carrier frequency (f central ) with respect to the frequency offset (Δf carr,p An alternative frequency offset applying means (11) is now described, further configured to apply a frequency offset of 1≦p≦P to said data signal carriers of said radio signal. An alternative formulation of equation (30) is given by the following equation:

number

[0212] Therefore, another interpretation of the application of a complex 1-bit quantized carrier offset, which results in another implementation, is to consider the chip stream obtained by multiplication of a spreading code equivalent to the primary code with a secondary code constructed from NRZ symbols as a longer spreading code, containing complex chips and whose period is equal to the least common multiple of the original spreading code period and the secondary code constructed from NRZ complex symbols.

[0213] In a similar manner, it is possible to interpret equation (29) as a modulation with an extended or longer spreading code composed of the original spreading code modulated with a secondary code composed of NRZ real symbols (reusing the structure described in (30)).

number

[0214] It can therefore be considered as the product of a spreading code (equal to the primary code) and a secondary code constructed from NRZ real symbols as a longer spreading code containing real chips and whose period is equal to the least common multiple of the original spreading code period and the secondary code constructed from NRZ real symbols.

[0215] It is recalled that the previous alternative embodiment, proposing the implementation of the carrier offset by a secondary NRZ code with complex or real symbols or a longer spreading code, is the result of the application of a 1-bit quantizer, in other words the use of the sign(●) operator. If a multi-level quantizer is applied, the result is that a multi-level quantized sinusoidal and / or complex exponential waveform has to be modulated onto the symbols of the secondary code modulated onto the spreading code. Based on this modification, a development similar to the one proposed for the 1-bit quantizer can therefore be proposed.

[0216] The previous implementation alternatives for the carrier offset applied for the generation of the transmitted signal components can also be directly proposed in the implementation of offset carrier cancellation at reception. Two main approaches are proposed to take into account the frequency offset applied to the radio signal components of interest at the receiver side. Their common objective is to suppress this offset in order to correlate with the spreading code delayed with the code delay hypothesis to be tested and modulated only with the Doppler hypothesis to be tested.

[0217] The first approach consists of including a frequency offset in the replica, and more precisely, adding a frequency offset to the Doppler hypothesis to be tested. Here, equation (4) describing the operation of correlation is therefore adapted by including the frequency offset corresponding to the signal component of interest in the argument of the exponential function that already includes the Doppler hypothesis to be tested.

number

[0218] If g0(t) represents a spreading code of length L and corresponds to the signal component of interest, then

number

[0219] Therefore, the local replica, also called the enhanced replica, is a complex offset carrier according to the frequency offset.

number

number

[0220] Alternative options can be proposed for generating the enhanced replica in a manner similar to the options proposed for the generation of the transmitted signal components. The first option consists in applying the real part of the enhanced replica, which is a real offset carrier depending on the frequency offset.

number

number

[0221] The second option is to quantize both the cosine and sine terms appearing in the complex enhanced replica expression, or to quantize only the real part of the enhanced replica expression. Nq A sign(●) operator can be used. In its simplest form, a sign(●) operator corresponding to 1-bit quantization is applied. The following equation (35) represents such an implementation for generating the local replica by multiplying the spreading code by a quantized complex carrier depending on the frequency offset, and the following equation (36) represents such an implementation for generating the local replica by multiplying the spreading code by a quantized real carrier depending on the frequency offset.

number

[0222] A further alternative option for generating the enhanced replica consists in considering that the sign (●) operator applied to a sine or cosine function results in an NRZ secondary code of symbols, which can be complex or real, which is then modulated onto a spreading code, in a manner similar to the generation of the transmitted radio signal components. A further and alternative option for generating the enhanced replica consists in considering that the combination of spreading sequences modulated with said NRZ code results in a longer spreading sequence. The following equation therefore represents the formula for the enhanced replica, generated by modulating a secondary NRZ code of complex symbols depending on said frequency offset onto a spreading code, or by generating a longer spreading code obtained by combining a secondary NRZ code of complex symbols depending on said frequency offset with said initial or original spreading code. To obtain this formula, the formula for the spreading code is

number

number

[0223] The following equations similarly represent the equations for an enhanced replica generated by modulating a secondary NRZ code of real symbols depending on the frequency offset onto a spreading code or by combining a secondary NRZ code of real symbols depending on the frequency offset with the initial or original spreading code to generate a longer spreading code obtained.

number

[0224] The second approach consists in first cancelling the frequency offset corresponding to the radio signal component of interest from the received signal, as if an additional downconversion to baseband had been applied. Here, two options are identified:

[0225] A first option for implementing this second approach consists in multiplying the received signal by a complex offset carrier depending on said frequency offset, which has an argument of frequency offset (-2πΔ carr,p ) by the exponential function

number

[0226] Alternatively, it is proposed to apply this step of further cancelling the frequency offset by multiplying the received signal by a real offset carrier which depends on the frequency offset. s'(t)=s(t)×cos(2πΔf carr,0 t), or s(t) × sin(2πΔf carr,0t) (40)

[0227] The second option for implementing the second approach is Quant Nq (●) consists in multiplying the received signal by a quantized complex carrier or exponential function obtained by the operator and depending on said frequency offset. In its simplest form, this consists in multiplying the received signal by a 1-bit quantized complex exponential function. s'(t)=s(t)×(sign(cos(2πΔf carr,0 t))-jsign(sin(2πΔf carr,0 t)) (41)

[0228] As an alternative option, you can also use Quant Nq It is also proposed to multiply the received signal by a quantized real carrier obtained with the (●) operator and depending on said frequency offset. In its simplest form, this consists in multiplying by a real term corresponding to a 1-bit quantization of the cosine or sine term. s'(t)=s(t)×(sign(cos(2πΔf carr,0 t))) (42)

[0229] According to one of the previous options for a further downconversion step to baseband, the further downconverted received signal s'(t) is correlated with a local replica generated only by the code delay hypothesis to be tested, and the Doppler hypothesis to be tested is calculated by the radio receiver again using equation (4).

number

[0230] In order to identify the proposed alternative methods used to implement the corresponding frequency offsets at the receiver side, it is proposed to represent in Figures 33a, 33b, 34a, 34b and 35a curves showing the maximum and highest percentiles of the distribution of CCF as a function of the (extended) interference Doppler residual for different frequency offsets of 0 KHz, 8.25 KHz, 16.5 KHz, 24.75 KHz and 33 KHz, and for the following cases: - the application of a conventional frequency offset with an exponential function (as in equation (17)) during transmission, and the conventional cancellation of the frequency offset by multiplying with an exponential function, shown in the leftmost diagram (also labeled "Analog Carrier"). - application of a conventional frequency offset with an exponential function (as in equation (17)) during transmission, and elimination of the frequency offset by multiplication with a 1-bit quantized complex exponential function, as shown in the central diagram (also labeled "Complex Digital Carrier"). - Application of a frequency offset with a 1-bit quantized complex exponential function during transmission, and cancellation of the frequency offset by multiplying with a 1-bit quantized real sine wave, shown in the diagram on the right (also labeled "Real Digital Carrier").

[0231] It should be noted that for frequency offsets where η is not an integer, a method has been applied to apply NRZ with plateaus of different lengths.

[0232] It is possible to verify that no significant degradation of the corresponding CCF insulation can be stated and, in particular, that the curve showing the maximum (and which usually serves as an envelope) does not increase. It is further outlined that this alternative scheme for implementing NRZ with possible different widths of plateaus depending on the carrier offset and thus on the η value does not significantly degrade the CCF insulation.

[0233] The previous alternative description of the embodiment of the present invention mainly described a method for designing, allocating and implementing frequency offsets for different radio transmitters transmitting radio signal components, as well as a method for processing the corresponding received radio signal components by a radio receiver in order to correlate at least one radio signal component of interest contained in a plurality of received radio signal components. Once this step of correlation has been performed, the next step required for the acquisition of the radio signal component of interest is similar to the steps described in the background section. This concerns in particular the generation of an acquisition detector based on the squared magnitude of the correlation function values ​​calculated for the code delay and Doppler offset hypotheses to be tested. To improve the acquisition performance, the acquisition detector can be further enhanced by summing different squared magnitudes of the correlation function values ​​calculated for the same code delay and Doppler offset hypothesis to be tested. The corresponding acquisition detector is therefore compared to a detection threshold. As an alternative acquisition procedure, which is usually applied when different code delay and Doppler offset hypotheses to be tested simultaneously are available, the ratio between the maximum value of the detection detector and a second maximum value calculated over the entire set of code delay and Doppler offset hypotheses is compared to another detection threshold. These alternative schemes for determining the correctness of the tested code delays and Doppler offsets based on detection thresholds set according to the specified PFA and PMD levels are also described in the Background section.

[0234] To illustrate the advantage of the proposed embodiment of the present invention, a measure C has been set up. This measure is defined as follows:

number

[0235] To justify the measure C to be applied for the embodiments of the invention, the following rationale is proposed. On the one hand, the spreading code length is proportional to the chip rate f cSince the GPS L5 radio signal component has to be linked to the L1 C / A radio signal component (for example, the GPS L5 radio signal component has a spreading code length of 10230 chips for a chip rate of 10.23 MCps, while the GPS L1 C / A radio signal component has a spreading code length of 1023 chips for a chip rate of 1.023 MCps, thus both result in the same spreading code period of 1 ms), and on the other hand other terrestrial or non-terrestrial radio navigation systems may apply completely different spreading code lengths (also related to the chip rate) and still meet the specified MAI and ranging performance requirements, it seems very arbitrary to consider first the spreading code length (expressed in chips) which is independent of the application. The second important term in the numerator is the radio receiver front-end bandwidth B, expressed in Hz. The front-end bandwidth can be considered as an important measure of the receiver complexity. In practice, a wider front-end bandwidth mandates the use of analog or digital filters, which may require a larger implementation load. Moreover, the sampling frequency is usually related or proportional to the front-end bandwidth according to the Nyquist law, which implies a larger power consumption of the radio receiver. Thus, by multiplying the radio receiver front-end bandwidth B expressed in units of Hertz by the spreading code period expressed in units of seconds, the product becomes integral, making a measure that is "independent" of the chip rate; a "hypothetical" radio navigation system based on a CDMA scheme and using a spreading code length with 1023 chips transmitted at 1 KCps would require a minimum radio receiver front-end bandwidth of 2 KHz (the width of the main lobe of the BPSK(1) signal waveform), which is equal to the product B [Hz] × (L × T c ) [s]. The value of this product is identical for the GPS C / A radio signal component, which has a spreading code length of 1023 chips, a chip rate of 1 MCps, and a radio receiver front-end bandwidth of 2 MHz.

[0236] Generally speaking, the product B [Hz] × (L × T c)[s] is better. In practice, having a lower bandwidth is beneficial for receiver complexity, but this is also true for the spreading code length (as it requires less memory and less computation for acquisition).

[0237] The denominator of measure C reflects the total number of radio signal components transmitted by the radio navigation system, since it is the product of transmitters Q with the number of radio signal components F transmitted by each transmitter. This can also be interpreted as the system capacity of the radio navigation system. To prove the location of the total number of radio signal components P in the denominator, the following explanation is given. Providing a larger capacity would be considered an advantageous feature of a radio navigation system, since the more transmitters there are, the larger the potential number of lines of sight that can be handled to improve positioning accuracy. Thus, intuitively the system capacity would be placed in the denominator (the larger the capacity, the smaller the measure C, and the better the overall performance). In the case of CDMA schemes, it is still not possible to increase the total number of radio signal components "indistinctly", otherwise the multiple access interference between all the radio signal components sharing the same frequency bandwidth would start to become unacceptable. This is why the second measure, defined by the specified cross-correlation isolation, would help to ensure that a reasonable number of radio signal components uniforms the MAI.

[0238] For the proof of the proposed measures, an illustrative example taken from one embodiment of the invention for a CDDMA scheme when considering the GPS C / A radio signal component as representative of the CDMA scheme, the GLONASS L1OCd radio signal component as representative of the FDMA scheme, and finally the same spreading code with a length of 31 chips is applied for the three main referenced use cases. Furthermore, and to simplify the comparison, it is proposed to consider the same number of satellites Q = 24 for all three radio satellite navigation systems (Q can be extended up to 6 or 8 additional operational satellites depending on the GPS or GLONASS system, and the accumulated number of satellites over different launches and generations, but this excess does not fundamentally change the main quantitative conclusions of this comparison, as will be shown later). Finally, it is emphasized again that the conclusions of the proposed comparison exercise applied to the aforementioned radio signal components from the GPS and GLONASS systems are extendable to other radio signal components, either from the same systems or from other systems such as the Galileo or COMPASS systems.

[0239] Considering the case of GPS C / A radio signal components transmitted in the L1 frequency band, the minimum bandwidth of the GPS C / A receiver front end is equal to 2.046 MHz, which corresponds to the main lobe width of the PSD of the BPSK(1) waveform of the GPS C / A radio signal components. Furthermore, the spreading code length is equal to 1023 chips transmitted at 1 MCps, resulting in a spreading code period of 1E-3s (expressed in seconds). Finally, since each GPS satellite transmits only one GPS C / A signal (F=1), and considering Q=24 GPS satellites as a baseline, C GPS This results in a scale of 2.046E6[Hz]×1E-3[s] / 24=85.25.

[0240] Considering the case of GLONASS L1OCd radio signal components transmitted in the L1 frequency band, the minimum receiver front-end bandwidth is equal to 12 MHz. In practice, each L1OCd signal is modulated with a BPSK (0.511) waveform, and therefore the minimum bandwidth per FDMA channel is equal to ~1 MHz. Furthermore, the diametrically opposed satellites use the same frequency channel. In conclusion, the total bandwidth, ensuring no overlap between FDMA channels, is equal to 24 x 1 MHz / 2 = 12 MHz. GLONASS L1OCd signals apply a spreading code length of 511 chips and a chip rate of 0.511 MCps, resulting in a spreading code period of 1E-3s (the same as for GPS C / A signals). Finally, since each GLONASS satellite transmits only one L1OCd signal (F=1), and considering Q=24 GLONASS satellites as a baseline, C GLONASS This yields a scale value of 12E6 [Hz] × 1E-3 [s] / 24 = 511. This value, which corresponds to the spreading code length of the L1OCd signal, can be derived using the parameters, assuming that two radio signal components are transmitted per FDMA channel and by two diametrically opposed satellites, and that the FDMA channel is a BPSK PSD (2f c ) with a minimum bandwidth B equal to the main lobe of FDMA Further assuming that the scale has C FDMA =(P / 2)×B FDMA xLxf c / P=P×(2×f c )×L×(1 / f c ) / P=L. It can be verified that the larger front-end bandwidth resulting from the FDMA scheme represents the main differentiator between the GPS and GLONASS cases and has an impact on the measure C.

[0241] The last case corresponds to the proposed embodiment of the present invention and the CDDMA scheme, applying the same spreading sequence, e.g. m-sequence, with a spreading code length of 31 chips, transmitted at a chip rate of 1 MCps, resulting in a spreading code period of ∼30.30E-6 s. Furthermore, the required bandwidth for this scheme is equal to 2.046 MHz + 12 × 8.5 KHz = 2.148E6 Hz. For this result, the offset between the CDDMA channels is given by (span D for typical interference residual Doppler of MEO satellite orbits, and frequency isolation margin ΔDopp isol It is further taken into consideration that the diametrically opposed satellites use the same CDDMA channel. Based on these numbers, and further taking into consideration that Q=24 satellites each transmit a single radio signal component (F=1), C CDDMA = 2.148E6 [Hz] × 30.30E-6 [s] / 24 = 2.71. It is noted that if the possibility of reusing the same CDDMA channel for the opposite satellite was not applied, the scale would not increase much at 2.84, since the occupied bandwidth would not increase much (2.046 MHz + 24 × 8.5 KHz = 2.148E6 Hz).

[0242] The previous comparison therefore highlights that the proposed embodiment of the present invention outperforms both GPS (CDMA) and GLONASS (FDMA) schemes based on the proposed measures. Nevertheless, for a fair comparison, an additional measure for cross-correlation isolation will also be considered. In fact, for the Gold code applied by the GPS C / A radio signal component, the maximum cross-correlation value is equal to ∼−22 dB, while this isolation reduces to ∼−15 dB when considering the same spreading code (e.g. m-sequence) with a length of 31 chips as an example for CDDMA (see FIG. 27). Therefore, for a fair comparison, a shorter spreading code length of ∼300 chips will be proposed for a set of Q different spreading codes applied for the CDMA scheme, sharing the same 2.046 MHz bandwidth and providing an isolation of approximately ∼−15 dB. For this second “hypothetical” radio navigation system using the CDMA scheme, the previous measure will be considered as follows: CDMA = 2.046E6[HZ] x 0,29E-3[S] / 24 = 25. This value, in the case of 25, is therefore proposed as the preferred value of the threshold for measure C, although another higher value of 50 may also be proposed as an alternative value for measure C. In practice, the specified measure 25 is based on configurations corresponding to existing GNSS systems such as GPS (CDMA) and GLONASS (FDMA), but it is pointed out that the proposed measure also depends on the application, and thus on the navigation or communication service.

[0243] [Reference 9]: "Resource Efficient Parallel Acquisition Architecture for Modernized GNSS Signals", J.Leclere, ECOLE POLYTECHNIQUE FEDERALE DE LAUSANNE, THESE NO6190 (2014)

[0244] [Reference 10]: "Performance Analysis of Modernized GNSS Signals Acquisition", M.Foucras.Thesis, 2015.Ecole Nationale d'Aviation Civile (ENAC)

[0245] To demonstrate the benefit of reducing the spreading code length in terms of receiver complexity reduction, it is proposed to apply a well-known technique based on exploiting the Fast Fourier Transform (FFT). One option of FFT-based acquisition allows to simultaneously test all code hypotheses in the code uncertainty region, since the spreading code periodicity allows to apply the cyclic property of the Fourier Transform, as shown in equation (Equation 45): where s represents the vector of samples of the received signal, while s rep represents the vector of samples of the replica spreading code. Before the FFT block, the exponential function e -J2πTsamPΔfDopp It is pointed out that multiplication with makes it possible to eliminate the Doppler being tested. For the sake of demonstration, it is finally proposed to consider only one coherent integration (i.e., no further incoherent summations). The same conclusion should hold for further incoherent summations. CCF(τ) = IFFT(FFT(S Rep )×conj(FFT(s)))τ∈{code uncertainty} (Equation 45)

[0246] Figure 36a shows the N s45 represents one implementation of (Equation 45) when considering a code length L of 1023 chips, resulting in L×U=2046 samples, in other words code hypotheses, and with U=2 samples per chip. Figure 36b now represents the corresponding implementation when applying a shorter code, with length L'=31 chips. In this case, the 1023 chips can be considered as 33 concatenations of the same spreading sequence of 31 chips. Thus, by computing the FFT product L' / L=33 times, the same cross-correlation function will be obtained, and FFT(s Rep )×conj(FFT(s')), where s' is the N' s = L' × U = 62 samples, and these 33 FFT products are accumulated (i.e. summed) before applying the IFFT operation. In this process, the FFT is calculated over 62 samples, again considering 2 samples per chip.

[0247] There are two main advantages to using a 62-point FFT instead of a 2046-point FFT. The first advantage is related to memory usage, since the number of registers implemented in the DSP that computes the FFT is (L x U) / (L' x U) = 2046 / 62 = 33 times smaller. This significant memory reduction is even true in spite of the presence of additional memory required for integration, which will have a small impact on the overall memory needs. The second advantage is related to the number of operations required to compute the correlation function. At a high level, the number of operations for the FFT is ∝N s ×log2N s is proportional to (N s ×log2N s ) / (N' s ×log2N' s)≈61. This semi-quantitative description does not claim to cover in detail how this principle can be reflected in each of the many different implementations of FFT-based acquisition. A non-exhaustive list of such implementations covers, for example, other techniques based on the Discrete Fourier Transform (DFT) or Finite Impulse Response (FIR) schemes, such as those presented in [Ref 9], which are calculated at several points, usually powers of 2, and apply the relevant Radix2, Radix4, Split Radix algorithms. Nevertheless, the general principle ([Ref 10]), according to which it is preferable to calculate and combine many FFTs calculated for short vectors, rather than a single FFT calculated for large vectors, is respected when considering short or very short codes. Short codes therefore represent an enabler for the application of this general principle. This advantage propagates to the remaining reduction in hardware complexity.

[0248] The description of the proposed solution provided above often refers to a satellite transmitter for transmitting radio (navigation) signal components as part of a Global Navigation Satellite System (GNSS) whose space segment includes Medium Earth Orbit (MEO) satellites. Illustrative and quantitative examples for the implementation of the proposed solution have been provided for such GNSS-MEO satellites. Nevertheless, it has already been emphasized that the proposed solution is not limited to this type of MEO platform, but is also applicable to other types of space-based platforms such as Low Earth Orbit (LEO) or Geostationary Earth Orbit (GEO) satellites, or any alternative near-earth platform such as High Altitude Platform Stations (HAPS), balloons or drones, as part of a radio navigation system. The solution can also be proposed for terrestrial "geostationary" platforms such as terrestrial navigation beacons, also called pseudolites or pseudolites.

[0249] The solution can also be made applicable for wireless communication systems in which MAI also represents a degradation effect to be overcome. Such wireless communication systems can comprise satellite-based platforms, such as, for example, satellite communication networks with satellites in GEO, MEO or LEO orbits, but also near-earth platforms, such as HAPS and drones, as well as terrestrial platforms, such as base transceiver stations (BTS) commonly used for terrestrial mobile network infrastructures, fixed or mobile radio transmitters in the case of wireless communication networks, or devices implemented in V2V or V2X communication networks.

[0250] Thus, even if the description and the corresponding preliminary quantitative performances used for the proof are mainly based on GNSS, some digressions can also be given when considering the aforementioned and alternative types of platforms or systems. For example, the general signal generation chain presented in FIG. 2 and proposed for GNSS satellites can be reused on board LEO, High Altitude Platform Systems (HAPS) or drones. The main difference will be the range of transmission power of the corresponding navigation signal, adapted for the antenna dimensions and possibly for the resulting spatial losses (function of the platform altitude) and guaranteeing a minimum and specified reception power level at the Earth's surface. This general signal generation chain can also be reused for pseudolites. Furthermore, for platforms that guarantee low or very low dynamics between the radio transmitter (as part of the communication or navigation system) and the radio receiver, such as GEO satellites or terrestrial platforms, the corresponding span of the observed Doppler offset is expected to be smaller or much smaller, which will allow to reduce the frequency offset implemented in each of the multiple Q radio transmitters and therefore to increase the number P of radio signal components for the same level of cross-correlation isolation.

[0251] Thus, the embodiments of the present invention described in the context of a radio navigation system can be made further applicable for wireless communication systems whenever a trade-off between transmitter and receiver bandwidth and complexity on the one hand, and MAI performance on the other hand, has to be resolved in order to capture the radio signal components of a wireless communication system.

Claims

1. Radio receiver (R ｘ ) in a plurality of radio signal components (S １ ,..S Ｐ ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ). １ . . S Ｐ ) are transmitted to a plurality of radio transmitters (T Ｘ１ . . . T ＸＱ ) are transmitted by respective wireless transmitters of the plurality of wireless signal components (S １ . . S Ｐ ) includes a data signal carrier for carrying a data signal component of the radio signal component, the data signal component including a spreading code, the spreading code including a predetermined number of chips (L), each chip of the spreading code having a duration (T ｃ ), said process comprising: - frequency offset (Δf ｃａｒｒ，ｐ , 1≦p≦P) with data signal carrier frequency (f ｃａｒ，ｐ = f ｃｅｎｔｒａｌ +Δf ｃａｒｒ，ｐ ) to obtain a data signal carrier having a center carrier frequency (f ｃｅｎｔｒａｌ ) with respect to the frequency offset (Δf ｃａｒｒ，ｐ , 1≦p≦P) to the data signal carriers of the radio signal components, and before transmission of each radio signal component, the radio transmitter (T Ｘ１ , . . . , T ＸＱ ), wherein the frequency offset corresponds to the wireless signal component of a plurality of wireless signal components; and transmitting at least one radio signal component to said plurality of radio transmitters (T Ｘ１ , . . . , T ＸＱ ) by a radio transmitter, each radio signal component including a data signal carrier carrying a data signal component modulated onto said data signal carrier; said plurality of radio signal components (S １ . . S Ｐ ) and stability (D ｏｓｃｉｌ，Ｒｘ ), and Ｘ１ . . . T ＸＱ ) for each wireless transmitter (V ＴＲＸ１ . . . V ＴＲＸＱ ) the wireless receiver (R ｘ ), each radio signal component being received by the radio receiver (R ｘ ) and an observed Doppler offset (D ) as a function of the velocity of the data signal carrier of the data signal carrier corresponding to each of the radio signal components. ＴＲＸ１ , . . . , D ＴＲＸＰ ) and receiving - Multiple radio transmitters (T) are detected by comparing the detection threshold with the capture detector. Ｘ１ . . . T ＸＱ ) Ｘｑ ) the plurality of radio signal components (S １ . . S Ｐ ) to a radio receiver (R) having a receiver front-end bandwidth (B) for receiving the at least one radio signal component of interest including the data signal carrier carrying the data signal component. ｘ ), wherein the acquisition detector is based on at least one correlation value obtained for a tested code and Doppler hypothesis of a set of code and Doppler hypotheses to be tested, and the correlation value is used to convert a received signal including a plurality of received radio signal components into a local replica of the data signal component including a spreading code corresponding to the signal component of interest, and into a local replica of the data signal component of the radio receiver (R ｘ ) is the coherent integration time (T ｉｎｔ ), the correlation values ​​including a combination of cross-correlation function values ​​corresponding to an interfering wireless signal component and one auto-correlation function value corresponding to a wireless signal component of interest; comprising the steps of: The predetermined number of chips (L) is determined by the spreading code length (L) and the chip duration (T ｃ ) the product of the radio receiver front-end bandwidth (B) and the total number of radio signal components (P) of said radio navigation system is such that the ratio does not exceed 25; - said frequency offset value is less than said chip rate of said spreading code; and - the frequency offset value is determined based on the receiver oscillator stability (D ｏｓｃｉｌ，Ｒｘ ) and the observed Doppler offset (D) for each of the interfering radio signal components of the plurality of radio signal components extended with a frequency offset corresponding to the interfering radio signal component. ＴＲＸ１ , . . . , D ＴＲＸＰ ) depending on a specified cross-correlation isolation applied to each of the cross-correlation function values ​​calculated for the interferometric Doppler residual values, the cross-correlation isolation being calculated as the difference between 1. A method for acquisition of at least one radio signal component of interest, comprising:

2. The method further comprises: filtering the received signal, including the plurality of received radio signal components, with the spreading code and further with the frequency offset (Δf ｃａｒｒ，ｐ 2. The method for acquisition of at least one radio signal component of interest according to claim 1, further comprising the step of correlating with a local replica comprising: a first replica of said first signal component and a second replica of said second signal component, said first replica including said first replica and said second replica;

3. The method further comprising: from the received signal comprising the plurality of received radio signal components, the respective frequency offsets (Δf ｃａｒｒ，ｐ , 1≦p≦P) to the wireless receiver (R ｘ ) to eliminate - comprising the plurality of received radio signal components, and each of said frequency offsets (Δf ｃａｒｒ，ｐ , 1≦p≦P) canceled received signal with a local replica of said data signal component including a spreading code corresponding to said signal component of interest; 2. The method for acquisition of at least one radio signal component of interest as recited in claim 1, further comprising the step of:

4. The spreading code is modulated with symbols, each symbol being the coherent integration time (T ｉｎｔ ) is the symbol duration (T ｓ ), which implies that the duration (T ｓ 4. The method for acquisition of at least one radio signal component according to claim 1, further comprising:

5. At least one radio signal component (S ｐ ) for transmitting a radio transmitter (T ｘｑ ), each of said radio signal components being a data signal carrier (D ｃａｒ，ｐ ), the data signal components include a spreading code, the spreading code including a predetermined number of chips (L), and each chip of the spreading code having a duration (T ｃ ) , and the wireless transmitter comprises: - frequency offset (Δf ｃａｒｒ，ｐ , 1≦p≦P) with data signal carrier frequency (f ｃａｒ，ｐ = f ｃｅｎｔｒａｌ +Δf ｃａｒｒ，ｐ ) to obtain a data signal carrier having a center carrier frequency (f ｃｅｎｔｒａｌ ) with respect to the frequency offset (Δf ｃａｒｒ，ｐ , 1≦p≦P) to the data signal carriers of the radio signal components prior to transmission of each radio signal component, the frequency offset corresponding to the radio signal components of a plurality of radio signal components; the frequency offset (Δf ｃａｒｒ，ｐ , 1≦p≦P)) on said data signal carrier of said corresponding radio signal component, said data signal component including a spreading code; - transmitting means (13) arranged to transmit said at least one radio signal component, each radio signal component comprising a data signal carrier carrying a data signal component modulated onto said data signal carrier of said corresponding radio signal component; Equipped with The predetermined number of chips (L) is determined by the spreading code length (L) and the chip duration (T ｃ ) the radio receiver front-end bandwidth (B) and the total number of radio signal components (P) of the radio navigation system, is such that the ratio does not exceed 25; the frequency offset value is less than the chip rate of the spreading code; - the frequency offset value is determined based on the receiver oscillator stability (D ｏｓｃｉｌ，Ｒｘ ) for each of the interfering data signal components of the plurality of radio signal components, augmented with a frequency offset corresponding to the interfering radio signal component. ＴＲＸ１ , . . . , D ＴＲＸＰ ) depending on a specified cross-correlation isolation applied to each of said cross-correlation function values ​​calculated for an interferometric Doppler residual value, the cross-correlation isolation being calculated as the difference between A wireless transmitter (T ｘｑ ).

6. The modulation means is adapted to modulate the spreading code with symbols, each symbol having a duration (T ｓ 6. A radio transmitter (T) according to claim 5, ｘｑ ).

7. The frequency offset application means (11) applies a center carrier frequency (f ｃｅｎｔｒａｌ ) with respect to the frequency offset (Δf ｃａｒｒ，ｐ 6. A radio transmitter (T) according to claim 5, further configured to apply a plurality of 1≦p≦P to the data signal carriers of the radio signal components. ｘｑ ).

8. The frequency offset application means (11) applies a center carrier frequency (f ｃｅｎｔｒａｌ ) with respect to the frequency offset (Δf ｃａｒｒ，ｐ 6. A radio transmitter (T) according to claim 5, further configured to apply a plurality of 1≦p≦P to the data signal carriers of the radio signal components. ｘｑ ).

9. The frequency offset application means (11) modulates the spreading code with a secondary NRZ code of complex or real symbols according to the frequency offset, thereby adjusting the center carrier frequency (f ｃｅｎｔｒａｌ ) with respect to the frequency offset (Δf ｃａｒｒ，ｐ 6. A radio transmitter (T) according to claim 5, further configured to apply a plurality of 1≦p≦P to the data signal carriers of the radio signal components. ｘｑ ).

10. The frequency offset application means (11) applies a central carrier frequency (f) to the central carrier frequency (f) by modulating a longer spreading code obtained as a combination of the secondary NRZ code of complex or real symbols according to the frequency offset modulated by the spreading code. ｃｅｎｔｒａｌ ) with respect to the frequency offset (Δf ｃａｒｒ，ｐ 6. A radio transmitter (T) according to claim 5, further configured to apply a plurality of 1≦p≦P to the data signal carriers of the radio signal components. ｘｑ ).

11. A plurality of radio signal components (S １ . . S Ｐ ) for the acquisition of at least one radio signal component of interest. ｘ ), wherein each of said plurality of radio signal components (S1 ..SP) is transmitted to a plurality of radio transmitters (T Ｘ１ . . . T ＸＱ ) are transmitted by respective wireless transmitters of the plurality of wireless signal components (S １ . . S Ｐ ) each radio signal component includes a data signal carrier (D) for carrying a data signal component of said radio signal component. ｃａｒ，ｐ ), the data signal components include a spreading code, the spreading code including a predetermined number of chips (L), each chip of the spreading code having a duration (Tc), and the radio receiver (R ｘ ) is stability (D ｏｓｃｉｌ，Ｒｘ ), and Ｘ１ . . . T ＸＱ ) for each wireless transmitter (V ＴＲＸ１ . . . V ＴＲＸＱ ), and the wireless receiver (R ｘ )but, receiving means (21) arranged to receive a plurality of radio signal components, each of which is connected to said radio receiver (R) for said corresponding radio transmitter; ｘ receiving means (21) having an observed Doppler offset responsive to said velocity of said first radio signal component and said data signal carrier frequency of said data signal carrier corresponding to said respective radio signal component; - Multiple radio transmitters (T) are detected by comparing the detection threshold with the capture detector. Ｘ１ . . . T ＸＱ ) Ｘｑ ) the plurality of radio signal components (S １ . . S Ｐ ), wherein the acquisition detector is adapted to acquire, by the radio receiver (Rx) having a receiver front-end bandwidth (B), the at least one radio signal component of interest comprising the data signal carrier carrying the data signal component from the at least one radio signal component of a set of code and Doppler hypotheses to be tested, the acquisition detector being adapted to analyze a received signal comprising a plurality of received radio signal components based on at least one correlation value obtained for a tested code and Doppler hypothesis of a set of code and Doppler hypotheses to be tested, the correlation value being determined by dividing the received signal comprising a local replica of the data signal component comprising a spreading code corresponding to the signal component of interest by the radio receiver (Rx) having a receiver front-end bandwidth (B). ｘ ) is the coherent integration time (T ｉｎｔ ), said correlation values ​​comprising a combination of cross-correlation function values ​​corresponding to interfering radio signal components and one auto-correlation function value corresponding to the radio signal component of interest; Equipped with said predetermined number of chips (L) is such that a measure calculated as the ratio between the product of the radio receiver front-end bandwidth (B) with the spreading code length (L) and with the chip duration (Tc) and the total number (P) of radio signal components of the radio navigation system does not exceed 25; the frequency offset value is less than the chip rate of the spreading code; - the frequency offset value is determined based on the receiver oscillator stability (D ｏｓｃｉｌ，Ｒｘ relying on a specified cross-correlation isolation applied to each of the cross-correlation function values ​​calculated for an interfering Doppler residual value calculated as a difference between a Doppler hypothesis being tested, taking into account also a Doppler hypothesis being tested, and the observed Doppler offset for each of the interfering radio signal components of the plurality of radio signal components extended with a frequency offset corresponding to the interfering radio signal component; A radio receiver (R) for the acquisition of at least one radio signal component of interest, characterized in that ｘ ).

12. The processing means (22) processes the received signal comprising the plurality of received radio signal components into the spreading code and further the frequency offset (Δf ｃａｒｒ，ｐ 12. A radio receiver (R) for the acquisition of at least one radio signal component of interest according to claim 11, characterized in that it is adapted to correlate with a local replica comprising a plurality of replicas (R 1 , 1≦p≦P) ｘ ).

13. The wireless receiver (R ｘ )but, said processing means (22) extracting from a received signal comprising said plurality of received radio signal components said respective frequency offsets (Δf ｃａｒｒ，ｐ , 1≦p≦P); and the processing means (22) is further adapted to correlate the received signal, comprising the plurality of received radio signal components and having the respective frequency offsets (Δf,p, 1≦p≦P) cancelled, with a local replica of the data signal component comprising a spreading code corresponding to the signal component of interest; A radio receiver (R) for the acquisition of at least one radio signal component of interest according to claim 11, further comprising: ｘ ).

14. 13. The radio receiver (R) according to claim 12, characterized in that the processing means (22) are arranged to generate the local replica by multiplying the spreading code by a complex or real offset carrier depending on the frequency offset. ｘ ).

15. 13. The radio receiver (R) according to claim 12, characterized in that the processing means (22) are arranged to generate the local replica by multiplying the spreading code by a quantized complex or real offset carrier depending on the frequency offset. ｘ ).

16. 13. A radio receiver (R) according to claim 12, characterized in that the processing means (22) are adapted to generate the local replica by modulating a secondary NRZ code of complex or real symbols depending on the frequency offset onto the spreading code. ｘ ).

17. 13. A radio receiver (R) according to claim 12, characterized in that the processing means (22) are adapted to generate the local replica by generating a longer spreading code obtained by combining with the spreading code a secondary NRZ code of complex or real symbols depending on the frequency offset. ｘ ).

18. 14. A radio receiver (R) according to claim 13, characterized in that the processing means (22) are arranged to cancel the frequency offset from the received signal by multiplying the received signal by a complex or real offset carrier depending on the frequency offset. ｘ ).

19. 14. A radio receiver (R) according to claim 13, characterized in that the processing means (22) are arranged to cancel the frequency offset from the received signal by multiplying the received signal by a quantized complex or real offset carrier depending on the frequency offset. ｘ ).

20. Radio receiver for radio navigation system (R ｘ ) at least one of the plurality of radio signal components (S １ . . S Ｐ ), the radio navigation system for the acquisition of the plurality of radio signal components (S １ . . S Ｐ ) are transmitted to a plurality of radio transmitters (T) of the radio navigation system. Ｘ１ . . . T ＸＱ ) are transmitted by respective wireless transmitters of the plurality of wireless signal components (S １ . . S Ｐ ) includes a data signal carrier for carrying a data signal component of the radio signal component, the data signal component including a spreading code, the spreading code including a predetermined number of chips (L), each chip of the spreading code having a duration (T ｃ ), and the radio navigation system comprises: At least one radio transmitter (T) according to claims 5 to 10 Ｘｑ ) The radio navigation system comprises: The radio receiver (R) according to any one of claims 11 to 19 ｘ ) 23. A radio navigation system comprising:

Images