GB2623556A

GB2623556A - Direct sequence spread spectrum coding scheme for a transmitter and receiver

Info

Publication number: GB2623556A
Application number: GB2215495.9A
Authority: GB
Inventors: David Macleod Malcolm
Original assignee: Qinetiq Ltd
Current assignee: Qinetiq Ltd
Priority date: 2022-10-20
Filing date: 2022-10-20
Publication date: 2024-04-24
Also published as: GB202215495D0; WO2024083969A1

Abstract

A transmitter and receiver arranged to implement a direct sequence spread spectrum sequence (DSSS) coding scheme, such as for a Global Navigation Satellite System (GNSS). The transmitter multiplies a DSSS spreading code, comprising a sequence of l chips, with an overlay code. The overlay code is selected based on the value of a plurality of data bits, representing additional data to be encoded on the DSSS signal. The overlay code comprises a sequence of m chips (m<l). Each chip of the overlay code is multiplied with a plurality of contiguous DSSS chips. On reception, a correlation process extracts the overlay code used, and hence the information encoded therein. This involves correlating a received signal with a local copy of the spreading code, in a segmented manner. The result of each of the segment correlations is further correlated with each possible instance of the overlay code. The further correlations identify the overly code selected by the transmitter and hence the value of the additional data. Embodiments provide for additional data to be transferred over a DSSS link.

Description

Direct Sequence Spread Spectrum Coding Scheme for a Transmitter and Receiver The present invention relates to data coding schemes, and corresponding decoding schemes, as may be used in transmitters and receivers. More particularly, it relates to 5 such schemes as may be used in Direct Sequence Spread Spectrum (DSSS) systems, such as Global Navigation Satellite Systems (GNSS) and others.

GNSS transmissions from a satellite comprise of data that is used, in combination with those of other satellites, to provide a receiver the ability to calculate its position, velocity, or the time (PVT). Each satellite sends a large bandwidth signal, that comprises of data bits modulated by a spreading code. The successful detection of the spreading code, and the data bits encoded therein provide a means for giving a high accuracy PVT measurement, along with the necessary satellite ephemeris information. The satellite ephemeris information is contained in the data bits, and is traditionally sent at a very low data rate, of typically 50bps or so. There is a desire to increase this data rate, so as to be able to send additional information such as redundant, or more accurate ephemeris data or commercial data. One method of doing this is to encode the spreading code using a Cyclic Code Shift Keying (CCSK) technique. This involves encoding the bits to be sent according to a particular circular shift of the spreading code. If, say, 16 (24) different shifts of the spreading code were used, then 4 bits are transmitted for each transmission of the spreading code, with an extra bit being available based upon the polarity of the transmission. This gives a very useful increase in the achievable data rate. A downside of this approach is the additional correlation of the received signal that is required in the receiver, being a factor of 2n increase over the case where CCSK is not used, where n is the number of bits being transmitted by the CCSK process. This can be a substantial amount of effort in receivers, which are often portable devices with limited battery and processing power availability.

The following prior art documents describe CCSK systems, and represent the state of the

art in the field:

A. Garcia-Pena, M.-L. Boucheret, C. Macabiau, J.-L. Damidaux, L. Ries, et al., "Implementation of Code Shift Keying signaling technique in GALILEO El signal", NAVITEC 2010, 5th ESA Workshop on Satellite Navigation Technologies and European Workshop on GNSS Signals, Dec 2010, Noordwijk, Netherlands. pp 1-8.

A. Garcia-Pena, D. Salos, 0. Julien, L. Ries, T. Grelier, "Analysis of the use of CSK for future GNSS Signals", ION GNSS 2013, Sept 2013.

Axel Garcia, Philippe Paimblanc, Olivier Julien, Lionel Ries, Thomas Grelier, "Analysis of different CSK configurations in an urban environment when using non-coherent demodulation", NAVITEC 2014.

Riccardo Andreotti, Andrea Emmanuele, Diana Fontanella, Marco Luise, "Code-Shift-Keying (CSK) with Advanced FEC Coding for GNSS Applications in Satellite Multipath Channel", NAVITEC 2014 It is therefore desired to provide an alternative to the prior art that overcomes or mitigates at least some of the problems or issues therewith.

According to a first aspect of the present invention there is provided a coding system for a Direct Sequence Spread Spectrum (DSSS) transmission link, the transmission link comprising a transmitter and a receiver, and the coding system comprising of a modulator in the transmitter, and a demodulator in the receiver; the transmitter being arranged to: t1) modify, in the modulator, a spreading code comprising of a sequence of / chips by multiplying it with a second signal comprising an instance of an overlay code, the instance being chosen according the value of a plurality of data bits from an information bit stream to be modulated onto the spreading code, wherein the overlay code comprises of in overlay code chips to be encoded onto the spreading code, where each chip of the overlay code is multiplied with a plurality of chips of the spreading code to produce a modified spread spectrum code; and t2) transmit the resulting modified spread spectrum code; and the receiver being arranged to: r1) receive the modified spread spectrum code transmitted by the transmitter; and r2) correlate the modified, received spread spectrum code with a local copy of the spreading code in a segmented manner, wherein each segment is of a size approximately that of the number of chips of the spreading code that are multiplied by each chip of the overlay code.

r3), correlating the resulting set of segment correlations with each possible instance of the overlay code, choosing the resulting instance having the highest amplitude, and identifying the value of the information bit steam based upon the comparison.

Embodiments of the invention therefore allow for an increased data transfer rate between a transmitter and a receiver, whilst having a reduced computational effort within the receiver as compared to a system using prior art CCSK modulation.

In some embodiments the computational effort within the receiver in correlating the incoming signal is the same or very similar to that of an equivalent binary phase shift keying (BPSK) DSSS system that is not implementing a CCSK system. In other words, such embodiments pose little or no additional effort in correlation computation over a system not transmitting additional information using a CCSK system. As is explained below, there is a small amount of additional effort required in processing the results of the segmented correlator, but embodiments of the invention are able to make a net saving of computational effort over prior art techniques, and in some embodiments a significant net saving of effort.

Preferably the overlay code comprises of a plurality of instances, each instance comprising of a sequence of chips, and each instance differing from the others in the values of approximately half of their chips. Each instance is preferably orthogonal, or near-orthogonal to each other instance.

In some embodiments of the invention the overlay code may comprise a Walsh code. Walsh codes have a length of n bits where n is a power of 2 (for example, 4, 8, 16 or 32). A Walsh code of a given length of n bits has the property that the various instances of it differ by n/2 bits from the other instances, and are therefore mutually orthogonal.

Of course, even where an orthogonal overlay code is used, in some embodiments the size of the spreading code may not be a multiple of n, which means that each bit of the overlay code cannot be multiplied by the same number of chips of the spreading code as every other bit of the overlay code, which means that true orthogonality of the modified spreading code will not be possible, but in most practical implementations, any deviation from this true orthogonality is likely to be very minor in its consequence. This is because in general there are a large number of chips in the spreading code, such as 1023 chips, so it is possible to choose an overlay code such that only one bit of the overlay code is used to modify one less chip of the spreading code, which has a very small effect on the orthogonality.

Using a Walsh code imparts the benefit that a particularly efficient means may be used to identify the particular bit transmitted, at step (r3). If the results of the segmented correlations are applied to an Inverse Walsh transform (also known as an Inverse WalshHadamard transform), the largest output of the Inverse Walsh transform will that corresponding to the particular overlay code instance that was used in the transmission process, which indicates the value of the transmitted information bit(s). Thus, in some embodiments an Inverse Walsh transform is used in the identification of the value of the bit stream at stage (r3).

Alternatively, some embodiments of the invention may use an overlay code that comprises of a Kasami code.

Other overlay codes may be used, where preferably the bit values that make up the overlay code differ in approximately 50% of cases between instances of the code having a given length n.

Advantageously, in some embodiments, step (t1) also includes a step of multiplying the output of the modified spreading code by a further bit of information data, that is arranged to change the polarity of the modified spreading code according to the value of the further bit. This increases the number of information bits transmittable by one.

In some embodiments of the invention the spreading code used does not repeat from one instance of length / chips to the next. Thus, it is useable in non-repeating codes such as cryptographically encoded codes. Of course, other embodiments may use repeated spreading codes.

In some embodiments of the invention, in step (r3), the receiver is arranged to store successive segmented correlations of the received, modified spread spectrum code using a shift register, and to form combinations of inversions of each, and comparing the combinations with each instance of the overlay code, to identify which overlay code instance has been used, and providing as an output the bit stream associated with the identified instance.

Where a Walsh code is used, then the comparison of the combinations may conveniently be performed using an Inverse Walsh transform (also known as an Inverse WalshHadamard transform).

By carrying out this process, wherein, for a given overlay code instance, the combinations are summed to produce an effective full correlation of the given overlay code instance, and a correlation of each of the possible overlay codes is formed. The one having the greatest value is assumed to be the overlay code instance transmitted, and the information bits associated with the particular instance are regarded as received, and are provided as an output.

According to a further aspect of the present invention there is provided a transmitter comprising a modulator arranged to implement the steps t1 and t2 of the first aspect.

According to a further aspect of the present invention there is provided a receiver comprising a demodulator arranged to implement the steps r1, r2 and T3 of the first aspect.

The invention will now be described, by way of example only, with reference to the following Figures, of which: Figure 1 shows at a top level some key components of a transmitter and receiver, of the type where an embodiment of the invention may be used; Figure 2 shows more detail of a modulator that may be used in a transmitter according to an embodiment of the present invention; and Figure 3 shows more detail of a demodulator as may be used in a receiver according to an embodiment of the present invention.

Shown in Figure 1, and in a top-level, simplified form, is a transmitter 100, and a receiver 102, as may be used in prior art systems, or which may employ an embodiment of the present invention. The transmitter in this instance is a satellite, whereas the receiver 102 is a portable GNSS receiver, arranged to receive signals from the satellite 100, and from other similar satellites. The transmitter 100 comprises of a signal generator 104, which produces a spreading code that is fed to a modulator 106 which, in prior art systems may shift the spreading code according to an information bit stream fed to it, and in systems according to embodiments of the present invention will multiply the spreading code with an overlay code. The modulated signal is then fed to an up-convertor 108, from there to an amplifier 110 then to an antenna 112. The signal is initially processed as a digital signal and is converted to an analogue one following the modulation process, as would readily be appreciated by the normally skilled person.

The receiver 102 comprises of a receive antenna 114, which receives the signal sent by the transmitter 100 and feeds it to a low noise amplifier 116 and to a down-converter 118 which brings the signal down in frequency to baseband. It is then demodulated in demodulator 120 to extract the timing and other information that was added during the transmission process. Digitisation can take place at different points depending upon the implementation. The demodulation step will, in the prior art, comprise of correlating the received signal (after conversion to base band) with each possible shift as imposed in the CCSK modulation process, so as to determine which shift (and hence which information bit(s)) were transmitted. Demodulation in the case of an embodiment of the present invention is described herein.

The transmitter 100 is arranged to transmit a GNSS navigation signal having information embedded therein comprising timing and positional information relating to the satellite's position in space. The signal being transmitted is a Direct Sequence Spread Spectrum (DSSS) signal comprising a repeated sequence of chips, that are sent at a rate that enables a suitably configured receiver to, in combination with signals received from the other satellites, compute its location. Such receivers are of course commonplace, being present in many mobile phones, as well as dedicated handheld and car navigation systems. There is a desire among GNSS operators to increase the amount of information that can be transmitted, and CCSK is a known way of providing this. Although it is relatively simple to implement in a transmitter, as explained above, there is an implementation cost in the receiver, with significant additional correlation being required to decode the CCSK signal and extract the additional transmitted information. In mobile phone, and other hand held devices, this is disadvantageous due to the additional processing power required, which impacts on receiver battery life.

Figure 2 shows in more detail a part of the transmitter, more specifically the modulator 106, as described above, as may be used in an embodiment of the invention. A 1-bit multiplier 200 has a first input from a signal generator (such as signal generator 104 of Figure 1) and a second input comprising of an instance of overlay code, from an overlay code generator 202. The overlay code generator 202 selects a code dependent upon bits within the information bit stream. The signal generator provides a prototype ranging or spreading code, that is to be modified as described. The result of the multiplication may then be multiplied in multiplier 204 by a single additional data bit taken from the information bit stream, that either leaves the code as is, or inverts the whole code, so codifying this additional data bit. This produces an output code that is then transmitted.

It can be seen then that whereas prior art systems use Ai different cyclic shifts of a spreading code to codify log2M = B bits in an information bit stream, an embodiment of the invention instead modifies the spreading code by multiplying it with one of Ad different instances of an overlay code (according to the values of the B bits of information in the information bit stream) that are stored or produced by the overlay code generator 202, to again transmit B bits of information (or B + 1 bits if the additional inversion bit is used at multiplier 204). Here, as the spreading code produced by code signal generator 104 is being modified (by more than just a cyclic shift) before transmission, it is known as a prototype ranging code. The spreading / prototype ranging code generated at 104, and the overlay code generated at 202, each comprise of a sequence of chips having binary values ±1.

Because embodiments of the present invention do not rely on M shifts of a spreading code to encode the additional information, it is not correct to call an implementation thereof a CCSK system. Instead, as they rely on orthogonal (or nearly-orthogonal) overlay code, it may be called instead an Orthogonal M-ary Overlay Code (OMOC) modulation scheme.

Both the prior art and the present invention result in the transmission of (when the spreading code has been well designed) orthogonal, or near orthogonal codes for transmission, resulting in an optimal or near-optimal ability to differentiate, in the receiver, between different information bits transmitted. However, the effort required to carry out the extraction in the receiver in an embodiment of the present invention is much reduced

as compared to the prior art.

In the present invention, a transmitted (output) code is generated by multiplying the chips of the prototype code by the chips of a further code, namely the overlay code mentioned above, which is of length in chips. There are 2' possible overlay codes of length In chips, from which a set of ill is selected for use, as discussed further below.

Within a transmitter arranged to produce a code according to this embodiment, the transmitter selects one of the selected set of M overlay codes. The transmitter may then also multiply the whole result by +1 to carry one further bit of data, as in conventional (biorthogonal) BPSK modulation. Each overlay code chip multiplies one or more prototype code chips.

As an example, consider In = 4 and 41= 2. Let the first instance of the overlay code be four +1s; then following the multiplication of the overlay code with the prototype code, the output code is the prototype code itself. Then for the second instance of the overlay code, consider as an example one in which +1 and -1 alternate: [+1 -1 +1 -1]. Then half the chips of the resulting output code are the same as the original prototype code and half are different from it. These two output codes may now be transmitted according to the value of an information bit stream that is being transmitted by the code.

It will be appreciated by the normally skilled person that if each overlay code chip multiplies exactly the same number of prototype code chips, this would make the second code orthogonal to the first one.

However in the case of GNSS signals it is usual for prototype ranging (i.e. spreading) codes to have a number of chips which is an integer multiple of 1023 = 210_ 1. For most values of overlay code length, in, it is then impossible for each overlay code chip to multiply exactly the same number of prototype code chips. However it is then desirable (as can be verified by computer simulation) for each overlay code chip to be used to multiply a similar number of prototype code chips.

If the length of a prototype ranging code which carries one data symbol is /, each overlay code chip is used to multiply approximately //m chips of the prototype ranging code. If (as below) in is made a power of 2, then if 1 is a multiple of 1023, a simple strategy is for the first m -1 overlay code chips to be used to multiply exactly (1+1)1 ni chips of the prototype ranging code, and the final overlay code chip to multiply (/+1)/m -1 chips.

The unequal number of chips mean that the resulting codes are not completely orthogonal to each other, but the discrepancy is very small and in practice is negligible, because in GNSS the orthogonality of codes is in any case slightly reduced by the differing relative delays and Doppler frequency shifts between the signals from different satellites. This slight lack of complete orthogonality is an example of "near-orthogonality" as mentioned herein, where the slight lack of orthogonality is small enough not to be significant in any detection process within a receiver. Although in this example the orthogonality of the output code will not be absolute, it will be very high. It will be appreciated that in other embodiments, where the orthogonality of the resulting output signal is lower, the invention still has efficacy.

Another possible choice of overlay code is a Gold code. These codes, which are widely used, for example in GNSS systems, have lengths of the form 2N-1, for example 31 (if N = 5). Each Gold code is nearly orthogonal to the others of that length. Length-31 Gold codes have the advantage that 31 divides exactly into 1023, so each bit of the overlay code can be arranged to multiply the same number of chips (that is, 33) of the ranging code.

The multiplication of the prototype code with the overlay code can very easily be applied in the transmitter, and can be used whether the prototype code is periodic or non-periodic.

It will also be appreciated that in a prior art CCSK implementation, the deviation from orthogonality of differently-shifted codes relative to each other depends on the autocorrelation function of the prototype code. In the OMOC method presented herein, it is achieved as a consequence of the fundamental properties of the overlay code. It is thus more convenient to choose a spreading code for such an embodiment, as its autocorrelafion properties do not need to be considered for each potential cyclic shift in the code.

Figure 3 shows in simplified form an embodiment of a receiver 300 arranged to decode a received OMOC signal. An input signal 302, from the front end of a receiver (e.g. the antenna 114 and amplifier 116 as shown in Figure 1) is provided to a mixer 302 which multiplies the input signal with a local oscillator signal 304 to bring it down to base band.

As is usual for receivers of digital signals, the frequency offset and the phase offset which are present in the received signal are removed by a Carrier frequency and phase wipeoff process, which employs a Frequency Locked Loop and/or Phase Locked Loop. Once the phase is wiped off, the modulating chips are present purely in the either the real part of the signal (or for some signals, purely in the imaginary part). To extract them, the real (or imaginary) part is then taken 306, and multiplied in multiplier 308 with a locally stored copy of the prototype ranging code 310. This is then fed into an accumulate-and-dump circuit 312. So far, this would also form part of a conventional BPSK or CCSK demodulator.

However instead of accumulating over the entire! chips, as would be done in a CCSK system, the first accumulation is only over the first section of chips, corresponding to those which were modulated by the first overlay code chip, and the resulting sum (which is called a "subcorrelation" or "partial correlation") is fed into a shift register 314 which, along with the output of the accumulate-and-dump function, is able to store in partial correlations. This process is repeated, filling the shift register with m subcorrelations, matching the in overlay code chips. The outputs of each register of the shift register, along with that of the accumulate-and-dump function are provided to an overlay code decoder.

This overlay code decoder then correlates these m values with each in turn of the Al possible orthogonal codes imposed at the transmitter. In a direct implementation of this correlation operation, each of the In sub-correlations stored in the shift register 314 is multiplied by the corresponding binary-valued chip of the overlay code, and the results of those multiplications are summed to give a correlation output. The decoder selects the correlation output which is largest in magnitude and outputs its sign together with data bits, representing a part of the information bit stream, which codify which of the /1/ overlay codes was used.

It can be seen that the total amount of correlation of input chips required in the receiver of this embodiment (that is to say the numbers of multiplications and additions required) is the same as a single correlation with / chips, whereas in a prior art CCSK system it is AI times greater.

All the multiplications in these correlations are by 1-bit coefficients, so they are implemented as just switchable addition/subtraction operations.

If the correlations in the overlay code decoder are implemented directly, as described above, then that requires Aix m 1-bit-multiplications and additions. This too is much smaller than the number required in a CCSK system. However a further significant reduction in computation may be achieved if Walsh codes are used for the overlay codes.

As stated above, it is advantageous, but not necessary, to use Walsh codes for the overlay codes. These are a class of binary orthogonal codes. These have power-of-two length, m. Each code is a Walsh function. The Walsh functions form an orthogonal set, which means that each one differs from any other in exactly mI2 chips. The set comprises /1Jfunctions, where A/ = in.

Using Walsh functions, the implementation of the modulator in the transmitter, and in the overlay code decoder in the receiver is straightforward. There are algebraic methods for generating Walsh functions which make an efficient real-time generator practical in many embodiments. It would alternatively be possible to store the AI different overlay codes in receiver memory.

If Walsh functions are used, the overlay code decoder may be implemented using an Inverse Fast Walsh Transform (also known as an Inverse Walsh-Hadamard Transform), whose structure is similar to that of a fast Fourier transform (FFT), but all the multiplications within it are by 1-bit coefficients, so they are implemented as just switchable addition/subtraction operations.

The whole transform requires 131 109201) of them. This is a very small computational load.

For example, given a spreading code length! of 4092 and if in is 64 (which, coupled with the overall sign multiplication, transmits 7 bits of data), the correlation process requires 4092 addition/subtractions and the Overlay Code decoder process requires 384. The process of selecting the maximum output requires another 64 giving a total of 4540. By contrast, CCSK would require up to 64 x 4092 + 64 = 261,952 additions or subtractions.

Although the invention has been described largely in relation to GNSS applications, it has utility in other systems where CCSK may be employed, such as communications systems.

It will be appreciated by the person skilled in the art that features in one aspect of the invention may be applied to other aspects of the invention, in any appropriate combination. In particular, method aspects may be applied to apparatus aspects, and vice-versa. It will further be understood that the various embodiments disclosed herein have been described for the purposes of illustration, and that modifications may be made without departing from the scope of the present disclosure.

Where appropriate, features implemented in hardware may be instead implemented in programmable hardware (such as FPGA etc.), firmware or software, and vice versa. References to "a computer" or the like, may relate to a single computer, or may refer to a group, or network of computers, that may be co-located, or may be distributed. Such a network may be connected using wired, or wireless data links.

Claims

Claims 1. A coding system for a Direct Sequence Spread Spectrum (DSSS) transmission link, the transmission link comprising a transmitter and a receiver, and the coding system comprising of a modulator in the transmitter, and a demodulator in the receiver; the transmitter being arranged to: t1) modify, in the modulator, a spreading code comprising of a sequence of / chips by multiplying it with a second signal comprising an instance of an overlay code, the instance being chosen according the value of a plurality of data bits from an information bit stream to be modulated onto the spreading code, wherein the overlay code comprises of In overlay code chips to be encoded onto the spreading code, where each chip of the overlay code is multiplied with a plurality of chips of the spreading code to produce a modified spread spectrum code; and t2) transmit the resulting modified spread spectrum code; and the receiver being arranged to: r1) receive the modified spread spectrum code transmitted by the transmitter; and r2) correlate the modified, received spread spectrum code with a local copy of the spreading code in a segmented manner, wherein each segment is of a size approximately that of the number of chips of the spreading code that are multiplied by each chip of the overlay code.r3), correlating the resulting set of segment correlations with each possible instance of the overlay code, choosing the resulting instance having the highest amplitude, and identifying the value of the information bit steam based upon the comparison.
2. A system as claimed in claim 1 wherein the overlay code comprises of a plurality of instances, each comprising of a sequence of chips, and each instance differing from the others in approximately half of their chips.
3. A coding system as claimed in claim 1 or claim 2 wherein the overlay code comprises of a Walsh code.
4. A coding system as claimed in claim 3 wherein an Inverse Walsh transform is used in the identification of the value of the information bit stream at stage (r3).
5. A coding system as claimed in claim 1 or claim 2 wherein the overlay code comprises of a Kasami code.
6. A coding system as claimed in any of claims 1 to 5 wherein in and / are chosen such that in divides /into approximately equal parts.
7. A coding system as claimed in any of the above claims wherein step (t1) also includes a step of multiplying the output of the modified spreading code by a further bit of information data, that is arranged to change the polarity of the modified spreading code according to the value of the further bit.
8. A coding system as claimed in any of the above claims wherein the spreading code used does not repeat from one instance of length / chips to the next.
9. A coding system as claimed in any of the above claims wherein, in step (r3), the receiver is arranged to store successive segmented correlations of the received, modified spread spectrum code using a shift register, and to form combinations of inversions of each, and comparing the combinations with each instance of the overlay code, to identify which overlay code instance has been used, and providing as an output the bit stream associated with the identified instance.
10. A transmitter comprising a modulator arranged to implement the steps t1 and U of any of claims 1 to 9.
11. A receiver comprising a demodulator arranged to implement the steps r1, r2 and r3 of claim any of claims 1 to 7.