Transmitter Signal Processor

Abstract

A transmitter signal processor (TSP) for use in combining, amplified, and transmitting multiple independent signals while maintaining an approximately constant envelope, the processor comprising a feedback network whereby an error signal between the sum of the signals and the output of a constant envelope process in the TSP is calculated, and for each input signal, this error signal is correlated with the input signal to model the effect of errors in the receiver, and a feedback signal is generated by multiplying each error signal by its corresponding input signal, summing the result for each such input signal, and then subtracting the feedback signal from the sum input. The constant envelope process may be all digital, such as a normalization process, and / or may comprise part of an analog transmit chain, such as an amplifier, frequency converter, or filter.

Description

[Technical Field]

[0001] The present invention relates to electronic amplifiers, and more particularly to amplifiers used to amplify signals for transmission, where multiple signals need to be transmitted together in an efficient manner. [Background technology]

[0002] Amplifiers are often used in applications where the power efficiency of the amplifier is important. One such application is in satellite systems, where there is often a very limited power budget available to supply all onboard electronic systems. Not only does the power source (which typically comprises solar panels) cost a lot of money, but the increased size and weight can adversely affect satellite performance in many ways, such as launch vehicle size, maneuverability, vulnerability to damage, etc. Furthermore, dissipating the waste heat resulting from amplifier inefficiency is a significant challenge for satellites in space, again resulting in additional cost and weight.

[0003] A Global Navigation Satellite System (GNSS), such as the US Global Positioning System (GPS), the European Galileo system, or the Chinese Beidou system, consists of several satellites (typically 15-30) each in orbit around the Earth. The trend in modern GNSS is to transmit a larger number of signals from a single satellite. These signals may support multiple different GNSS "services," where, for the purposes of this application, each service is associated with one or more signals.

[0004] Each of the individual signals can be a constant envelope signal, but when signals are combined together, it is very difficult to maintain a constant envelope, especially when the combined signal is subsequently filtered. The result can be undesirable intermodulation signals that are a waste of energy, compared to the ideal solution of producing a constant envelope combined signal that can then be fed into a constant envelope amplifier.

[0005] One method of processing multiple signals in a satellite signal transmitter uses a separate transmit amplifier for each signal, with each signal being a constant envelope signal for maximum efficiency, then combines these amplified signals before sending them to the transmit antenna. This therefore results in constant envelope high power amplification. However, combining high power signals for transmission through a single antenna is a technical challenge, and overall this solution has size, cost, and weight disadvantages. Note that a constant envelope signal is one in which the amplitude of the signal is not modulated. This may be synonymously known as a constant amplitude signal.

[0006] A solution to these problems is sought, which takes the non-constant amplitude multiplex of signals and modifies them before they are passed through the amplifier, with the goal of maximizing the fraction of the output power that carries the desired signal, which correspondingly means minimizing the fraction of the output power that goes into unwanted erroneous signals known as intermodulation signals.

[0007] An alternative approach is to use a single power-efficient, constant envelope amplifier. A known technique that operates on this basis is called Phase Optimized Constant Envelope Transmission (POCET) and is described in U.S. Patent No. 8,774,315. This works by using a pre-calculated table of composite signal phase values, values ​​from the table being chosen through an optimization process that minimizes or reduces envelope variations for a phase-modulated carrier, subject to constraints within the signal, for multiple input signals.

[0008] International Publication No. WO 01 / 058026 discloses a technique in which signals from individual sources in a transmission system are combined and amplified in a power amplifier, and simultaneously the signals are combined and subtracted from a portion of the amplified combined signal to provide an error signal that is a measure of the distortion produced by the amplification process. The error signal is then adjusted in phase and amplitude and then combined with the power amplified signal to compensate for the distortion in a feedforward system. [Prior art documents] [Patent documents]

[0009] [Patent Document 1] U.S. Patent No. 8,774,315 [Patent Document 2] International Publication No. 2001 / 058026 [Non-patent literature]

[0010] [Non-Patent Document 1] Allen Katz, John Wood, and Daniel Chokola, "The Evolution of PA Linearization," IEEE Microwave Magazine, February 2016, p. 32 Summary of the Invention [Problem to be solved by the invention]

[0011] It is an object of the present invention to provide an alternative approach to efficient generation of a transmit signal. [Means for solving the problem]

[0012] According to a first aspect of the present invention, there is provided a transmitter having a transmitter signal processor (TSP) for processing a plurality of signals for transmission to generate a constant envelope output signal, the transmitter signal processor having a sum input comprising a sum of the plurality of signals for transmission, further inputs consisting of each signal for transmission, and an output, the TSP being configured to calculate an error signal between the sum input and the output of the TSP, and for each input signal, to model the effect of errors in the receiver by correlating the error signal with the input signal, and to use the sum of the effects for each input signal to generate a correction signal which is fed back in a feedback loop to the input of the TSP.

[0013] Embodiments of the present invention therefore use feedback techniques to improve amplification-induced distortion of signals where a constant envelope is desired.

[0014] Some embodiments of the present invention comprise a transmitter, wherein a feedback loop generates a combined feedback signal that is subtracted from a sum input to generate a "constant envelope input signal", the feedback loop calculating for each input signal: a) a correlator configured to generate a correlation between (i) the input signal and (ii) an error signal comprising a measurement of the difference between the total input and output of a constant envelope processor within the TSP; b) a multiplier configured to multiply the result of the correlation in step (a) with the input signal to generate a multiplier output; wherein the signal processor is further configured to sum the outputs of the multipliers in step (b) for each input signal to generate a combined feedback signal, and to feed the combined feedback signal back to a summation input of the signal processor; The constant envelope process comprises either a normalization function, or an analog signal processing chain, or a digital model thereof.

[0015] Embodiments may provide a transmitter capable of processing multiple (eg, 2, 3, 4, 5, or more input) digital input signals for transmission and generating an output signal for transmission having a substantially constant envelope.

[0016] Those skilled in the art will understand that the correlator in (a) is performing a process similar to that which may occur in a receiver of a signal transmitted by a transmitter. This is evident from the fact that the total input signal in (a) is of a similar form to that of the reference signal used in the correlator in a well-designed receiver, and the transmitted signal is (partially) received at the receiver. The output of the correlator thus provides a representation of the effects of distortion caused by the constant envelope process as seen at the receiver. This correlator output is then multiplied by the input signal in (b) (along with contributions from similar signals generated for each of the other input signals) to form a combined feedback signal, which is subtracted from the total input signal to the TSP. Thus, in effect, the distortion error caused by the constant envelope process is continuously subtracted from the input, which at least partially cancels said distortion.

[0017] In some embodiments, the output signal may be a digital signal, which is then converted to an analog signal in a digital-to-analog converter (DAC), amplified, and up-converted in frequency if necessary before being transmitted.

[0018] Other embodiments can include the outputs of the DAC and amplifier and / or upconverter as part of a feedback loop of a transmitter signal processor (TSP), allowing the TSP to take into account the distortion added by these components, thus leading to an improved, more constant envelope signal transmitted by the transmitter. Alternatively, digital models of one or more components, such as the DAC and upconverter and / or amplifier, can be used in the feedback loop of the TSP rather than the corresponding analog components. Thus, in some embodiments of the present invention, the constant envelope process comprises an analog signal processing chain, or a digital model thereof, having at least one of an amplifier or frequency upconverter.

[0019] In embodiments incorporating analog elements such as a frequency up-converter and / or amplifier, the feedback loop further comprises a digital-to-analog converter along with a frequency down-converter, where the up-converter is in a constant envelope process, allowing analog signals that may be at different frequencies from those of different signals provided to the inputs of the TSP to be at the same frequency as those input signals and more conveniently processed within the feedback loop.

[0020] In some embodiments, processing within the TSP occurs entirely in the digital domain, and the constant envelope process may comprise a normalizer configured to calculate the output o(t)=βw(t) / |w(t)|, where β is the desired constant amplitude and w(t) is the total input, or a scaled version of it.

[0021] Advantageously, in some embodiments, the correlator comprises a multiplier configured to multiply the complex conjugate of signal (i) by signal (ii) and to accumulate the result in an accumulator.

[0022] Advantageously, a high-pass filter may be included in the correlator between the multiplier and the accumulator to remove low frequency components from the output of the multiplier, preventing such low frequency components (including signals at or near DC) from accumulating and causing errors.

[0023] Advantageously, the transmitter may further include scaling means for scaling the input signals before they are summed at the sum input. The scaling means may comprise one or more multipliers. The multiplier(s) may preferably be implemented digitally. The multipliers may be configured to scale each input signal independently. The scaling provided by the scaling means may be adjustable in operation of the TSP. The adjustment to the scaling may be determined by measuring the DC component removed by the high-pass filter.

[0024] According to a second aspect of the present invention, there is provided a method of combining a plurality of digital input signals for transmission by a transmitter, the method operative to generate a constant envelope output signal, the transmitter having a transmitter signal processor (TSP) configured to sum the digital inputs to generate a sum input and to calculate an error signal between the sum input and the output of the TSP, and for each digital input signal, to model the effect of errors that will occur in the receiver by correlating the error signal with the digital input signal, and to generate a correction signal using the sum of the modeled effects for each digital input signal and to feed back the correction signal in a feedback loop, and to subtract the feedback signal from the sum input of the TSP.

[0025] Advantageously, in some embodiments, the method comprises: feeding the sum signal into a constant envelope processor in the TSP, the constant envelope processor comprising at least a normalization function or an analog signal processing chain, or a digital model thereof, and generating an output; The feedback signal is: i) measuring the instantaneous error signal by taking the difference between the total input and the output; ii) for each input signal, modeling the effect of the instantaneous error signal by correlating the instantaneous error signal with the input signal and multiplying the output of the correlation with the input signal; iii) generated by summing the result of the multiplication in step (ii) generated for each input signal with the result of this summation including the feedback signal; The constant envelope process further includes methods that include either a normalization function or an analog signal processing chain.

[0026] Advantageously, the constant envelope processor may be configured to normalize the constant envelope input signal using a normalizer.

[0027] In some embodiments, the constant envelope process comprises a digital-to-analog converter and at least one of a frequency up-converter and an RF amplifier, and for purposes of (i) of this second aspect, the output is derived from a down-converted and digitized version of the output from the RF amplifier and / or up-converter.

[0028] In some embodiments, the constant envelope process comprises a digital model of at least part of an analog processing chain, including at least a power amplifier, of the transmitter in which the method is implemented. Advantageously, the digital model may have inputs from the analog processing chain and may be adapted to modify parameters of the digital model based on these inputs.

[0029] Embodiments of the invention will now be described in more detail, by way of example only, and with reference to the following figures: [Brief explanation of the drawings]

[0030] [Figure 1] FIG. 1 illustrates a high-level representation of a transmitter configured to transmit multiple independent signals through a common upconversion and amplification process. [Figure 2] FIG. 1 illustrates a high-level representation of a prior art process for achieving a constant envelope output signal. [Figure 3] FIG. 1 illustrates a high-level representation of an improved prior art process for achieving a constant envelope output signal. [Figure 4] FIG. 1 illustrates a high-level architecture of one embodiment of the present invention. [Figure 5] FIG. 1 illustrates a more detailed architecture of one embodiment of the present invention, which can operate entirely in the digital domain. [Figure 6] FIG. 1 shows partial details of an alternative embodiment of the present invention in which some elements are operable in the analog domain. [Figure 7] FIG. 10 shows partial details of an alternative embodiment of the present invention in which digital models of analog components are used. DETAILED DESCRIPTION OF THE INVENTION

[0031] 1 shows a simple transmitter architecture 100 in which three independent digital input signals s1(t)-s3(t) are scaled in multipliers α1-α3 and then summed in summer 102 to generate a combined digital signal w(t). This is then converted to analog form in digital-to-analog converter (DAC) 104, before being upconverted by mixing with a local oscillator signal (not shown) in mixer 106 and amplified in amplifier 108. This is then fed into an antenna (not shown) for transmission. In general, of course, in a given system, different numbers of input signals s i (t) can exist. A typical application may involve GNSS satellites, and signals s i (t) may include different navigation signals, such as open navigation signals, as well as commercial, public regulated service, and safety of life navigation signals, and pilot signals, which may be at different frequencies, bandwidths, or phases from one another.

[0032] The signal w(t) is the signal desired for transmission. However, since each signal itself varies, for example in amplitude or phase, the signal s i The envelope of w(t) can vary due to the independence of (t). Variations in the envelope of w(t) can lead to inefficiencies, particularly in RF amplifiers, such as the generation of intermodulation signals as mentioned earlier, which can be costly in terms of additional cooling requirements or excessive power input needed to achieve a given performance when the amplifier is in a challenging environment such as a satellite.

[0033] FIG. 2 shows a simple prior art technique 200 for generating a constant envelope signal. It is similar to the architecture of FIG. 1, but with the addition of a new block in the signal path between summer 102 and DAC 104. The new block comprises a scaling means χ (which can be ignored for the purposes of this paragraph) and a "constant envelope" (CE) block that takes a signal w(t) and produces an output o(t)=βw(t) / |w(t)| of constant envelope (or more constant than its input). It therefore operates to normalize the input signal to an amplitude β. While all known techniques for generating constant envelopes distort the output signal to some degree, this simple approach performs less favorably than other techniques, including embodiments of the present invention.

[0034] A measurement of the error produced by this technique can be made by subtracting the output signal of the CE block from the input signal in subtractor 202 to produce the instantaneous error signal e(t). When implementing the technique as shown in FIG. 2, the power of the error e(t) is determined by the root mean square of w(t) being β 2 It is generally minimized by design by adjusting the scaling factor χ so that it is equal to the root mean square of o(t), where χ 2 {E[|α1| 2 |s1(t)| 2 ]+E[|α2| 2 |s2(t)| 2 +E[|α3| 2 |s3(t)| 2 ]]}=β 2

[0035] Alternatively, and equivalently, the three scale factors α, α, and α can be adjusted by a common scale factor of χ. We assume that the scale factors are fixed and this is all set up once and for all. It should be noted that known implementations of any system of the type shown in FIG. 2 do not actually use the error signal e(t) in any real-time correction process, but rather merely indicate here where the error is measured.

[0036] This simple prior art approach has two limitations: There is no control over the power density spectrum of the noise added by the operation of the CE block; and There is no control over the actual signal power seen by the receiver matched to each individual signal.

[0037] An improved prior art technique 300 is shown in Figure 3. This is the technique used by prior art document U.S. Patent No. 8,774,315, which, as mentioned above, uses pre-calculated tables to generate output signals based on individual input signals, which are conveyed to the CE component by dashed connections 302. These tables take possible combinations of input signals, such as a discrete number of different phases for each (for phase-modulated signals), and pre-calculate the ideal output signal for each possible input combination.

[0038] 4 shows, at a top level, a system 400 according to one embodiment of the present invention. The underlying principle of this embodiment, and all embodiments of the present invention, is the use of negative feedback of the error signal e(t) to correct for errors introduced by constant envelope processes. The error signal e(t) is equivalent to the error signal shown in FIG. 2, but in this embodiment the error signal is used to generate a feedback signal, as described below.

[0039] Signals s1(t)-s3(t) are input signals that need to be combined and transmitted from the power amplifier. They are scaled by corresponding scaling factors α1-α3 and summed in summer 402 to provide a summed output. Further combined scaling can be performed on the summed output in 404, or alternatively, this scaling can be done by incorporating an appropriate common factor into the individual scaling factors α1-α3. This output is signal w1(t), which provides the first input to summing node 406 to calculate error signal e0(t). In this way, the instantaneous error caused by the transformation to constant envelope (CE) is calculated.

[0040] Signal w1(t) is also provided to a further summing node 408 in the transmit path, where a feedback signal is subtracted from it to produce signal w2(t). This signal is then fed into a CE process to produce output o(t), which in this embodiment is given by o(t) = w2(t) / |w2(t)|. Signal o(t) is fed into summing node 406 as its second input, where it is subtracted from signal w1(t) to produce e0(t). Signal o(t) is a constant envelope (digital) signal in this embodiment, which is then converted to an analog signal, upconverted in frequency, and amplified in a power amplifier as needed for transmission.

[0041] The feedback signal is generated by taking the error signal and, for each individual input signal si(t), applying it to a correlator 410, which computes it as s i(t) to model the effect that errors may have on the correlation process that may occur within the receiver, which of course (GNSS and other Direct Sequence Spread Spectrum applications) has its own locally stored reference copy of signal s i (t). Figure 4 only illustrates the correlation and feedback process associated with signal s i (t), and as will be explained later, in practice each signal s i Note that (t) now has its own correlator to provide its own contribution to the feedback signal. The correlator output is then multiplied by a multiplier 412 to produce a signal s i (t) is multiplied by s for the feedback correction signal i (t)'. The corrections themselves are distorted by the CE process due to their subtraction from the signal w1(t), but the continuous negative feedback still reduces the cumulative error effect.

[0042] Figure 5 shows the generation of the feedback signal in more detail. The embodiment 500 shown is the same as that of Figure 4, but the additional elements shown make the operation of the system clearer. Like reference numbers indicate like functional blocks. As explained earlier, the instantaneous transmitter error e0(t) is calculated and correlated with each reference signal. This is shown in the figure only for signal s1(t); processing for other signals uses replicated blocks. Arrow 502 indicates, for example, the transmission of other signals s i 4. The correlator 410 includes a conjugator 506 for conjugating the reference signal before it is multiplied by the error signal in multiplier 508. It is then accumulated in accumulator 510 to generate the error signal e1(t).

[0043] An ideal system would add to the output signal o(t) a correction for the next sample that, when demodulated by the associated receiver correlator, would be equal to the negative of the accumulated error on that signal. This would cause the accumulated error on that signal to go to zero after the next sample. This is the purpose of multiplier 412, which multiplies the accumulated error from accumulator 510 by a reference signal s1(t), and then adds the result in adder 504 to the other input signal s1(t). i (t) and then subtracting the result of the summation from the desired signal w1(t) in adder 406 to yield a modified desired signal w2(t).

[0044] The distorting action of the CE block means that the correction is also distorted, and therefore the level of correction actually achieved for the next sample is reduced, but a significant degree of correction survives the CE block, and better modeled performance than prior art techniques is achieved.

[0045] A detail to note is that the scale factor χ at 404 (or equivalent adjustments made to α1-α3) will not be optimally adjusted for the signal components s i The output amplitude that can be actually achieved for (χ α i s i ) can never be equal to α . Therefore, the error associated with that signal component can grow without bound. This problem is solved by including a high-pass filter at the point indicated by the star 0 in the correlator 410. By measuring the DC component removed by the high-pass filter, the source signal scale factor (α i ) can be adjusted by a slow fitting algorithm (for example, the LMS steepest gradient algorithm).

[0046] The above embodiments operate entirely in the digital domain, in that all processing and CE processes operate in the digital domain and provide their output to a DAC, which then feeds into any necessary upconversion and amplification for transmission. Other embodiments operate at least partially in the analog domain and use a CE process with a feedback path that comes from elements in the analog path, usually after the amplification stage. This allows imperfections in the amplifier to be directly taken into account by the processing that occurs in the feedback path. These imperfections may include any soft clipping that occurs in the amplifier and / or any AM-to-AM and AM-to-PM distortion that may occur, where AM is amplitude modulation and PM is phase modulation.

[0047] 6 shows a top-level portion 600 of an embodiment of the present invention having a CE process in analog form, which comprises part of the transmit chain of a transmitter. Note that the details of the feedback process are the same as for the above embodiment and will not be described in further detail in connection with this figure.

[0048] Therefore, this diagram only shows the analog equivalent of the (digital) CE block of Figure 5, together with the error summer 406. The connections from w1(t) and o(t) to summer 406, together with the output from summer 406, show the paths to the feedback circuit, which may comprise, for example, the feedback circuit described in relation to Figure 4.

[0049] Here, signal w(t) is passed to a DAC, and the resulting analog signal is upconverted to the transmit frequency in upconversion mixer 604 and then passed to RF amplifier 606 for amplification to the desired transmit output power to generate analog power signal 608 for subsequent transmission via the antenna. This is the step where the method of the present invention corrects most of the distortion. Coupler 610 takes a small amount of this transmit energy from the output of the RF amplifier and converts it back to a baseband signal in downconversion mixer 612, which then converts this back to a digital signal in analog-to-digital converter 614. This is then fed into a feedback loop to generate error signal e(t), as in the previous embodiment.

[0050] Unlike the all-digital approach of the previous embodiment, this embodiment may introduce a small delay in the signal path. In GNSS applications, the signal s i (t) comprises a sequence of chips, and provided that any added delay is small compared to one chip duration (which is approximately 0.1 μs at higher GNSS chip rates), this delay can be ignored without causing any significant problems. However, compensating for such a delay is not difficult, especially if it is roughly constant. To do so, the delay is measured at design time and during the test phase of system manufacturing. That amount of delay is then inserted into the digital correction circuitry in conjugate box 506 of FIG. 5 (this is shown in an all-digital scenario, but the feedback signal generation will be the same in this embodiment, as mentioned earlier).

[0051] A further potential problem with the embodiment of FIG. 6 is that splitting the signal from the amplifier output (at coupler 610) in the RF measurement process, downconverting the signal in frequency, and converting it to digital form can introduce noise. FIG. 7 shows an alternative embodiment aimed at reducing the effects of any such noise that may be present. This embodiment shares similarities with that of FIG. 6 in that the output of the analog transmit chain is used to influence the feedback process. The analog transmit chain, coupler, downconverter, and digitizer are similar to those shown in FIG. 6 and therefore have the same reference numbers. However, in this embodiment, the output of digitizer 614 is not used directly but instead feeds into digital amplifier model 616. This digital amplifier model is a model of the RF amplifier. Because RF amplifier characteristics do not change very rapidly, it has adjustable parameters that are slowly adapted over a "relatively long" period based on the input of the downconverted and digitized signal from the actual power amplifier. This period may be measured, for example, on the order of tens or hundreds of milliseconds, or even seconds. Their adaptation is controlled by feedback of the error, which is the difference between the (downconverted and digitized) amplifier output p(t) and the model output o(t).

[0052] The adjustable parameters can take the form of a mapping function from input envelope to output envelope and a mapping function from input envelope to output phase error. Alternatively, they can be parameters of a Volterra function model of the power amplifier. Such a model is described in "The Evolution of PA Linearization," by Allen Katz, John Wood, and Daniel Chokola, IEEE Microwave Magazine, February 2016, p. 32.

[0053] This embodiment offers the advantage that noise or errors in the combined output o(t) are not directly fed into the feedback process described above, but are smoothed out by long-term adaptation in the model, which itself is digital and operates with low output noise.

[0054] Those skilled in the art will appreciate that the novel techniques described herein are quite different from existing methods because they use pre-calculated tables whereas the new method uses real-time feedback of errors.

[0055] Embodiments of the present invention will typically be implemented in software, e.g., one or more digital signal processors or microprocessors, which are programmable hardware in which one or more processors are controlled by computer code, or may be operated in firmware / hardware, such as in one or more application-specific integrated circuits (ASICs) or field-programmable gate arrays (FPGAs). Such devices will typically include or be connected to appropriate memory and storage, as will be understood by those skilled in the art. The present invention may therefore extend to a software program storably disposed within a computer and consisting of instructions that cause a processor to perform various elements described herein. Some embodiments may be implemented in a combination of hardware or software, and the transmitter signal processor may comprise a combination of analog and digital circuitry configured to process signals in both the analog and digital domains. In particular, the analog circuitry forming part of the transmitter signal processor may consist of one or more amplifiers (including power amplifiers), frequency shifters, filters, etc., that form part of the signal processing chain for signals to be transmitted by the antenna. Embodiments of the present invention have utility in many areas of signal transmission, typically in space or airborne applications, such as GNSS satellites, and may also be used more widely for terrestrial radio receivers, where increased power efficiency results in longer battery life and reduced waste heat.

Claims

1. 1. A transmitter having a transmitter signal processor (TSP) for processing a plurality of signals for transmission to generate a constant envelope output signal, the transmitter signal processor having a sum input including a sum of the plurality of signals for transmission, further inputs consisting of each signal for transmission, and an output, the TSP being configured to calculate an error signal between the sum input and the output of the TSP, and for each input signal, to model the effect of errors in the receiver by correlating the error signal with the input signal, and to use the sum of the effects for each input signal to generate a correction signal that is fed back in a feedback loop to the input of the TSP.

2. A feedback loop generates a combined feedback signal that is subtracted from the sum input to generate a "constant envelope input signal," and for each input signal, the feedback loop: a) a correlator configured to generate a correlation between (i) the input signal and (ii) an error signal comprising a measure of the difference between the total input and output of a constant envelope processor within the TSP; b) a multiplier configured to multiply the result of the correlation in step (a) with said input signal to generate a multiplier output; wherein the signal processor is further configured to sum the outputs of the multipliers in step (b) for each input signal to generate a combined feedback signal, and to feed the combined feedback signal back to a summation input of the signal processor; The transmitter of claim 1 , wherein the constant envelope processor comprises either a normalization function, or an analog signal processing chain, or a digital model thereof.

3. 3. The transmitter of claim 2, wherein the constant envelope processor consists of, or comprises a digital model of, an analog signal processing chain having at least one of an amplifier or a frequency up-converter.

4. 4. The transmitter of claim 3, wherein the feedback loop further comprises a digital-to-analog converter in conjunction with a frequency down-converter, and the up-converter is in a constant envelope process.

5. 3. The transmitter of claim 1 or 2, wherein processing in the signal processor occurs entirely in the digital domain, and the constant envelope process consists of a normalizer configured to calculate an output o(t) = βw(t) / |w(t)|, where β is a desired constant amplitude and w(t) is the total input or a scaled version thereof.

6. 6. The transmitter of claim 1, wherein the correlator comprises a multiplier configured to multiply the complex conjugate of signal (i) by signal (ii) and to accumulate the results in an accumulator.

7. 7. The transmitter of claim 6, wherein a high-pass filter is included in the correlator between the multiplier and the accumulator to remove low frequency components from the output of the multiplier.

8. 8. A transmitter according to any preceding claim, wherein the transmitter further comprises scaling means for scaling the input signals before they are summed at the summation input.

9. 1. A method of combining multiple digital input signals for transmission by a transmitter, the method operating to generate a constant envelope output signal, the transmitter having a transmitter signal processor (TSP) configured to sum the digital inputs to generate a sum input, calculate an error signal between the sum input and the output of the TSP, and for each digital input signal, model the effect of errors that will occur in a receiver by correlating the error signal with the digital input signal, and generate a correction signal using the sum of the modeled effects for each digital input signal, and feed back the correction signal in a feedback loop, and subtract the feedback signal from the sum input of the TSP.

10. The summed input signal is fed into a constant envelope processor in the TSP, which includes at least a normalization function or an analog signal processing chain, or a digital model thereof, and generates an output; The feedback signal is i) measuring the instantaneous error signal by taking the difference between the total input and the output; ii) for each input signal, modeling the effect of the instantaneous error signal by correlating the instantaneous error signal with the input signal and multiplying the output of the correlation with the input signal; iii) summing the result of the multiplication in step (ii) generated for each input signal with the result of this summation including the feedback signal. The method of claim 9 , wherein the compound is produced by

11. 11. The method of claim 10, wherein the constant envelope processor comprises a normalizer for normalizing the constant envelope input signal.

12. 12. A method according to claim 10 or 11, wherein the constant envelope process further comprises a digital to analogue converter and at least one of a frequency up-converter and an RF amplifier, and for the purposes of claim 9(i), the output is derived from a down-converted, digitised version of the output from the RF amplifier and / or up-converter.

13. 12. The method of claim 10 or 11, wherein the constant envelope process further comprises a digital model of at least part of an analog processing chain, including at least a power amplifier, of a transmitter in which the method is implemented.

Images