Response linearizing circuit, particularly for microwave radiofrequency systems, and method of linearizing the output of an apparatus

[0002] 1. Field of the Invention

[0003] The present invention relates to the linearization of non linear networks and in particular concerns a fully adaptive linearizing circuit not using predistortion systems, particularly for microwave radiofrequency systems, and which is independent of the type of modulation. The present invention also concerns a method for linearizing the output of an apparatus, for instance a power amplifier.

[0004] 2. Description of the Prior Art

[0005] The nonlinearity phenomena affect many fields of the electronic engineering. An important aspect, which is substantially the subject-matter of the present invention, is the one related to the nonlinearity of the amplifiers used in the electronic systems applied to telecommunication and in particular, others being not excluded, to digital signal transmission.

[0006] Present radio transmission systems, using complex modulations such as the multilevel ones, require more and more complex transceivers. In particular, in current digital signal transmissions, signals with high peak to average power ratios must be transmitted. In order to obtain a good quality of transmission by using these types of signals, it is necessary to keep the distortion value of the processed signal below a prefixed level.

[0007] This fact often forces the designers to use, for instance in transmission systems, class-A power amplifiers with high Back-Off values, hence with a low energy efficiency.

[0008] The main known linearizing techniques can be classified into four main different groups or methods: a) Feed-forward; b) Feed-back; c) Baseband numeric predistortion; and d) IF or RF analog predistortion. Each specific solution, excluding the Feed-back one which is adaptive by its very nature, can in turn be implemented according to an adaptive or non-adaptive technique.

[0009] In the adaptive technique, the linearizing apparatus observes the output produced by the power amplifier and, in dependence on the detected values, updates its internal parameters, so it minimizes the magnitude of the produced distortions. It is easy to understand how this is the ideal solution from the conceptual viewpoint, because it adapts through a suitable algorithm, the linearizing function to the actual behaviour of the system; in practice, the realization is unfortunately complex and, above all, expensive.

[0010] In the non-adaptive technique, on the contrary, the linearizing device is adjusted once and for all during the calibration, so as to assure the most satisfactory possible mean behaviour. The system is simpler, but with lower performances, since there is not feed back for controlling the actual status of the distortions coming out of the transmitter.

[0011] U.S. Pat. No. 5,742,201 discloses a predistortion linearizing device for an RF or microwave power amplifier. The circuit couples the input of an amplitude and phase controller through a feedback signal path to receive a portion of the amplified output signal. A second input receives a delayed portion of the input over a second signal path. The gain and phase regulations performed at the input to be amplified are controlled by a controller. However such known device has the following main feedbacks:

[0012] the system does not act instantaneously but with a delay due to outside controllers, and

[0013] the feedback loop is based upon the estimation of two or more signal parameters and not upon a sole signal.

[0014] It is, therefore, the primary object of the present invention to provide a fully adaptive linearizing circuit which does not use predistortion systems, particularly for microwave radiofrequency systems, which is independent of the type of modulation.

[0015] A further object of the present invention is to provide a fully analog device of a Feed-back type that is utilizable both at intermediate frequency IF and at radio frequency RF.

[0016] Still a further object of the present invention is to provide a linearizing device that does not lead to a decrease in the gain of the amplifier chain.

[0017] These and further objects are achieved by a linearizing circuit according to the invention having the features set forth in the independent claim 1 and by a method having the features set forth in claim 8. Further advantageous characteristics of the present invention are set forth in the dependent claims.

[0018] The basic idea of the present invention consists in picking-up only the distorted portion of the output signal and feeding it back to the amplifier input with the correct amplitude and phase levels in order to completely cancel the output portion of the signal distortions.

[0019] The invention will certainly result in being clear from the following a detailed description, given by way of a mere non limiting and exemplifying example, to be read with reference to the annexed illustrative sheets of drawings.

[0020] The various figures show:

[0021] FIG. 1 shows the basic diagram of the linearizing circuit according to the present invention; and

[0022] FIG. 2 shows a possible different embodiment of the circuit according to the present invention.

[0023] Like reference characters will be used throughout the various figures to designate the same components or functionally equivalent parts.

[0024] The present invention falls under the fully analog predistortion category and it is utilizable both at intermediate frequency IF and at radio frequency RF. Moreover, it is structured to be able to operate adaptively, being of the Feed-back type.

[0025] With reference to FIG. 1, the circuit CI, according to the invention, for linearizing the output OU of an apparatus TR, for instance an amplifier/transmitter TR, comprises:

[0026] a divider DI_IN of the input signal IN;

[0027] a signal adder circuit AD2 at the input (i.e. upstream) of the apparatus TR;

[0028] an output signal OU divider (coupler) DI_OU;

[0029] a feedback loop FL with an amplitude control circuit AC;

[0030] a signal loop SL; and

[0031] a closure loop CL.

[0032] The signal loop SL comprises an amplifier AM, a delay line DL and a phase controller PC. The closure loop CL comprises a first adder circuit AD1, an amplifier AM, a loop filter LF and a phase controller PC.

[0033] The operating principles of the linearizing circuit in accordance with the present invention will be described hereinafter.

[0034] At divider DI_IN, the input signal IN is divided into two parts, IN1, IN2, that may also be unequal. The signal portion IN1 proceeds towards the amplifier TR passing through the second signal adder circuit AD2. Downstream of the amplifier TR to be linearized, a portion of the signal is picked up by means of the output signal OU divider DI_OU. This portion of signal after a suitable level adjustment through circuit AC, is fed to one of the two inputs of the first adder circuit AD1 of the closure loop.

[0035] The signal portion IN2 coming out of the divider DI_IN, through the signal loop SL, enters the second input of the second adder circuit AD2. While passing through the loop SL, the signal portion IN2 is amplitude, delay and phase-regulated by means of the amplifier AM, the delay line DL and the phases controller PC, respectively. Such amplitude, delay and phase regulations result in that:

[0036] the signal at the two inputs of the adder AD1 has the same amplitude;

[0037] the signal at the two inputs of the adder AD1 has a 180-degree phase difference; and

[0038] the signal at the two inputs of the adder AD1 has the same slope of the group delay characteristic.

[0039] In this way, at the output of the adder AD1 we obtain the difference between the output signal sample and the input signal sample. This difference is nothing but the distortion produced by the amplifier (or in general by the apparatus) TR to be linearized.

[0040] At this point, the feedback loop is closed through the adder AD1 of the loop CL for the distorted signal alone. By adjusting the amplitude on the closing loop CL, the proper loop gain is determined and, by adjusting the phase (PC), the proper negative feedback is obtained. The loop filter LF serves to stabilize the feedback loop FL.

[0041] It will be easily understood that the circuit according to the present invention is different from the conventional feedback circuits since the gain of the amplifier to be linearized is unchanged. This in turn results in an evident and unquestionable economical advantage.

[0042] Therefore, calling G_Lthe gain of the loop and G_Athe gain of the amplifier, it follows that the input signal is output multiplied by G_A, instead of being reduced to G_A/|1−G_L|, but with a distortion level decreased by a factor 1/|1−G_L|.

[0043] Implementatively, the signal band, the group delay slope and the loop gain are connected each other by the need to have a stable system. In order for the system to be suitable, because of the feedback loop FL, it is necessary to restrict the signal band so as to ensure sufficient gain and phase margins, in accordance with the theory. In fact, given an amplifier to be linearized with a given group delay and defined an extent of distortion reduction equal to about the reciprocal of the loop gain (1/G_L), a filter is thus defined that ensures the stability and consequently the utilizable band (to be noted that the useful band is only 33% of the filter band, the two distortion lobes having to be taken into account). Therefore, there is a compromise between the improvement to be achieved and the useful band that depends mainly upon the delay of the amplifier to be linearized.

[0044] FIG. 2 shows a possible different circuit embodiment according to the present invention. In practice it is shown how, by acting at an intermediate frequency, a diagram substantially equivalent to the one of FIG. 1 can be used. As can be seen, the differences concern the addition of a local oscillator LO which performs a frequency shift in the feedback loop FL and right before the transmitter TR.

[0045] The method for linearizing the output of an apparatus TR comprises the steps of subdividing the input signal IN into at least a first (IN1) and a second (IN2) portions and picking up a portion of the signal OU coming out of the apparatus TR to be linearized and it is characterized in that it further comprises the steps of summing the picked up portion of the output signal with the second portion IN2 of the input signal IN thus outputting a signal; and summing said output signal with the first portion IN1 of the input signal upstream of the apparatus to be linerized.

[0046] It will now be apparent that the advantageous characteristics, compared with similar systems such that described in the above cited U.S. Pat. No. 5,742,201, are:

[0047] 1. The two controllers are not necessary;

[0048] 2. The device according to the present invention is a pure analog system and, as such, it acts quickly with a delay only due to the circuit topology and not to outside controllers; and

[0049] 3. The feedback loop operates with a single signal and not with the estimate of two or more parameters thereof.

[0050] The invention further includes a computer program comprising computer program code means adapted to perform all the steps of the described method when said program is run on a computer. Furthermore, it includes a computer readable medium having a program recorded thereon, said computer readable medium comprising computer program code means adapted to perform all the steps of the described method when said program is run on a computer.

[0051] It is apparent that several modifications, adaptations, variations and replacement of parts by other which are functionally equivalent can be carried out to the embodiments illustrated and described in detail above without departing from the scope defined by the following claims which are intended to be an integral part of the present description.