TECHNICAL FIELD

The present invention discloses a receiver device with adaptive IQ correction.

BACKGROUND

A well known problem in receiver technology is to obtain I and Q signals which are perfectly matched in phase, i.e. have exactly 90 degrees difference, and which have identical amplitudes. The problem is often caused by imperfections in components, mainly in analogue components.

Since the problem is well known, there are also many solutions which have been proposed to solve the problem. Known proposed solutions include solutions which involve a priori information about the signal, or solution which require calibration of the receiver at startup. In addition, so called "blind compensation" techniques have also been proposed, i.e. techniques where no prior knowledge about the signal is available.

Drawbacks to these previous solutions to "the l-Q problem" include the need for a pilot tone if calibration is used or the need for a known sequence of data if an "a priori" solution is chosen, which may lead to delays or a lower throughput due to the need to schedule the known sequence in addition to the "payload". Regarding the use of blind compensation techniques, such techniques have generally turned out to have poor performance.

SUMMARY It is an object of the invention to provide a receiver with improved performance regarding l-Q performance. This object is addressed by means of a receiver device arranged to receive an input signal. The receiver device comprises a first and a second receiver, and the first and second receivers are arranged to receive and demodulate respective first and second parts of the input signal into respective output baseband signals which comprise I and Q components. The receiver device also comprises an adaptive IQ correction component arranged to receive the output baseband signals from the first and second receivers and to correct IQ errors in the output baseband signal from the first receiver by means of the received output baseband signals. In the receiver device, the first and second receivers are arranged to have different transfer functions within one and the same frequency range and the corrected output signal from the first receiver is arranged to be the output signal from the receiver device. In embodiments of the receiver device, the adaptive IQ correction component comprises an adaptive filter which is arranged to correct IQ errors in the output baseband signal from the first receiver by means of coefficients which are adaptively adjusted. In embodiments of the receiver device, the adaptive filter is an adaptive Least Mean Square filter.

In embodiments of the receiver device, the adaptive IQ correction component is arranged to adaptively determine said coefficients by means of the output baseband signal from the first receiver, the output baseband signal from the second receiver and the output signal from the adaptive filter.

In embodiments of the receiver device, the adaptive filter comprises four filters, two for each of the I and Q components of the input signal to the adaptive filter, one of which is used directly and the other is used as a cross connection.

In embodiments of the receiver device, the first and the second receivers have different transfer functions within one and the same frequency range by virtue of the first receiver being a homodyne receiver and the second receiver being a non-homodyne receiver.

In embodiments of the receiver device, the second receiver is a heterodyne receiver, suitably a so called superheterodyne receiver.

In embodiments of the receiver device, the first and the second receivers are arranged to receive said respective first and second parts of the received signal by means of the receiver device comprising a splitter connected to the inputs of the first and second receivers.

In embodiments of the receiver device, the first and the second receivers are arranged to receive said respective first and second parts of the received signal by means of the receiver device comprising a coupler arranged to couple a part of the received signal to the second receiver.

BRIEF DESCRIPTION OF THE DRAWINGS

The invention will be described in more detail in the following, with reference to the appended drawings, in which Fig 1 shows an overview of a receiver device with adaptive IQ correction, and

Figs 2 and 3 show more detailed views of a filter used in the receiver device of fig 1.

DETAILED DESCRIPTION

Embodiments of the present invention will be described more fully hereinafter with reference to the accompanying drawings, in which embodiments of the invention are shown. The invention may, however, be embodied in many different forms and should not be construed as being limited to the embodiments set forth herein. Like numbers in the drawings refer to like elements throughout.

The terminology used herein is for the purpose of describing particular embodiments only, and is not intended to limit the invention.

Fig 1 shows an embodiment of a receiver device 100. As shown, the receiver device 100 comprises a first 105 and a second 1 10 receiver, each of which is arranged to receive and demodulate respective first, A, and second, B, parts of an input signal "RX in" which is received by the receiver device 100, for example from an antenna, suitably through or via a Low Noise Amplifier, an LNA.

The first 105 and second 1 10 receivers are, in this embodiment, arranged to receive their first A and second B parts of the received signal by means of a directional coupler 125 which is comprised in the receiver device 100 before the two receivers 105, 1 10, so that a (suitably minor) part B of the signal "RX in" is diverted to the second receiver 1 10, and the rest, part A, of the received signal goes to the first receiver 105. Another way of providing both receivers 105, 1 10, with their respective parts of the signal "RX in" is to use a splitter instead of the directional coupler 125. In such a case, it should be pointed out that the first A and second B parts of the received signal need not be of equal proportions.

Each receiver 105, 1 10, receives and demodulates its part A, B, of the signal "RX in" to baseband ("BB") level, and outputs a demodulated baseband signal which comprises I- and Q- components. Suitably, each of the receivers 105, 1 10, comprises an analogue to digital converter, an ADC, so that there is within each receiver 105, 1 10, an interface between the analogue and digital "domains", with the final conversion of the signal to a baseband signal being carried out in the digital domain, so that the output baseband signals from the two receivers 105, 1 10, are digital signals. Equally suitably, the input signals A and B to the respective receivers 105, 1 10, should have differing frequency offsets at the input to their respective ADC, which can, for example, be accomplished by means of using differing mixer frequencies at mixers in the receivers, in which mixers the input signals to the respective receivers are "down-mixed" before being input to their respective ADC. "Differing offsets" can of course also be expressed as saying that one of the receivers has an offset relative to the other at the input to its ADC.

As mentioned above, in the receiver device 100, it is the first receiver 105 which is the "main receiver", i.e. the receiver which is to have IQ errors in its output signal corrected, which means that the second receiver 1 10 can be seen as a "reference receiver". For these reasons, the demodulated output IQ signals from the first receiver 105 are shown as Xi(n) and Xcj(n), and the demodulated output IQ signals from the second receiver 1 10 are shown as refi(n) and refo(n). The letter "n" is used to indicate a time sequence. As is also shown in fig 1 , the receiver device 100 comprises an adaptive IQ correction component 120 which is arranged to receive the output I and Q signals from the first receiver 105, and to correct IQ errors (both in phase and in amplitude) in the output baseband signal from the first receiver 105 and to output a corrected baseband signal comprised of I and Q components; this output signal then becomes the output signal from the entire receiver device 100. For this reason, the output signal from the adaptive IQ correction component 120 is indicated as yi(n) and yo(n) in fig 1. In order to achieve the correction of IQ errors in the output signal baseband from the first receiver 105, the adaptive IQ correction component 120 is adaptive, and is arranged to also receive the demodulated baseband output signal from the second receiver 1 10, i.e. the signals shown as refi(n) and refo(n). In order to achieve the correction of IQ errors in the output baseband signal from the first receiver 105, a principle of the invention is that the two receivers 105, 1 10, utilize different receiver principles in their demodulation of their respective parts A and B of the signal "RX in". The use of different receiver principles can also be expressed as saying that the two receivers 105, 1 10, have different transfer functions within one and the same frequency range, where the frequency range can be more or less arbitrarily chosen. The transfer functions should differ at last frequency-wise, i.e. the two receivers 105, 1 10, have different transfer functions within one and the same frequency range In one embodiment, the first receiver 105 is a homodyne receiver, and the second receiver 100 is a non-homodyne receiver, an example of which would be a heterodyne receiver, suitably but not necessarily a so called superheterodyne receiver. In other embodiments, the second receiver 1 10 is a homodyne receiver with a frequency offset of the signal "RX in". Turning now to a more detailed description of the adaptive IQ correction component 120, in one embodiment the adaptive IQ correction component 120 comprises an adaptive filter, which in one embodiment is an adaptive so called Least Mean Square filter, i.e. an adaptive LMS filter, shown as 130 in fig 2. The I and Q output signal of the adaptive LMS filter 130 is used as the output I and Q signals from the entire adaptive IQ correction component 120, as indicated in fig 2, by means of the output signals from the adaptive LMS filter 130 being shown in fig 2 as yo(n) and yi(n). The adaptive LMS filter 130 is embodied by applying a direct FIR filter and a cross branch FIR filter to each of the I and Q signals from the first receiver 105, i.e. to the signals xi and XQ. Thus, there are four adaptive FIR filters in the LMS filter 130, as follows: the direct FIR filters are shown as 131 (Q- branch) and 134 (l-branch) in fig 2, and the cross branch FIR filters are shown as 132 (Q-branch) and 133 (l-branch) in fig 2.

All of the four FIR filters are adaptive, i.e. comprised of coefficients which are adaptively updated in a manner which will be described in more detail below. If we denote each such FIR filter W with the corresponding sub-scripts, we thus get four combinations, the direct filters 134 ("W _N ") and 131 ("W _QQ ") and the cross branch filters 133 ("W _Q i") and 132 ("Wiq"), as is also shown in fig 2. If we let each of the four FIR filters have M "taps", we can write the function of the filters 131 -134: WQQ = [W _QQ (1 ), W _QQ (2) ... WQQ(M)] (1 )

W| _Q = [W|Q(1 ), W|Q(2) ... WIQ(M)] (2)

WQ, = [WQI(1 ), WQ,(2) ... WQI(M)] (3)

W„ = [W„(1 ), W„(2) ... Wii(M)] (4) Since the output from the adaptive LMS filter 130 is used as the output from the adaptive IQ correction component 120, at time n denoted as yi(n) and y _Q (n), the LMS filter 130 gives us:

y _Q («) = ½ («) - [ _/ («) ^■ x _I (n - m) + WQ _Q (n) - x _Q (n - m)]

Thus, the output signals yi and y _Q from the LMS filter 130 are used as the corrected output signals (I and Q) from the entire receiver device 100. In addition, the signals y and y _Q are also used in order to update the filter coefficients of equations (1 )-(4) above, i.e. the coefficients in the four vectors of M length, or M "taps".

The updating of the filter coefficients is shown in fig 3: as we see, the output signals yi and VQ are used together with the I and Q output signals refi and refo from the reference receiver, i.e. the second receiver 1 10, in order to calculate I and Q "error signals" ecj and ei at time n, as follows:

e _[ (n) = y _[ (n) - ref _[ (n)

The error signals ecj(n) and ei(n) are used together with the signals xi (n) and yo(n) to update the coefficients in the FIR filters 131 -134, i.e. to determine the coefficients in the FI R filters 131 -134 for time (n+1 ). The updating of the coefficients is carried out in an updating block 135 which is shown in fig 3, and which updates the coefficients as follows: H¾(« + 1) = H¾(«) + («)¾ («) (5) w™ (« + 1) = w™ («) + μ(ή)β _ΰ (ή)χ™(η) (6)

+ !) = + ( ⁿ ) (7) w _u ^m {n + 1) = vt£(n) + μ(η)β ₇ (η (η) (8)

In equations 5-8 above, μ(η) is the "update step size", i.e. how large a "step" to take in each update of the coefficients above and m e [0.. - 1] . As shown in fig 3, the updating block 135 is connected to the adaptive filter 130 in a feedback loop, in order to provide the adaptive filter 130 with the updated coefficients. It can also be mentioned that the update step size μ(η) can be time dependent, or, in other words, can be made to vary over time.

In the drawings and specification, there have been disclosed exemplary embodiments of the invention. However, many variations and modifications can be made to these embodiments without substantially departing from the principles of the present invention. Accordingly, although specific terms are employed, they are used in a generic and descriptive sense only and not for purposes of limitation. The invention is not limited to the examples of embodiments described above and shown in the drawings, but may be freely varied within the scope of the appended claims. For example, in embodiments with the adaptive filter, the filter can be a Kalman filter or a so called Recursive Least Square filter, an "RLS" filter.