US20040137868A1

US20040137868A1 - Automatic gain control

Info

Publication number: US20040137868A1
Application number: US10/250,473
Authority: US
Inventors: Christian Nystrom; Per Kondradsson; Ari Grasbeck
Original assignee: National Semiconductor Corp
Current assignee: National Semiconductor Corp
Priority date: 2001-02-16
Filing date: 2002-02-13
Publication date: 2004-07-15
Also published as: US7082291B2; SE0100528D0; EP1360759A1; SE521838C2; US20030148745A1; SE0100528L; WO2002067421A1

Abstract

The invention relates to gain control for an amplifier for received radio signals by means of an automatic gain control signal (c). A down converter (103) produces signal components (I; Q) that are amplified in a respective amplifier (104, 105). Then, the resulting signal components (I_P-LP; Q_P-LP) are digitised in A/D-converters (110; 111) for further processing in a digital signal processor (112) and are also fed to a respective signal level detector (113; 114) whose output signals ([I]; [Q]) are combined (115) into a combined signal (V_Σ). This signal (V_Σ) is forwarded to a gain control unit (116), which produces the automatic gain control signal (c). The automatic gain control signal (c) controls the gain of the amplifiers (104; 105) such that the resulting signal components (I_P-LP; Q_P-LP) are maintained at signal level ([I]; [Q]) less than a predetermined limit level being adapted to the A/D-converters (110; 111).

Description

THE BACKGROUND OF THE INVENTION AND PRIOR ART

The present invention relates generally to automatic gain control in a radio frequency receiver. More particularly the invention relates to a method of controlling gain of an amplifier for received radio signals in a radio receiver, according to the preamble of claim 1. The invention also relates to a computer program, for instance in the form of a digital processor algorithm, according to claim 12, a computer readable medium according to claim 13 and an arrangement according to the preamble of claim 14.

A problem that arises in most radio communications receivers concerns the wide variation in power level of the radio signals received at the antenna. This variation is due to a variety of causes. For example, the distance between the transmitter and the receiver can vary considerably. Different transmitters may also utilise different power levels. Since, disregarding the influence of any screening objects, the received signal power decreases as the square of the distance to the transmitter, wide variations in received power level are likely to arise in many situations. Furthermore, these variations may occur very rapidly due to changes in the radio conditions. Movements of the receiver station and/or transmitter station or repositioning of objects between the stations are typical situations in which the conditions for the radio channel can change dramatically.

In radio design it is therefore common practice to include an automatic gain control (AGC) circuit in the receiver. The AGC circuit utilises feedback to maintain a fixed (or at least as stable as possible) signal power level within the receiver even though the signal level at the antenna varies widely. The AGC is achieved by using an amplifier whose gain can be controlled by an external signal, e.g. a voltage or a current.

In analogue receivers it is known to incorporate AGC circuits that operate on a down converted intermediate frequency (IF) signal, i.e. a signal component, which has been frequency transformed down from a received radio frequency (RF) signal and which is to be further frequency transformed down in a following frequency down conversion step.

Receivers in which the radio signal is digitally processed, in most cases after frequency down conversion, usually perform the AGC operation by digitally assisted processing. Thus, the AGC loop implies both analogue to digital conversion and digital to analogue conversion. For many of today's applications this gives a satisfying compensation for the power level variations in the received radio signals.

However, besides capable A/D- and D/A-converters, digital AGC also requires an amount of processing power, which in turn is correlated with power consumption and costs. For large signal bandwidths this effect becomes especially pronounced.

Moreover, a large bandwidth places relatively demanding requirements on the A/D- and D/A-converters, particularly if a high digital resolution is necessary.

In a radio communication system for so-called bursty communication, for instance in the form of data packets, short pieces of information are passed between transmitter stations and receiver stations at irregular and generally unpredictable time instances. A particular station can, in most such systems, act both as a transmitter station and as a receiver station. With some exceptions, this means that every station in the system is a potential receiver of a radio message at any time. The station must therefore be capable of tuning its AGC circuit to received radio signals very rapidly.

The IEEE: 802.11a, 802.11b and ETSI: Hiperlan/2 constitute specific examples of wireless LAN protocols where extremely quick and accurate AGC-tuning is demanded. (IEEE=The Institute of Electrical and Electronics Engineers, ETSI=The European Telecommunications Standards Institute, LAN=Local Area Network, Hiperlan=High performance radio local area network). 802.11b specifies packet data exchange at speeds up to 11 Mbps/channel (under direct sequence modulation at 2.4 GHz) and Hiperlan/2 makes possible wireless access to the Internet and real time video services at speeds of 54 Mbps (at 5 GHz). In order to meet the hardest requirements of these standards a radio receiver must be capable of calibrating its receiver circuitry to the power level of received radio signals within 10 μs from start of transmission. This means that the AGC function must control the receiver amplifier gain to a suitable level within 10 μs or less.

However, it is both technically complicated and expensive to accomplish an AGC function involving digital signal level detection with sufficient accuracy and within such short time limits by utilising today's A/D converters, D/A converters and digital signal processors.

SUMMARY OF THE INVENTION

It is therefore an object of the present invention to solve the problems above and thus provide an improved solution for controlling the gain of an analogue variable gain amplifier in a radio receiver, which is quick and accurate enough to meet the requirements of the modern packet data protocols.

According to one aspect of the invention the object is achieved by a method for controlling gain of an amplifier for received radio signals in a radio receiver as initially described, which is characterised by varying a time constant of an automatic gain control signal in response to a time derivative parameter of the at least one down converted signal. The time constant is at least varied such that the automatic gain control signal is adapted to a power variation rate of the received radio frequency signals. The time constant may, of course, also be varied on basis of arbitrary additional parameters depending on, for instance, the protocol and frame structure according to which the particular receiver operates.

According to another aspect of the invention these objects are achieved by a computer program directly loadable into the internal memory of a digital computer, comprising software for controlling the method described in the above paragraph when said program is run on a computer.

According to yet another aspect of the invention these objects are achieved by a computer readable medium, having a program recorded thereon, where the program is to make a computer perform the method described in the penultimate paragraph above.

According to an additional aspect of the invention the object is achieved by an arrangement for controlling gain of an amplifier for received radio signals in a radio receiver as initially described, which is characterised in that a gain control signal generator is adapted for varying a time constant of the automatic gain control signal in response to a time derivative parameter of at least one down converted signal. The variation of the time constant is thus adjusted to a power variation rate of the received radio frequency signals.

The invention thereby provides a very fast, flexible and reliable AGC, which places moderate demands both on the analogue and the digital components. As a consequence, the solution may be realised at a comparatively low cost. The power consumption also becomes reasonable.

Moreover, the solution is scalable with respect to frequency range and power level, such that it may be applied to many technologies in addition to the packet data protocols mentioned. Hence, the invention is also applicable to known TDMA- and CDMA-standards for public mobile data communication, such as GSM, EDGE, GPRS, IS-95B, IS-136, cdma2000, W-CDMA and IMT-2000. (TDMA=Time Division Multiple Access, CDMA=Code Division Multiple Access, GSM=Global System for Mobile Communication, EDGE=enhanced data rate for GSM Evolution, GPRS=General Packet Radio Service in GSM, IS-95B: a packet mode version of the direct sequence CDMA-standard IS-95 used in North America, IS-136: a TDMA cellular system standard predominantly used in North America, cdma2000: a proposed standard in the USA for the next generation of public mobile communication, WCDMA=Wideband CDMA, IMT-2000: harmonisation initiative in International Mobile Telecommunications in 2000 between USA, Europe and Asia under the Radiocommunication Standardization Sector of the International Telecommunication Union (ITU-R)).

BRIEF DESCRIPTION OF THE DRAWINGS

The present invention is now to be explained more closely by means of preferred embodiments, which are disclosed as examples, and with reference to the attached drawings. [0018]
FIG. 1 shows a block diagram over an arrangement according to an embodiment of the invention, [0019]
FIG. 2 illustrates a first aspect of the automatic gain control algorithm according to the invention by means of a power diagram, [0020]
FIG. 3 shows a diagram over a combined signal on which an automatic gain control is based according to an embodiment of the invention, [0021]
FIG. 4 shows a diagram over a time derivative parameter on which the automatic gain control may also be based according to an embodiment of the invention, and [0022]
FIG. 5 illustrates, by means of a flow diagram, an embodiment of the method according to the invention.[0023]

DESCRIPTION OF PREFERRED EMBODIMENTS OF THE INVENTION

FIG. 1 shows a block diagram over a receiver arrangement, which comprises a gain control loop according to an embodiment of the invention. The arrangement includes at least one [0024] radio frequency amplifier 102 for receiving radio frequency signals RF and producing corresponding amplified signals RF_P. A down converter 103 generates down converted signals I and Q from the amplified radio frequency signals RF_P. According to a preferred embodiment of the invention, the down converter 103 produces two quadrature signal components I and Q respectively by multiplying the amplified signal RF_Pwith a reference frequency LO from a local oscillator 121 respective with a phase shifted version of the reference frequency LO.
The quadrature signal components I and Q are passed to a set of [0025] amplifiers 104; 105, whose gain is controllable in response to an automatic gain control signal c and which generate corresponding amplified signal components I_Pand Q_P. According to a preferred embodiment of the invention the amplified signal components I_Pand Q_Pare also filtered and further amplified in subsequent low pass filters 106; 107 and amplifiers 108; 109 respectively. As a result, filtered and amplified signals I_P-LP; Q_P-LPare created, which represent information contained in the received radio frequency signals RF. The signals I_P-LP; Q_P-LPare received by respective A/D-converters 110; 111 where they are converted into a digital format I_D; Q_Dfor further processing in a digital signal processor 112.
Nevertheless, the filtered and amplified signals I[0026] _P-LP; Q_P-LPare also included in the gain control loop and thus fed back to indirectly control the gain of the amplifiers 104; 105. A first signal level detector 113 receives a first signal component Q_P-LPand produces in response thereto a first signal level [Q] representing the level or envelop of the first signal component Q_P-LP, i.e. a first order approximation. Naturally, any higher order of approximation of the signal [Q] may likewise be made, such as a second order power estimation. Correspondingly, a second signal level detector 114 receives a second signal component I_P-LPand produces a second signal level [I] representing the level or envelop of the second signal component I_P-LP.
A [0027] combiner 115 receives both the first and the second signal level [Q]; [I] and generates a combined signal V_Σ, which represents the sum of the signal levels [Q] and [I]. The combined signal V_Σ is passed on to a gain control signal generator 116 that produces an automatic gain control signal c on basis of i.a. the combined signal V_Σ.
The gain [0028] control signal generator 116 may include an integrator, which influences a time constant τ of the automatic gain control signal c in response to time properties of the combined signal V_Σ, such that a quickly varying combined signal V_Σresults in a comparatively short time constant τ, and vice versa, a slowly varying combined signal V_Σ results in a comparatively long time constant τ. During search for a radio frequency signal RF to receive or at other instances when a message is expected, the gain control signal generator 116 allocates a comparatively low value to the time constant τ (i.e. a short time constant τ). Otherwise, the gain control signal generator 116 sets a time constant τ value that is adapted to an actual power variation rate of the received radio frequency signals RF. More generally, the gain control signal generator 116 is adapted for varying the time constant τ in response to a time derivative parameter of the at least one down converted signal, i.e. I_P-LPor Q_P-LP, which describes a power variation rate of the received radio frequency signals RF. Naturally, this does not preclude that the time constant may be varied on basis of arbitrary additional parameters, which for instance, depend on the protocol and frame structure according to which the particular receiver operates.
On basis of a suitable input level of the first and second signal components I[0029] _P-LPand Q_P-LPto the dynamic range of the A/D- converters 110 and 111, the digital signal processor 112 produces a desired value V_desof the combined signal V_Σ. This value V_desis also fed to the gain control signal generator 116 to influence the automatic gain control signal c.
The automatic gain control signal c is fed in parallel to both the [0030] amplifiers 104 and 105 where it controls the gain, i.e. the amplification factor. The automatic gain control signal c is repeatedly updated to a value that is expected to make the amplifiers 104 and 105 respective 108 and 109 deliver the down converted signals at levels [I]; [Q] sufficiently close to a desired signal level. In practice, this also means that the automatic gain control signal c influences the amplifiers 104, 105,108 and 109 to deliver the down converted signals at levels [I]; [Q] less than a predetermined limit level. Preferably, the predetermined limit level is also optimised with respect to the dynamic range of the A/D- converters 110 and 111.
A [0031] derivation unit 117 also receives the combined signal V_Σ. This unit produces a time derivative parameter V_Σ by estimating a time derivative of the combined signal V_Σ, for instance by high pass filtering the combined signal V_Σ. The time derivative parameter V_Σis received in a ramp detector 118, for instance a comparator, together with a threshold value V_Tfrom the digital signal processor 112. If the time derivative parameter V_Σexceeds the threshold value V_T, the comparator 118 generates a pulse signal R. The digital signal processor 112 receives the pulse signal R and interprets the presence of the pulse signal R as an indication of a message start in the received radio frequency signals RF. The pulse signal R is activated i.a. when the RF power level rises sufficiently quickly from a relatively low, or even zero/noise floor level, to a relatively high level. By default, the time constant τ here has a comparatively low value and receiver is therefore well prepared for receiving the incoming message.
In an alternative embodiment of the invention, one or both of the [0032] derivation unit 117 and the ramp detector 118 are realised by functions within the digital signal processor 112.
In either case, after receiving the pulse signal R, the [0033] digital signal processor 112 computes a relevant starting value for the time constant τ (i.e. a short value) and delivers a corresponding digital value τ_set-Don an output. This value τ_set-Dis forwarded to a conversion unit 119 being adapted to the digital signal format used. The conversion unit 119 creates a corresponding analogue parameter τ_set. Hence, the unit 119 may be a conventional D/A-converter, a low pass filter or a pulse generator depending on the resolution requirements and the format of the digital signal τ_set-D. If necessary, a following time constant generator 120 produces an actual time constant τ on basis of the analogue parameter τ_set, for instance, by means of a table look-up in a bank of stored time constant values τ.
During reception of a message (e.g. a data packet) the signal level of the received radio frequency signals RF should only vary to small extent. The automatic gain control signal c should namely have calibrated the gain of the [0034] amplifiers 104 and 105 to this signal level, such that the time derivative parameter V_Σ is approximately zero, or at least below the threshold value V_T. Consequently, the pulse signal R should now be inactive. If, nevertheless, the digital signal processor 112 registers a pulse signal R during reception of a message, this is interpreted as an indication of an erroneous setting of the controllable gain in the amplifiers 104 and 105. Of course, this can in turn be a consequence of an external fact, such as a radical change in the radio environment. In any case, the affected message will normally be damaged or lost and will thus probably have to be re-transmitted.
This could imply that the receiver enters a search mode, in which a relevant portion of the radio frequency spectrum is scanned for possible radio frequency signals RF to receive. Alternatively, the receiver resets one or more of its parameters and attempts to re-establish the message reception. In any case, the [0035] digital signal processor 112 allocates a comparatively short time constant τ_set-D.
The pulse signal R in turn constitutes a basis for the time constant value τ[0036] _set-D. The pulse signal R history, and indirectly the threshold value V_Thistory, namely indicate that a particular stage in a typical message reception cycle has been reached. A short time constant value τ_set-Dshould always be applied in the beginning of the reception of a message. A large rise in the pulse signal R accompanied by a rise in the power level of the received radio frequency signals RF constitute indications of a message start and thus a short time constant value τ_set-Dis likely to be successful.
A relatively long time constant value τ[0037] _set-Dshould be applied whenever the radio frequency signals RF are received at a relatively high and constant power level and if simultaneously the absence of a pulse signal R indicates that a message is currently being received. A long time constant value τ_set-Dhere means that the time constant value τ_set-Dmay also be infinitely long. The received radio frequency power level can be determined via the digital signals I_D; Q_Dand the value of the automatic gain control signal c. The value of the automatic gain control signal c can in turn be detected by the digital signal processor 112 via an A/D-converter (not shown).
A relatively short time constant value τ[0038] _set-Dshould be applied whenever no radio frequency signals RF are being received or in other situations when a new message is expected, for instance, in case of a message being lost or damaged. Consequently, based on variations in the power level of the received radio frequency signals RF and the pulse signal R, the digital signal processor 112 can allocate a adequate time constant value τ_set-Dfor the automatic gain control signal c. After having received a complete message, the digital signal processor 112 typically allocates a short time constant value τ_set-Dand thus prepares the receiver for reception of a new message.
FIG. 2 shows a power diagram that illustrates a first aspect of the automatic gain control algorithm according to the invention. The horizontal axis represents time t and the vertical axis shows a transmitted/received signal power level in decibel. Naturally, the power values differ substantially depending on whether they reflect transmitted or received signal power. However, the interrelationship between the relevant levels and thresholds are substantially the same in both cases. A first power level PSS denotes the average signal level during transmission/reception of the payload information in the message. The transmitter should reach the level P[0039] _SSat latest at a rise time t_rafter initiating the transmission and the receiver should reach its level P_SSat a somewhat later time t_tafter that the transmission has been initiated. However, it is preferable if also the receiver has reached its level P_SSalready at the end of the rise time t_r. The period between the rise time t_rand the time t_tis namely utilised for calibrating certain units in the receiver, such as an equaliser. Thus, a so-called training sequence is typically sent between t_rand t_t. The 802.11a-protocol specifies a relatively short rise time t_rof 2 μs from initiating of the transmission. Since the receiver station cannot know when a transmission is in fact started, it uses a ramp detector, which here is illustrated by the threshold value V_Tfor the time derivative V_Σ of the combined signal V_Σ. Hence, the receiver station deduces that a new message is to be received whenever the time derivative V_Σ rises above the threshold value V_T. The receiver station construes that it has detected a so-called signal power ramp, which starts at point in time t_R, when the threshold value V_Tis exceeded. Provided that the rise time t_ris a system parameter known by the receiver station (e.g. t_r=2 μs), the receiver station can expect the signal power ramp to end within a time period equal to the rise time t_rafter t_R.
When the rise time t[0040] _rhas come to an end, the power P of the radio frequency signal preferably levels out at the first power level P_SS. During the period from t_rto t_t, when for instance training sequence is sent, the signal power level P typically shows a variation of ±ΔP_t=3 dB around the first power level P_SS(i.e. peak-to-average (PAP)=3 dB). According to 802.11a t_tis 8 μs and 802.11b t_tstipulates a t_tof 20-48 μs.
As of the time t[0041] _tand onwards the radio frequency signals should have an average power P level close to the first power level P_SSuntil the transmission is ended. The variation from this level is normally ±ΔP_D=10 dB (i.e. PAP=10 dB).
Even though the expected variation in power level P is lower between t[0042] _rand t_tthan after t_t, it is generally preferable if the time constant τ of the automatic gain control signal c is shorter between t_rand t_tthan after t_t. It is namely not desirable if the receiver “compensates” for the variations in the signal power after t_t, since such power variations may represent payload information in the message. Regardless of the time constant, the automatic gain control signal c should always have a value such that a down converted signal level is maintained below a limit level. This level is illustrated by means of a corresponding power value P_limin the diagram.
FIG. 3 shows a diagram over an exemplary combined signal V[0043] _Σ on which an automatic gain control c may be based according to an embodiment of the invention. The horizontal axis represents time t and the vertical axis shows the combined signal level V_Σ, for instance in volts. The FIG. 1 shows an embodiment of the invention where the combined signal V_Σ contains a sum of the signal levels [Q]; [I] of a first down converted signal component Q and a second down converted signal component I. If the down converted signal is a sine wave with a constant amplitude |s|, the combined signal level V_Σ will then vary between {square root}{square root over (2)}|s| and |s| also when the received signal level is constant. Consequently, a certain degree of variation in the combined signal level V_Σ must be tolerated without activation of the ramp detector 118. Of course, an alternative level detection and/or signal combination may be performed, which results in a smaller “undesired” variation in the combined signal level V_Σ. It is, nevertheless, still preferable that a moderate variation (i.e. non-zero time derivative) in the combined signal level V_Σ be accepted.
FIG. 4 shows a diagram over an exemplary time derivative parameter V[0044] _Σ according to an embodiment of the invention. Provided that the combined signal V_Σ is produced according to the design in FIG. 1 and that the down converted signal is a sine wave with a constant amplitude, the time derivative parameter V_Σ will also vary to some extent, for instance between a highest value Δ and a lowest value −Δ. In order not to generate a pulse signal R (or alter the time constant τ) in response to these variations, the threshold value V_Tmust lie sufficiently above Δ to also exclude variations due to the down converted signal being a non-sine wave signal or a signal composed of more than one sine wave. A pulse signal R should only be generated when the RF power level rises relatively quickly. However, as soon as the derivative parameter V_Σ exceeds the threshold value V_T, here at t=t_R, the ramp detector 118 produces a pulse signal R.
In order to sum up, a method for controlling gain of an amplifier for received radio signals in a radio receiver according to an embodiment of the invention will now be described with reference to a flow diagram in the FIG. 5. It should be borne in mind that even though the steps in the flow diagram are executed sequentially, this is merely true with respect to an infinitesimal fraction of a received signal. While, for instance a certain segment of the signal is being down converted, a somewhat later segment of the same signal is being received, and so on. [0045]
Radio frequency signals are received and amplified in a [0046] first step 501. A following step 502 down converts the signals into quadrature components, which are then amplified. Subsequently, a step 503 computes a time derivative parameter on basis of the quadrature components. A following step 504 determines an automatic gain control signal having a time constant value, which is based on the time derivative parameter computed in the step 503. Finally, the amplification of the quadrature components is controlled in a step 506, on basis of the automatic gain control signal, such that down converted signal components to be digitised are maintained at signal level less than a predetermined limit level.
All of the process steps, as well as any sub-sequence of steps, described with reference to the FIG. 5 above may be controlled by means of a computer program, such as a digital signal processor algorithm, being directly loadable into the internal memory of a general computer, a digital signal processor, a baseband processor or an ASIC (Application Specific Integrated Circuit), which includes appropriate software for controlling the necessary steps when the program is run on a computer/digital signal processor. The computer program can likewise be recorded onto arbitrary kind of computer readable medium. [0047]
The term “comprises/comprising” when used in this specification is taken to specify the presence of stated features, integers, steps or components. However, the term does not preclude the presence or addition of one or more additional features, integers, steps or components or groups thereof. [0048]
The invention is not restricted to the described embodiments in the figures, but may be varied freely within the scope of the claims. [0049]

Claims

1. A method for controlling gain of an amplifier for received radio signals in a radio receiver, the radio receiver having a controllable gain responsive to an automatic gain control signal (c), the method comprising:

amplifying received radio frequency signals (RF),

level detecting at least one down converted signal (I_P-LP; Q_P-LP) resulting from amplified radio frequency signals (RF_P), and

generating the automatic gain control signal (c) in response to at least one down converted signal level ([I]; [Q]) such that the controllable gain maintains the at least one down converted signal level ([I]; [Q]) at less than a predetermined limit level, characterised by

varying a time constant (τ) of the automatic gain control signal (c) in response to a time derivative parameter (V_Σ) of the at least one down converted signal (I_P-LP; Q_P-LP) such that the automatic gain control signal (c) is adapted to a power variation rate of the received radio frequency signals (RF).

2. A method according to claim 1, characterised by

allocating a relatively small value to the time constant (τ) if the time derivative parameter (V_Σ) has a comparatively high value, and

allocating a relatively large value to the time constant (τ) if the time derivative parameter (V_Σ) has a comparatively low value.

3. A method according to claim 2, characterised by generating a pulse signal (R) if the time derivative parameter (V_Σ) exceeds a threshold value (V_T), the pulse signal (R) at least being indicative of a message start in the received radio frequency signals (RF).

4. A method according to claim 3, characterised by during reception of a message the pulse signal (R) being indicative of an erroneous setting of the controllable gain.

5. A method according to any one of the claims 3 or 4, characterised by setting the threshold value (V_T) on basis of a power level (P_RF) of the received radio frequency signals (RF).

6. A method according to any one of the claims 3-5, characterised by digitally processing the pulse signal (R) to produce a time constant value (τ_set-D), the time constant value (τ_set-D) forming a basis for the time constant (τ) of the automatic gain control signal (c).

7. A method according to claim 6, characterised by updating the time constant value (τ_set-D) after a specific time period.

8. A method according any one of the claims 1-7, characterised by

combining two signal components of the at least one down converted signal (I_P-LP; Q_P-LP) to form a combined signal (V_Σ), and

producing the time derivative parameter (V_Σ) by estimating a time derivative of the combined signal (V_Σ).

9. A method according to claim 8, characterised by estimating time derivative of the combined signal (V_Σ) by high pass filtering the combined signal (V_Σ).

10. A method according to any one of the claims 8 or 9, characterised by deriving the two signal components (I_P-LP; Q_P-LP) by means of quadrature demodulation of the received radio frequency signals (RF).

11. A method according to any one of the claims 6-10, characterised by controlling the generation of the automatic gain control signal (c) on basis of the time constant (τ), a desired value (V_des) and an average value of the combined signal (V_Σ).

12. A computer program directly loadable into the internal memory of a digital computer, comprising software for performing the steps of any of the claims 1-11 when said program is run on a computer.

13. A computer readable medium, having a program recorded thereon, where the program is to make a computer perform the steps of any of the claims 1-11.

14. An arrangement for controlling gain of an amplifier for received radio signals in a radio receiver, comprising

at least one amplifier (102) for amplifying received radio frequency signals (RF),

a gain control loop (112-120) for controlling gain of at least one amplifier (104; 105) in response to an automatic gain control signal (c), the gain control loop (112-120) including

at least one signal level detector (113, 114) for receiving at least one down converted signal (I_P-LP; Q_P-LP) resulting from amplified radio frequency signals (RF_P) and producing in response thereto at least one down converted signal level ([I]; [Q]), and

a gain control signal generator (116) for receiving the at least one down converted signal level ([I], [Q]; V_Σ) and producing in response thereto the automatic gain control signal (c) such that the gain control loop (112-120) maintains the at least one down converted signal level ([I]; [Q]) at less than a predetermined limit level,

characterised in that the gain control signal generator (116) is adapted for varying a time constant (τ) of the automatic gain control signal (c) in response to a time derivative parameter (V_Σ) of the at least one down converted signal (I_P-LP; Q_P-LP) to compensate for a power variation rate of the received radio frequency signals (RF).

15. An arrangement according to claim 14, characterised in that the gain control signal generator (116) is adapted for

16 An arrangement according to claim 15, characterised in that the gain control loop (112-120) includes a ramp detector (118) for generating a pulse signal (R) if the time derivative parameter (V_Σ) exceeds a threshold value (V_T).

17. An arrangement according to claim 16, characterised in that the gain control loop (112-120) includes a digital signal processor (112) for receiving the pulse signal (R) and producing in response thereto a time constant value (τ_set-D), the time constant value (τ_set-D) forming a basis for the time constant (τ) of the automatic gain control signal (c).

18. An arrangement according to any one of the claims 16 or 17, characterised in that the digital signal processor (112) is adapted to set the threshold value (V_T) on basis of a power level (P_RF) of the received radio frequency signals (RF).

19. An arrangement according to any one of the claims 14-18, characterised in that the gain control loop (112-120) includes

at least two signal level detectors (113, 114), which each produces a respective down converted signal level ([I]; [Q]),

a combiner for receiving the down converted signal levels ([I]; [Q]) and producing in response thereto a combined signal (V_Σ), and

a derivation unit (117) for receiving the combined signal (V_Σ) and producing in response thereto a time derivative parameter (V_Σ).

20. An arrangement according to claim 19, characterised in that the derivation unit (117) includes a high pass filter.

21. An arrangement according to any one of the claims 19 or 20, characterised in that the radio receiver comprises a quadrature demodulator (103) and a frequency oscillator (121) for deriving two signal components (I_P-LP; Q_P-LP) by means of quadrature demodulation of the amplified received radio frequency signals (RF_P).

22. An arrangement according to any one of the claims 17-21, characterised in that

the gain control loop (112-120) includes means (119; 120) for receiving the time constant value (τ_set-D) and producing the time constant (τ) in response thereto,

the digital signal processor (112) is adapted for producing a desired value (V_des) indicative of an appropriate combined signal level, and

the gain control signal generator (116) is adapted for receiving the time constant (τ), the desired value (V_des) and a time limited average value of the combined signal (V_Σ) and producing in response thereto the automatic gain control signal (c).