A method of compensating for non-linearities in an end amplifier incorporated in a radio transmitter

Claim 1

A method suitable for an end amplifier (F) having a transfer function H R , Hφ, said amplifier being included in a radio transmitter of the quadrature type for linear digital modulation based on table look-up of an incoming signal vector α of given length containing the binary values of a momentary and an accumulated phase angle of an incoming information signal to the radio transmitter, said method comprising the steps of storing (ST,CT) the digital values I(t, α), Q(t, α) of the sine and cosine waveforms associated with the quadrature components to be supplied to the end amplifier, said digital values being are adressed by means of said signal vector α, multiplying (M1, M2) and adding said quadrature components in order to project these on two carrier wave components (sine w c t, cos w c t) before supplying the thus obtained quadrature modulated radio signal r(t, α) to the end amplifier (F), and modifying said digital quadrature component values I(t, α), Q(t, α) in dependence on amplitude and phase values which are supplied to the end amplifier for various input signals, characterized in storing (MH1,MH2), before operation of the radio transmitter a plurality of the inverse amplifying values (H R ⁻¹) and the corresponding phase value (H φ ) of the transfer function of the end amplifier for various values of the incoming information signal to the radio transmitter, and modifying said quadrature compoment values by a) calculating an absolute value of said quadrature modulated radio signal r(t,α) from said quadrature component values as non-compensated and, for each instant (t₁), b) adressing said plurality of stored inverse amplifying values and corresponding phase values in dependence on said calculated absolute value, c) forming (MS) sine- and cosine component values (sin H φ , cos H φ ) of said phase value adressed in step b), and d) multiplying the inverse amplifying values obtained in step b) by the sine-and cosine values obtained in step c) combined with the non-modified quadrature component values I(t,α), Q(t,α), respectively, such as to get modified quadrature component values i(t,α), q(t,α) with amplitude and phase modified in dependence on said stored values only and at each instant of the signal vector α which compensate the radio signal for said non-linearities in the end amplifier (F).