T 1996/22 (Power-tracking signal/QUALCOMM) 23-01-2024

Reasons for the Decision

1. AUXILIARY REQUEST 2b

1.1 Claim 1 - inventive step (Article 56 EPC)

1.1.1 Using the wording of claim 1 of auxiliary request 2b, document D3 discloses:

1 An apparatus of a wireless device ([0034]: "... base stations ... mobile terminals ..."), the apparatus comprising:

1.1 a power tracker (Figs. 3 and 4: blocks 104, 106, 108) configured to determine a single power tracking signal (Figs. 3 and 4: "control signal CTRL") based on a plurality of inphase, I, and quadrature, Q, components (Figs. 3 and 4: "abs(x) 104, 106" implies I and Q components) of all of a plurality of transmit signals (Figs. 3 and 4: "signals S1, S2") being sent simultaneously,

1.1.1 on a plurality of carriers at different frequencies (Fig. 3: "LO+IF1", "LO-IF2"; Fig. 4: "LO1, LO2"),

1.1.2 the power tracking signal has a bandwidth that is smaller than the overall bandwidth of the pluralities of carriers (Figs. 3 and 4, [0059], [0070] and [0071]);

[deleted: the power tracker is further configured to]

1.1.3 [deleted: determine an overall power of the plurality of transmit signals based on the I and Q components of the plurality of transmit signals at baseband,]

1.1.4 [deleted: the overall power is the sum of the powers of each of the transmit signals in the form of I]1**(2)[deleted: (t)+Q]1**(2)[deleted: (t)+...+I]k**(2)[deleted: (t)+Q]k**(2)[deleted: (t), wherein I]k**(2)[deleted: (t)+Q]k**(2)[deleted: (t) denotes the power of the k-th transmit signal in sample period t,]

1.1.5 [deleted: determine the power tracking signal based on the overall power of the plurality of transmit signals;]

1.2 a power supply generator (Figs. 3 and 4: power amplifier supply voltage module PA') configured to generate a single power supply voltage (Figs. 3 and 4: "supply voltage Vsup") based on the single power tracking signal;

1.3 a power amplifier, PA (Figs. 3 and 4: "power amplifier PA"), configured to receive the single power supply voltage and a single modulated radio frequency, RF, signal (Fig. 3: "RF signal SRF", Fig. 4: "RF signal S'RF") comprising all of the plurality of transmit signals, and configured to amplify the single modulated RF signal based on the single power supply voltage to produce a single output RF signal.

1.1.2 Thus, the subject-matter of claim 1 differs from the apparatus of D3 in features 1.1.3 to 1.1.5.

Contrary to the conclusions set out in Reasons 63 of the decision under appeal, the board agrees with the proprietor that "overall power" cannot be equated with "overall voltage". "Overall power", however broad, is still "power", and certainly not the power of an individual signal. There is a notorious difference between "power" and "voltage" or "amplitude". Even if the computation of the "absolute value" of signals S1 and S2 depicted in Figures 3 and 4 of D3 typically involved the intermediate calculation of their individual powers, document D3 does not disclose the calculation of an "overall power", let alone the use of the specific formula of feature 1.1.4.

1.1.3 Opponent 1 submitted that D3 used for the power tracking an "envelope signal". Such signal was typically an oscillating signal which outlines its amplitude, i.e. extremes, in a smoothed way. The smoothing property of the envelope signal could be achieved for instance with a low-pass filter (cf. paragraph [0050] of the opposed patent). The common approach applied in the opposed patent and in the system of D3 to smooth the envelope was to tap the baseband signal, i.e. before the signal was upconverted. When the signal was composed of many independent individual signals, as was the case here, the individual signal had to be considered in the computation of the envelope signal. Document D3 (Fig. 3) and Equation (2) of the opposed patent used the arithmetical mean or "L1-norm" of the amplitudes of the complex baseband signals, whereas Equation (1) of the opposed patent used the quadratic mean or "L2-norm" of the amplitudes of the complex baseband signals.

1.1.4 According to the proprietor, opponent 1's obviousness arguments ignored the technical effect of determining the "overall power" of the transmit signals according to feature 1.1.4 as a basis for the calculation of the "power tracking signal". This calculation - as applied for instance in Equation (1) of the opposed patent - led to a reduced bandwidth when compared with the power tracking signal calculated using Equation (2) of the opposed patent or the approach of Figure 3 of D3. The different methods for the determination of the "power tracking signal" described in the opposed patent were not just interchangeable. It was apparent to the skilled person that the use of the overall power led to a significant reduction of the bandwidth of the power tracking signal. Using the overall power instead of the individual voltages of the transmit signals as a basis struck a balance between efficiency and noise at the power tracker. On the basis of Equation (1), the approximation of the envelope of the output RF signal was less accurate, which meant that the power tracker was "lazier" and that there was a further "smoothing" effect which went hand in hand with a reduced bandwidth of the "power tracking signal". Document D3 did not suggest in any way to use the overall power of the transmit signals according to feature 1.1.4 as a basis for determining the power tracking signal. Rather, D3 taught to preferably use the embodiment according to Figure 2, in which the power tracking signal (uncontestedly) did not have a reduced bandwidth with the technical benefit of being basically identical with the envelope signal. Only exceptionally, namely in case of large frequency spacings between the input signals, D3 suggested to determine the "power tracking signal" at baseband level leading to a reduction of the bandwidth. This corresponded to the embodiment of Figure 3 of D3. Only for this specific purpose, D3 disclosed summing the absolute values (voltages) of the individual signals in order to determine the power tracking signal. Based on this, there was simply no motivation at all for the skilled person to look for other ways to determine the power tracking signal, which would even lead to a further reduction of the bandwidth, i.e. an even less accurate approximation of the envelope of the RF output signal. Thus, the skilled person would not have modified the embodiment covered by Figure 3 of D3 in a way to arrive at the teaching of claim 1 according to auxiliary request 2b. The opposition division endorsed this line of argumentation in Reasons 75 of the decision under appeal. Finally, according to paragraph [0051] of the opposed patent, the "digital power tracking signal" might be computed in different manners dependent on the number of transmit signals being sent simultaneously or other factors. For instance, Equation (1) would be more appropriate for a higher number of transmit signals, as the power tracker would then become busier.

1.1.5 The board considers that the technical effect indicated by the proprietor and by the opposition division - inasmuch as it is founded on the assertions in paragraph [0049] of the opposed patent - cannot be credibly associated with features 1.1.3, 1.1.4 and 1.1.5. More specifically, paragraph [0049] of the opposed patent states the following:

"... [t]he digital power tracking signal computed in equation (1) or (2) has a bandwidth that approximates the bandwidth of the widest transmit signal (instead of the overall bandwidth of all transmit signals being sent simultaneously). Having the bandwidth of the power tracking signal being smaller than a modulation bandwidth may allow for a more efficient power tracking circuitry and may also result in less noise being injected into PA 560 via the power supply."

It follows that paragraph [0049] of the opposed patent provides no conclusive information about how the bandwidth of the power tracking signal resulting from Equation (1) ("the digital power tracking signal is then obtained by taking the square root of the overall power") compares with the one resulting from Equation (2) ("the voltages of all transmit signals are then summed to obtain the digital power tracking signal"). At any rate, this paragraph suggests that both "power tracking signals" should have a similar bandwidth, i.e. that of the transmit signal having the broadest bandwidth. And the advantages identified therein ostensibly concern both Equations (1) and (2).

1.1.6 As regards any "further smoothing" beyond the bandwidth being approximated to the bandwidth of the widest transmit signal, the board agrees with opponent 1 that this would have to be achieved by means of a subsequent low-pass filter, as indicated in paragraph [0050] of the opposed patent. However, such a filter is not present in claim 1. Nor does the opposed patent give any indications as to whether or not the "power tracking signal" computed according to Equation (1) would indeed enable narrower filtering than the signal computed according to Equation (2).

1.1.7 In summary, the board does not dispute that - applied to the same plurality of transmit signals - the apparatus of D3 and the claimed one would generate different "power tracking signals". The question at stake is whether or not these differences would actually translate into a tangible and noticeable technical effect or improvement, such as the bandwidth reduction asserted by the proprietor. The board is not convinced that this is the case. In the absence of a credible technical effect associated with features 1.1.3, 1.1.4 and 1.1.5 of claim 1, the use in D3 of a "sum of powers" instead of the disclosed "sum of absolute values" cannot be considered to contribute to the presence of an inventive step (see e.g. G 1/19, Reasons 49 and 124).

1.1.8 Finally, both in the opposed patent (see e.g. paragraph [0055]) and in Figure 3 of D3, the "power tracking signal" is intended to follow the envelope of the output RF signal. Even if the board were to assume - arguendo - that the calculation of the "overall power" according to features 1.1.3, 1.1.4 and 1.1.5 did indeed lead to a reduced bandwidth of the power tracking signal and to the technical effects ascribed thereto by the proprietor, i.e. "lazier" power tracker and better balance between efficiency and noise, the advantages - as well as the drawbacks - derived from the mathematical properties of the known envelope-tracking algorithms - including "absolute value" and "squaring" - would have been apparent. Starting from document D3, a person skilled in the art would have chosen one tracking algorithm or the other, according to the practical circumstances, without the need of any inventive step.

1.2 It follows that auxiliary request 2b is not allowable under Article 56 EPC.

2. Since there is no allowable claim request on file, the patent must be revoked.