Leonard R. Kahn (Kahn) appeals the judgment of the United States District Court for the Southern District of New York that Claims 53 and 54 of U.S. Patent No. 4,018,994 (the '994 patent) are invalid and not infringed by General Motors Corporation (GM). See Kahn v. General Motors Corp., No. 88-CV-2982, 1997 WL 37581 (S.D.N.Y. Jan.29, 1997). We affirm-in-part and reverse-in-part.

Kahn filed an action in the United States District Court for the Southern District of New York on April 29, 1988 alleging that GM's AM stereo radio receivers infringed Claims 53 and 54 of the '994 patent. After nine years of litigation and twelve judicial opinions,1 including three appeals to this court,2 it appears that this litigation may have finally reached a conclusion. The most recent installment of this case was argued before this court on November 5, 1997.

Kahn's patent application was filed May 2, 1975. The '994 patent, entitled "Compatible AM Stereophonic Receivers", subsequently issued on April 19, 1977.4 The claimed invention addresses the problem of minimizing the distortion within an AM stereo receiver. The Kahn receiver is designed to receive transmissions in which the two stereo audio signals, the left (L) and right (R) signals, are transmitted by an independent single sideband system. The information is encoded by envelope or amplitude modulation of the carrier wave by the stereo sum (L+R) signal component. According to the specification of the '994 patent, the most important feature of this invention is that the receiver derives a distortion cancellation component from the L+R signal and its higher harmonics and uses it to inversely amplitude modulate the stereo difference (L-R) signal. This substantially reduces the distortion of the L-R signal caused by amplitude modulation of the carrier wave. This concept is embodied in Claims 53 and 54 of the '994 patent, which read as follows:

53. In a compatible AM stereo receiver of the type for receiving a carrier wave amplitude modulated with a stereo sum (L+R) signal component and phase modulated with a stereo difference (L-R) signal component, and utilizing detection means for deriving the L+R audio signal and demodulating means for deriving the L-R audio signal and means for combining the L+R and L-R signals to isolate the L and R signal outputs, the improvement comprising:

means for deriving a distortion cancellation component from the stereo sum signal component, and

means for combining said distortion cancellation component with a signal including the stereo difference signal component information to substantially reduce distortion of the stereo difference signal component.

54. A receiver according to claim 53, wherein said means for deriving the distortion cancellation component comprises means separating a portion of the L+R audio signal, and said means for combining the separated distortion cancellation component with said signal including the stereo difference signal component information comprises means inversely amplitude modulating the modulated carrier wave with such L+R audio signal portion prior to demodulation thereof.

(emphasis added)

Simple quadrature modulation is often used to encode and transmit a stereo signal to compatible quadrature receivers. Quadrature modulation essentially involves the transmission of two AM signals on the same channel. This allows both the "left" speaker signal and the "right" speaker signal to be transmitted together on the same channel. Before transmission, the L+R and L-R signals are phase shifted relative to one another and combined into a single signal, which is then transmitted to a receiver. The receiver is able to decode this combined signal into its components and present a stereo signal at the speakers.

The accused Compatible Quadrature Amplitude Modulated (C-QUAM) device works in a similar manner. The signal created by quadrature amplitude modulation, however, is not compatible with monophonic AM radio receivers, which cannot decode the quadrature modulated signal, but can only detect envelopes. Therefore, at the transmitter, the C-QUAM system first normalizes the signal so that it can be decoded by standard monophonic AM circuitry. To ensure that these monophonic receivers can also properly derive a monophonic signal from the stereo transmission, the C-QUAM system multiplies the L+R and L-R signals by the cosine of the phase angle (cos(G)) between the two signals to make the signal compatible with standard AM radio receivers. This adjusts the envelope of the quadrature signal to make it compatible with standard monophonic AM radio receivers, which can then interpret the signal as a monophonic signal without having to decode the C-QUAM stereo signals. The resulting L+R and L-R signals are also phase shifted relative to one another in the transmitter. The C-QUAM transmission is in effect the hypotenuse of the triangle created by the (L+R)*cos(G) and (L-R)*cos(G) signal vectors. Upon reception of a quadrature signal by a C-QUAM receiver, the cos(G) signal is recovered by comparing the envelope to a portion of the quadrature signal. The received C-QUAM signal is then divided by cos(G) to restore the L+R and L-R signals to their original form. These stereo signals can then be demodulated and sent to the speakers.

Kahn claims that the Motorola circuit used in the GM C-QUAM receivers derives a distortion cancellation component from the L+R signal and uses it to inversely modulate the L-R signal to substantially reduce the distortion in the L-R signal as claimed in Claims 53 and 54 of the '994 patent.

In determining whether there has been infringement, a two step analysis is required. First, the claims must be correctly construed to determine the scope of the claims. Second, the claims must be compared to the accused device. See Markman v. Westview Instruments, Inc., 52 F.3d 967, 976, 34 USPQ2d 1321, 1326 (Fed.Cir.1995) (in banc), aff'd, 517 U.S. 370, 116 S. Ct. 1384, 134 L. Ed. 2d 577, 38 USPQ2d 1461 (1996); Vitronics Corp. v. Conceptronic, Inc., 90 F.3d 1576, 1581-82, 39 USPQ2d 1573, 1576 (Fed.Cir.1996); Hormone Research Found., Inc. v. Genentech, Inc., 904 F.2d 1558, 1562, 15 USPQ2d 1039, 1042 (Fed.Cir.1990). To establish literal infringement, a plaintiff must demonstrate that every limitation in the claim is literally met by the accused device. See Intellicall, Inc. v. Phonometrics, Inc., 952 F.2d 1384, 1388-89, 21 USPQ2d 1383, 1387 (Fed.Cir.1992); see also Key Mfg. Group, Inc. v. Microdot, Inc., 925 F.2d 1444, 1449, 17 USPQ2d 1806, 1810 (Fed.Cir.1991). Demonstration that every limitation of the claim is literally met by the accused device must be shown by a preponderance of the evidence. See Envirotech Corp. v. Al George, Inc., 730 F.2d 753, 758, 221 U.S.P.Q. (BNA) 473, 477 (Fed.Cir.1984).

Claim construction is a matter of law and is reviewed by this court de novo. Markman, 52 F.3d at 974. Claim 53, the only independent claim at issue in this appeal, is written in means-plus-function language as permitted by 35 U.S.C. § 112, p6. Claims written in means-plus-function form are interpreted to cover the structure set forth in the specification and its equivalents. See 35 U.S.C. § 112, p6; Intellicall, 952 F.2d at 1388; In re Iwahashi, 888 F.2d 1370, 1375, 12 USPQ2d 1908, 1912 (Fed.Cir.1989). The only claim limitation at issue in this appeal is the limitation in Claim 53, which requires "means for deriving a distortion cancellation component from the stereo sum signal component." The district court found that the parties had stipulated that the proper construction of the term "deriving" was "getting something from something else." Kahn, 1997 WL 37581, at * 5. Kahn does not appeal this finding. We therefore adopt this interpretation of the term "derive".

Unlike the ordinary situation in which claims may not be limited by functions or elements disclosed in the specification, but not included in the claims themselves, in writing a claim in means-plus-function form, a party is limited to the corresponding structure disclosed in the specification and its equivalents. See 35 U.S.C. § 112, p6. A structure disclosed in the specification is only deemed to be "corresponding structure" if the specification clearly links or associates that structure to the function recited in the claim. See B. Braun Med., Inc. v. Abbott Lab., 124 F.3d 1419, 1424, 43 USPQ2d 1896, 1900 (Fed.Cir.1997). The duty to link or associate structure in the specification with the function is the quid pro quo for the convenience of employing § 112, p6. See O.I. Corp. v. Tekmar Co., 115 F.3d 1576, 1583, 42 USPQ2d 1777, 1782 (Fed.Cir.1997).

There is sufficient structure disclosed in the '994 patent associated with the claim limitations. Kahn points to the structure associated with each limitation of Claim 53. Kahn claims that the limitation requiring "means for deriving a distortion cancellation component from the stereo sum signal component" is found in Figures 1, 2 and 3 and is labeled therein as parts 40, 44, 54, 54' and 96'. Kahn also claims that this limitation is disclosed in the specification of the '994 patent at column 5, lines 1 to 13. Turning to the Figures 1, 2 and 3 of the specification of the '994 patent, the above mentioned references disclose that part 40 refers to an amplifier, part 44 refers to a summer, and part 54 refers to a frequency doubler. Parts 54' and 96' also refer to frequency doublers in alternative embodiments of the claimed invention. Column 5, lines 1 to 13 of the written description disclose that the output from the envelope detector (L+R) is fed into the amplifier 40 and is summed together (along with the output from the frequency doubler 54) by the summer 44. The resulting summed signal is used to inversely amplitude modulate the stereo difference signal (L-R) to reduce distortion. The specification discloses a method for feeding the L+R signal and its higher harmonics taken from the frequency doubler to an inverse modulator, which then is used to eliminate the distortion caused by the amplitude modulation of the carrier wave. Kahn claims that the feedback circuit used in the GM receiver is an equivalent of the above described structure. We disagree with this characterization.

B. Comparison of the Claims to the Accused Device

Literal infringement of a claim exists when every limitation recited in the claim is found in the accused device. See Strattec Security Corp. v. General Automotive Specialty Co., 126 F.3d 1411, 1418, 44 USPQ2d 1030, 1036 (Fed.Cir.1997); Johnston v. IVAC Corp., 885 F.2d 1574, 1580, 12 USPQ2d 1382, 1384 (Fed.Cir.1989) (collecting cases). The district court found that Kahn failed to sustain his burden of proving that the GM system embodies every element of Claim 53 because he did not show that the GM circuit included "means for deriving a distortion cancellation component from the stereo sum signal component." Kahn, 1997 WL 37581, at * 3. For the same reasons, the district court also found that the GM circuit did not infringe dependent Claim 54.

The feedback circuit used in the GM receiver consists of an envelope detector, a comparator and an inverse modulator. The comparator uses as one input the output from an envelope detector, or the L+R signal. On the other input is the output from the demodulator, which is the L+R signal plus the cos(G) signal. The output from the comparator is the cos(G) signal, which is connected to the inverse modulator and is used to inverse modulate the input signal to remove the cos(G) signal.

Kahn also claims that part of the distortion cancellation signal in Claim 53 is an element of background noise derived from the L+R signal. This background noise is used to reduce or eliminate the background noise present in the L-R signal. Kahn claims that the accused device also uses the background noise derived from the L+R signal to reduce the distortion caused by background noise in the L-R signal. Kahn is incorrect. To the extent that the background noise is derived from the L+R signal, the noise is eliminated by the comparator, which uses the exact same L+R signal including background noise to isolate the cos(G) signal from the L+R signal. The background noise is therefore present on both inputs to the comparator, and each cancels the other from the signal, leaving the cos(G) signal as the sole input to the inverse modulator.

Despite Kahn's claim that the feedback system used in the GM circuit is the feedforward equivalent of the structure disclosed in the '994 patent, there are significant differences between the two structures. The structure of the '994 patent includes a frequency doubler, which has no counterpart equivalent in the GM receiver that performs a similar function. Likewise, the summer described in the '994 specification has no counterpart in the GM circuit. It is clear that the GM feedback circuit employing a comparator is not an equivalent to the amplifier, frequency doubler, summer combination disclosed in the '994 specification. These two circuits utilize very different structures, for different purposes. As a result, we hold that there is no counterpart equivalent structure in the accused device that meets the means-plus-function limitation of Claim 53, which requires "means for deriving a distortion cancellation component from the stereo sum signal component."

The absence of even a single limitation of Claim 53 from the accused device precludes a finding of literal infringement. See Key Mfg., 925 F.2d at 1449. Because a limitation contained in Claim 53 is not met by the accused device, Claim 53 is not literally infringed. Similarly, because Claim 54 contains limitations not contained in Claim 53, Claim 54 is narrower than Claim 53 and is therefore also not infringed by the accused device.

Kahn also claims that the GM circuit infringes under the doctrine of equivalents. Under the doctrine of equivalents, an accused device "that does not literally infringe upon the express terms of a patent claim may nonetheless be found to infringe if there is 'equivalence' between the elements of the accused product or process and the claimed elements of the patented invention." Warner-Jenkinson Co. v. Hilton Davis Chem. Co., --- U.S. ----, ----, 117 S. Ct. 1040, 1045, 137 L. Ed. 2d 146, 41 USPQ2d 1865, 1868 (1997); see also Graver Tank & Mfg. Co. v. Linde Air Prods. Co., 339 U.S. 605, 608, 70 S. Ct. 854, 856, 94 L. Ed. 1097 (1950); Pennwalt Corp. v. Durand-Wayland, Inc., 833 F.2d 931, 934, 4 USPQ2d 1737, 1739 (Fed.Cir.1987) (in banc).

The district court found that the GM circuit was substantially different from the system claimed in Claims 53 and 54. See Kahn, 1997 WL 37581, at * 4. We review a decision of infringement under the doctrine of equivalents by a district court for clear error. See Pennwalt, 833 F.2d at 936. We agree with the district court that the accused device is not an equivalent of the claimed invention and does not infringe under the doctrine of equivalents. The record before us is devoid of any evidence that would support a finding that the limitations specified in the claim that are not found exactly in the accused device are met by equivalents in the accused device. In fact, it is clear from the record that the accused device performs a substantially different function in a different way to achieve an entirely different result.

The doctrine of equivalents is applied to each individual element of a claim, not to the invention as a whole. See Warner-Jenkinson, --- U.S. at ----, 117 S. Ct. at 1049. Focusing on the individual elements of Claim 53, it is clear that the function of the limitation in Claim 53 which requires "means for deriving a distortion cancellation component from the stereo sum signal component" is to reduce distortion in the input signal. Kahn claims that the comparator and other related components in the accused device correspond to this claim element. However, these components function not to reduce distortion in the input signal, but to remove the cos(G) signal in order to recover the original L-R signal. These are not substantially equivalent functions. The "means for deriving" limitation of Claim 53 functions to reduce distortion caused by amplitude modulation, while the comparator and related components in the accused device enable the circuit to remove an unnecessary signal added merely for backwards compatibility with existing systems.

The limitation in Claim 53 requiring "means for deriving a distortion cancellation component from the stereo sum signal component" also uses a significantly different structure to accomplish its function. As stated above, the amplifier, frequency doubler, summer combination which define the "means for deriving" are significantly different from the feedback structure of the accused device, which employs a comparator. The way these elements accomplish their respective functions cannot be said to be substantially the same.

Finally, the result is substantially different. The result of the structure defined by the "means for deriving" limitation in Claim 53 is the reduction of distortion in the input signal. The result of the allegedly equivalent structure in the accused GM circuit is merely the elimination of the cos(G) component of the input signal. The L+R signal plays no part in the inverse modulation of the L-R signal as it is completely eliminated by the comparator, which outputs only the cos(G) signal.

For the foregoing reasons, the above described structure of the GM circuit cannot be considered an equivalent to that disclosed in the specification of the '994 patent corresponding to the limitation in Claim 53 requiring "means for deriving." As a result, there can be no infringement under the doctrine of equivalents.PATENT VALIDITY

The district court found Claims 53 and 54 of the '994 patent invalid under 35 U.S.C. § 103 (1994) as obvious over U.S. Patent No. 2,420,230 issued to Crosby on May 6, 1947, (the Crosby patent) and U.S. Patent No. 3,080,453 issued to Avins on March 5, 1963, (the Avins patent). See Kahn, 1997 WL 37581, at * 7. The Avins patent, but not the Crosby patent, was before the examiner during the prosecution of the '994 patent. The district court based its finding in large part upon the expert testimony proffered by GM. Because the Avins patent combined with the Crosby patent would not have made using the envelope signal with the addition of higher harmonics of the envelope signal obvious to one of ordinary skill in the art, we reverse the decision of the district court.

When a district court holds a patent invalid for obviousness, we review the ultimate legal conclusion as a matter of law de novo. See Stiftung v. Renishaw PLC, 945 F.2d 1173, 1182, 20 USPQ2d 1094, 1102 (Fed.Cir.1991). However, we review the underlying factual findings such as differences between the prior art and the claimed invention, the level of ordinary skill in the art, and objective evidence of secondary considerations of patentability for clear error. See id. "A finding is 'clearly erroneous' when although there is evidence to support it, the reviewing court on the entire evidence is left with the definite and firm conviction that a mistake has been committed." United States v. United States Gypsum Co., 333 U.S. 364, 395, 68 S. Ct. 525, 542, 92 L. Ed. 746 (1948).

GM argued, and the district court found, that Claims 53 and 54 were invalid because they would have been obvious in light of the Avins and Crosby references. See Kahn, 1997 WL 37581, at * 6. The record clearly supports the district court's finding that Kahn admitted in his deposition that each element in the preamble of Claim 53 could be found within the Avins patent. We turn now to the teachings of the Crosby patent, which was issued almost thirty years before the '994 patent issued.

The Crosby patent teaches the use of a distortion correction signal derived from the envelope signal to inversely modulate the phase modulated component of the input signal in order to eliminate some of the distortion that inevitably occurs as the signal is recovered from the envelope. The Crosby patent is essentially directed towards a process for reducing the distortion caused to an FM signal by an AM envelope. The specification of the Crosby patent discloses that, by using inverse modulation, the limiter often used to eliminate the AM envelope can be dispensed with because the elimination of amplitude modulation by the balanced detector is made more complete. Essentially, the Crosby system modulates the envelope in inverse proportion to the envelope amplitude. In so doing, the Crosby system reduces the amplitude of the envelope when it is at a high point and amplifies the amplitude of the envelope when it is at a low point. The Crosby patent states that "[t]he inverse modulation ... opposes the undesired amplitude modulation which is present, and tends to cancel it." (Crosby, col.2, ll.29-32). As a result, the envelope is more completely eliminated without the need to use a limiter. The Crosby patent teaches that a better alternative to using a limiter to eliminate the L+R envelope is to inversely modulate the envelope using the envelope signal itself.

Under 35 U.S.C. § 103, "[a] patent may not be obtained ... if the differences between the subject matter sought to be patented and the prior art are such that the subject matter as a whole would have been obvious at the time the invention was made to a person having ordinary skill in the art to which said subject matter pertains." The determination under § 103 is whether the claimed invention as a whole would have been obvious to a person of ordinary skill in the art at the time the invention was made. See In re O'Farrell, 853 F.2d 894, 902, 7 USPQ2d 1673, 1680 (Fed.Cir.1988). Obviousness may not be established using hindsight. See W.L. Gore & Assocs., Inc. v. Garlock, Inc., 721 F.2d 1540, 1551, 220 U.S.P.Q. (BNA) 303, 312-13 (Fed.Cir.1983). In determining obviousness, the invention must be considered as a whole and the claims must be considered in their entirety. See Medtronic, Inc. v. Cardiac Pacemakers, Inc., 721 F.2d 1563, 1567, 220 U.S.P.Q. (BNA) 97, 101 (Fed.Cir.1983). An issued patent is presumed to be valid and, hence, nonobvious. See 35 U.S.C. § 282 (1994). Thus, in this case, it is GM's burden to prove by clear and convincing evidence that the claimed invention would have been obvious in view of the prior art. See Glaverbel Societe Anonyme v. Northlake Mktg. & Supply, Inc., 45 F.3d 1550, 1555, 33 USPQ2d 1496, 1499 (Fed.Cir.1995); see also Polaroid Corp. v. Eastman Kodak Co., 789 F.2d 1556, 1558, 229 U.S.P.Q. (BNA) 561, 562-63 (Fed.Cir.1986). The presentation of evidence that was not before the examiner does not change the presumption of validity, although the burden may be more or less easily carried because of the additional evidence. See Applied Materials, Inc. v. Advanced Semiconductor Materials Am., Inc., 98 F.3d 1563, 1569, 40 USPQ2d 1481, 1485 (Fed.Cir.1996).

Kahn agreed in his deposition that the Avins patent teaches every limitation of Claim 53 except the limitation that requires "means for deriving a distortion cancellation component from the stereo sum signal component." As we noted above, Claim 53 is written in means-plus-function form pursuant to 35 U.S.C. § 112, p6. Means-plus-function claims are limited to the structure disclosed in the specification and "equivalents thereof." 35 U.S.C. § 112, p6. The specification of the '994 patent clearly discloses a structure that isolates the L+R envelope signal and uses the envelope signal to inversely modulate the L-R signal to reduce distortion. This is clearly disclosed by the Crosby patent. However, as we discussed above, the structure shown in the specification of the '994 patent also discloses a frequency doubler and a summer, which are used to add higher harmonics to the original envelope signal. The claim is limited to this structure as well. Indeed, one of the bases for the finding of noninfringement of the accused device was the limitation of the claim to a structure for including higher harmonics of the L+R signal in the distortion cancellation component. "It is insufficient to establish obviousness that the separate elements of the invention existed in the prior art, absent some teaching or suggestion, in the prior art, to combine the elements." Arkie Lures, Inc. v. Gene Larew Tackle, Inc., 119 F.3d 953, 957, 43 USPQ2d 1294, 1297 (Fed.Cir.1997). The district court erred in failing to limit Claim 53 to the structure in the '994 specification for adding higher frequency components to the distortion cancellation signal. Such a limitation precludes a finding of obviousness based on the combination of the Avins and Crosby patents.

Crosby clearly discloses the use of the envelope signal to eliminate distortion in the L-R signal. Combined with the Avins patent, the Crosby patent teaches all of the elements of Claim 53 save one--the addition of higher harmonics to the distortion cancellation component. There is nothing in the prior art that discloses that the use of higher harmonics to reduce distortion would be desired or even useful to reduce the distortion in the L-R signal.

As discussed above, the Crosby invention used only the L+R envelope signal to eliminate the effects of amplitude modulation by the transmitter. The purpose of inverse modulation in the Crosby patent is to cancel the L+R envelope as completely as possible. There is nothing in the Crosby patent to suggest to one of ordinary skill in the art that doubling the frequency of the envelope and adding it to the distortion cancellation component would improve the noise characteristics of the resulting signal. Hence, it would not have been obvious to one of ordinary skill in the art to add higher harmonics to the Avins-Crosby combination to achieve the result of Claim 53.

GM has not proven by clear and convincing evidence that the invention defined by Claim 53 would have been obvious to one of ordinary skill in the art. As a result, we find that the district court erred in finding that the claimed invention would have been obvious over the combination of Avins and Crosby. Similarly, because Claim 54 depends from Claim 53, the invention defined by Claim 54 also would not have been obvious in light of the prior art. Accordingly, the judgment of the district court with respect to the validity of Claims 53 and 54 of the '994 patent is reversed.CONCLUSION

We hold that the accused device does not literally infringe either Claim 53 or 54 because the accused device does not contain an exact or equivalent structure to that disclosed in the specification of the '994 patent. The functions and structure of each circuit are so different that there is no equivalency under the doctrine of equivalents. We also hold that the invention of Claims 53 and 54 of the '994 patent would not have been obvious based on the combination of the Avins and Crosby patents because the combination of Avins and Crosby would not have taught one of ordinary skill in the art to utilize a structure for adding higher harmonics to the distortion cancellation signal. The judgment of the district court is affirmed with respect to its finding of noninfringement and reversed with respect to its finding that the invention of Claims 53 and 54 would have been obvious based on the prior art.

AFFIRMED-IN-PART and REVERSED-IN-PART.