CONFIDENTIAL — ATTORNEYS EYES ONLY
`
`IN THE UNITED STATES DISTRICT COURT
`
`FOR THE EASTERN DISTRICT OF TEXAS
`TYLER DIVISION
`
`NETWORK-1 SECURITY SOLUTIONS,
`
`v
`
`INC.,
`
`v.
`
`Plaintiff,
`
`D~LINK CORPORATION, et al. ,
`
`Jury Trial Demanded
`
`Civil Action No. 6:05-cv—291-LED
`
`Defendants.
`
`EXPERT REBUTTAL REPORT
`
`OF JAMES KNOX, Ph.D.
`
`122904.00601/35776962v.1
`
`

`

`CONFIDENTIAL — ATTORNEYS EYES ONLY
`
`1. All opinions and facts stated herein are true and correct to the best of my
`knowledge. If called upon to testify, I could and would testify to the truth of the
`following.
`I have worked in the design of computer hardware and sofiware for over 40
`2.
`I received a Ph.D. in Electrical Engineering from the University of Texas in 1978,
`years.
`a Master’s degree in Computer Science from the University of Texas in 1971, and a
`Bachelor’s degree in Electrical Engineering from the University of Texas in 1969.
`I have
`taught Electrical Engineering and Computer Science at the University of Texas, and am
`currently the owner of a computer technology company called TriSoft located in Austin,
`Texas. During my career, I have designed and implemented a variety of communications
`systems (both hardware and software), including systems for military, commercial, and
`security users. A copy of my resume is attached as Exhibit 4.
`3.
`I have been retained by Network-l to provide my expert opinions in this case.
`My compensation is $160 per hour. My compensation is not based on the outcome of
`this litigation.
`4.
`If called at trial, I expect to testify on the matters set forth herein. It is also
`possible I may testify in response to testimony given by experts for parties other than D-
`Link in rebuttal to the positions set forth in this report. Also, as set forth above, it is my
`understanding that the parties are still exchanging information in this case, and I reserve
`the right to expand on or modify the positions I set forth in this report based on any
`information received from parties other than Network-l after the submission of this
`report.
`
`I have reviewed the Initial Expert Reports of Gregory Ennis and Steven B.
`5.
`Carlson in the case Network-1 Security Solutions, Inc. v. D-Link Corporation, et a1.
`6. Within the framework of this review I have relied on knowledge of the US.
`Patent 6,218,930 (“the Katzenberg patent”), of the Court’s Markman claim construction
`regarding this case, and the documents identified in my previously submitted expert
`report.
`I have further reviewed the additional patents and patent applications cited within
`the Ennis report. In addition, I have considered the state of the art as was known at the
`time of the filing of the Katzenberg patent, and of the Katzenberg Provisional Patent
`Application which preceded it.
`7.
`It is my opinion that the arguments made by Mr. Ennis and Mr. Carlson are
`not valid and that their conclusions are therefore invalid.
`I am not persuaded by their
`statements to alter my own assessment that the Katzenberg patent represents a valid
`teaching of the invention stated, enabling one of ordinary skill in the art to create and
`utilize the described invention. Further, I find no compelling argument that there exists
`prior art which would invalidate the Katzenberg patent.
`In arriving at this conclusion I
`have taken into account the filing date of the Katzenberg patent w the date of the
`referenced Provisional Application for Patent, Serial Number 60/123,688 (“the
`Katzenberg provisional”), as well as the level of ordinary skill in the art at the time the
`Katzenberg provisional was filed. It is also my opinion that a person of ordinary skill in
`the art in March 1999, upon reviewing the Katzenberg patent, would understand that the
`inventors were in possession of the claimed invention, and each of its elements.
`
`l22904.00601/35776962v.l
`
`

`

`CONFIDENTIAL — ATTORNEYS EYES ONLY
`
`I have reviewed the testimony of Gregory Ennis in his deposition on February
`8.
`21 , 2007. It is my understanding that Network-1 sought to depose Mr. Carlson before the
`due date of this report so that I could take his further testimony into account, but that D-
`Link refused to permit such a deposition. If Mr. Carlson’s deposition provides additional
`information that is relevant to the interpretation of his report and that was not available at
`the time of this report due to D-Link’s refusal to schedule the Carlson deposition, I
`reserve the right to serve a Supplemental Report addressing that information.
`
`PERSON OF ORDINARY SKILL IN THE ART
`
`9. D-Link’s new experts, Mr. Ennis and Mr. Carlson, offer a definition of a
`person of ordinary skill in the art that changes the definition agreed upon by myself and
`D-Link’s initial expert, Rich Seifert, which was a degree in Electrical Engineering (EE)
`with approximately three years of relevant industrial experience.1
`10. D-Link’s new experts assert that a person of ordinary skill also includes
`persons with Bachelors of Science (BS) degrees in physics or computer science and
`requires only two years experience.
`I do not agree.
`1 1. The field of data communications, and the Katzenberg patent, involve
`implementation of both electronic hardware and the use of microprocessors that operate
`using software. D-Link’s expert, Mr. Ennis, acknowledges that the relevant art includes
`an ability to both design and build electronic circuits, computer networking, Ethernet data
`communications, an ability to build circuits that use microprocessors and A/D converters,
`and a knowledge of power sources and power supplies. Deposition of Greg Ennis,
`February 21, 2007 (“Ennis Tr.”), 39:13 to 41 :12.
`12. A normal electrical engineering (EE) curriculum includes basic course
`requirements in both hardware and software design. Therefore, an EE graduate can be
`expected to have certain core abilities in each of these areas.
`13. The course requirements for a BS in physics or a BS in computer science, in
`the 1999—2000 time frame, required little, if any, in the way of electronics hardware
`courses, focusing instead on the atomic level of semiconductor theory (holes and
`electrons, field charges, and stochastic processes — rather than the use of semiconductor
`devices and the design of electronic circuits). The typical graduate with a physics or
`computer science credential can be expected to have far less knowledge and ability in the
`area of electronics hardware design than an EE, and far less ability to understand a patent
`disclosure involving a combination of hardware and software.
`;
`14. Mr. Ennis acknowledges that he does not know whether physics majors
`typically take courses where they actually assemble microprocessor circuits, and admits
`that a computer science curriculum does not commonly include courses in building
`electronic circuits. Ennis Tr. at 43:6 to 43:21. Thus, it seems clear that a person with
`either a physics or computer science degree would not have the educational background
`to build electronic circuits.
`
`15. Additional knowledge and understanding of core subjects is usually gained
`through work experience. However, the work assignments given to an EE graduate are
`typically different than the assignments given to computer science and physics majors.
`Because recent graduates in computer science and physics typically do not have the
`
`‘ Testimony of Rich Seifert, 8/11/2006, p. 138.
`
`122904.00601/35776962v.1
`
`

`

`CONFIDENTIAL — ATTORNEYS EYES ONLY
`
`coursework background to be competent in building complex electronic circuits, they
`typically do not receive work assignments that require these tasks during their first few
`years of work. Computer science and physics graduates would more likely be assigned to
`the software side of a project than to gain experience building electronic hardware. Thus,
`even after several years of work experience in their respective areas of competence, in
`March of 1999 a computer science or physics graduate could not be presumed to have the
`requisite knowledge to be a person of ordinary skill in the art in the technical area of the
`Katzenberg patent, which includes building electronic hardware.
`16. I note that neither Mr. Ennis nor Mr. Carlson earned a degree in Electrical
`Engineering. Thus, neither has personally experienced the EB curriculum or personally
`experienced the level of knowledge possessed by a BB with three years of relevant
`industrial experience.
`17. Even if Mr. Ennis and Mr. Carlson were correct in their effort to reduce the
`
`level of ordinary skill from that previously agreed to by myself and Mr. Seifert, a person
`of ordinary skill in March 1999 would have substantial capabilities in creating electronic
`hardware and software to implement the disclosure in the Katzenberg patent. Even Mr.
`Ennis agrees that a person of ordinary skill in this field in March, 1999 would have the
`ability to both design and build electronic circuits, knowledge of computer networking
`and particularly Ethernet data communications, an ability to build circuits that use
`microprocessors and A/D converters, and a knowledge of power sources and power
`supplies. Ennis Tr., 39:13 to 41:12.
`
`LEVEL OF DETAIL
`
`18. In Section 4.2 of his report, Mr. Ennis argues that the Katzenberg patent is
`“excessively brief’ and is therefore insufficiently enabling. At seven pages, I would
`agree that the patent is brief, but I most definitely do not agree that it is insufficiently
`detailed or contains too little information to enable one of ordinary skill in the art to
`implement the invention as described in the Katzenberg patent.
`19. Mr. Ennis compares the size of the Katzenberg patent with a number of other
`physically larger patents and implies that the relative amount of data given makes the
`validity of the Katzenberg patent questionable. This is not a logical conclusion. Even
`Mr. Ennis acknowledges that each patent must be evaluated on its on merits, and that no
`cOnclusions can be drawn merely from the length of a patent document. Ennis report at
`3 :34-35.
`
`20. I understand that to meet legal requirements, the Katzenberg patent must
`describe the invention in sufficient detail to allow a person of ordinary skill in the art in
`March 1999 to make and use the invention without undue experimentation.
`'
`21. I thus understand that the intended audience for the Katzenberg Patent
`disclosure is a person of ordinary skill in the art at the time of the patent filing, not a lay
`person. As admitted by Mr. Ennis, the knowledge and details that are conventional or
`otherwise already known to this audience as of March 1999 need not be included. Ennis
`Tr. at 167:23 to 168:9. In fact, it is preferable to provide only necessary detail, and omit
`information that is already available to the audience to avoid obscuring the actual
`invention.
`
`122904.00601/35776962v.1
`
`

`

`CONFIDENTIAL — ATTORNEYS EYES ONLY
`
`DATA SIGNALING PAIRS
`
`22. In paragraphs 39-41 of his report, Mr. Carlson argues that claims 1, 6 and 9
`are invalid because “No 10/ l 00 Ethernet data communication is possible over a single
`twisted pair.” First of all, this statement is technically incorrect. Systems such as
`Switched lOBase2 (see Exhibit 1) which perform at 10 Mbps could be considered under
`the ‘930 patent. Such systems are amenable to the PoE methodology as taught by the
`‘930 patent (admittedly with practical limitations) and permit full Ethernet
`communications over a single pair of wires. Indeed, such wiring (“thin coax”) is
`available for both 10 and 100 Mbps Ethernet, and interfaces are sold for multi-speed
`operation (10/100Base2 Ethernet.
`[See Exhibit 2, IBM Model 30H3934 10/100Base2
`Network Printer Server.]
`23. As another point, the wording of claim 1 indicates “at least one” data
`signaling pair. In the present day, and now more common, application of CAT-5 twisted
`pairs (10BaseT and 100BaseT) it is indeed true that two twisted wire pairs are used for
`Ethernet communication.
`I see no conflict with the language of Claim 1, as this does
`indeed fall within the requirement of “at least one data signaling pair...” The preferred
`embodiment of the Katzenberg patent shows a system that meets the requirements of
`these claims, i.e. uses at least one data signaling pair.
`24. As another point, claim 2 specifically indicates the use of two data signaling
`pairs. Thus, even if Mr. Carlson were correct in asserting that claim 1 is defective, claim
`2 does not have the alleged defect.
`25. The ‘930 patent suggests that other wiring is possible at col. 2 lines 44-46.
`“Cable 12 is preferably Category 5 wiring such as 100BaseX suitable for 100 Mb/s data
`communications over a switched Ethernet node.” Clearly if Category 5 wiring is
`“preferable” then other forms of wiring are also possible.
`26. In paragraph 40, Mr. Carlson makes a related argument, that “Supplying
`power for a 10/100 Ethernet system is not possible using only a single data signaling
`pa1r.”
`
`27. Mr. Carlson again ignores consideration of other than Cat-5 Base-T cabling.
`Phantom power over Ethernet coax was known at the time of the filing, as was the ability
`to block the DC saturation of the transformers with capacitors (a problem referenced by
`Mr. Seifert). This would allow for the secondary power supply to supply power from the
`data node over a single data signaling pair to the access device.
`28. Mr Carlson incorrectly assumes that the return path must necessarily be a
`second data signaling pair. While this'1s certainly the preferred embodiment, as shown'1n
`Figure 2 of the patent, nothing precludes the use of a logic ground, or a spare pair that1s
`not used for data communications, as a return path for such power.
`29. In my opinion, a person of ordinary skill in the art as of March 1999, upon
`reviewing the Katzenberg patent, would understand that the inventors were in possession
`of each element of the claimed invention, including its use of data signaling pairs.
`
`PHANTOM POWER ISSUES
`
`122904.00601/35776962v.1
`
`

`

`CONFIDENTIAL — ATTORNEYS EYES ONLY
`
`30. In Paragraphs 43-48 of his report, Mr. Carlson asserts that the Katzenberg
`patent is flawed because it does not discuss implementation issues relating to phantom
`power.
`I disagree.
`31. As has been pointed out (frequently by D-Link corporation, and within the
`“Background of the Invention,”) common implementation issues of phantom power were
`already known to those of ordinary skill in the art in March 1999. For those issues
`unique to Base-T Ethernet, Figure 2 discloses the use of center tapped transformers which
`isolate DC power from the high frequency AC communications. This same center tap
`technique also limits magnetic saturation of the transformer.
`32. In Paragraph 42, Mr. Carlson notes that power supply noise can “compromise
`data integrity of the 10/100 Ethernet link unless it is properly controlled and specified.”
`Mr. Carlson’s statement is true, regardless of whether the power is over Ethernet or
`simply through a poorly filtered and regulated power supply directly feeding the access
`device. This issue would have been most definitely understood by those skilled in the art
`in March 1999 and in no way invalidates the ‘930 patent.
`33. Mr. Carlson goes to great length to show that the much later issued IEEE
`802.3af standard, comprising several hundred pages, has more detailed implementation
`information than is contained within the ‘930 patent. The 802.3af standard is a detailed
`interoperability standard produced through many thousands of hours of labor and
`negotiation by engineers evaluating competing technologies. While it does not
`specifically constrain the components to be used, it has very specific requirements for all
`aspects of Ethernet PoE interactions between products of different manufacturers. This
`includes much detail regarding interoperability that was not required for a person of
`ordinary skill in the art in March 1999 to make and use the invention claimed in the
`Katzenberg patent. For example, a variety of mechanisms and preselected conditions
`could have been chosen and still remain within the scope of the ‘930 invention. The
`802.3af standard instead goes into great detail to define how a single method can be
`implemented in a standardized way by competing companies.
`34. It must be remembered that the 802.3af standard is just that, a standard. Many
`items covered within that standard are necessary so that the elements will work with each
`other, even from different manufacturers. This is not necessary within the ‘930 invention
`where it is necessary only that the chosen data nodes and access devices must work with
`each other. In short, it is from patents such as ‘930 that standards such as 802.3af are
`developed. While the standard has detail about achieving consistent operation between
`different manufacturer’s products, it does not necessarily provide detail about the
`construction and operation of circuits within those products, leaving specifics of internal
`design to individual manufacturers.
`35. The standard applied by Mr. Carlson in evaluating whether the Katzenberg
`patent has an enabling disclosure is not consistent with my understanding of the legal
`standard as set forth above. Mr. Carlson’s stated methodology is to compare the level of
`detail in the Katzenberg patent with the level of detail in a much later developed industry
`standard, rather than considering what a person of ordinary skill in the art already knew
`and would have understood from the disclosure in March 1999. Therefore, in my opinion
`Mr. Carlson’s analysis is based on an improper methodology and is not scientifically
`valid.
`
`122904.00601/35776962v. l
`
`
`
`

`

`CONFIDENTIAL — ATTORNEYS EYES ONLY
`
`36. Mr. Carlson also states that his analysis of enablement is based on his
`understanding that all products which practice the IEEE 802.3 af standard are within the
`scope of the patent claims. Mr. Carlson thus appears to indicate that his opinion is
`consistent with a standard other than the Court’s interpretation of the claims.
`I
`understand that enablement should be determined relative to the Court’s construction of
`
`the claims in its Markrnan ruling. Therefore, in my opinion Mr. Carlson’s methodology
`is not scientifically valid and cannot be considered reliable.
`37. In my opinion, a person of ordinary skill in the art as of March 1999, upon
`reviewing the Katzenberg patent, would understand that the inventors were in possession
`of each element of the claimed invention, including the use of phantom power.
`
`ACCESS DEVICE
`
`38. In Section 4.3 of his report, Mr. Ennis argues that the access device is
`insufficiently described. As noted before, only those features that are not already known
`to those of ordinary skill in the art need to be described. Thus, the Katzenberg patent is
`not required to provide a design guide for conventional features of Ethernet access
`devices.
`
`39. Those skilled in the art as of March 1999 would understand that the access
`
`device can be a Wireless Access Point (WAP), an Ethernet camera, a VOIP telephone, or
`other type of access device. The basic functionality of these devices is well known and
`they are commercially available. Their power requirements are also known to persons of
`ordinary skill in the art, and are provided within the manuals accompanying such devices.
`Internal operating voltage for the IC’s, necessary only if one is designing the internal
`access device power supply, could be 5 volts, or 3.3 volts, or some other value, and this
`information is readily available from the IC manufacturer’s datasheets.
`,
`40. What is required is that power be made available to a power supply within the
`access device, of quantity and value sufficient to allow that power supply to operate the
`remainder of the access device. Additionally, this power must be supplied, we are taught,
`over at least one data signaling pair.
`.
`41. Figure 2 of the Katzenberg patent, and the related text, provide sufficient
`information for a person of ordinary skill in March 1999 to implement power delivery to
`the access device. The connections to supply power from the data node are shown, as are
`the center tapped transformers and the data signaling pairs. The use of the power within
`the access device, once it has been delivered through the inventive methods, is
`conventional and already understood by those skilled in this field as of March 1999.
`42. Another feature of the invention is its ability to identify access devices that
`can accept remote power. A person of ordinary skill in this field in March 1999 would
`understand from the Katzenberg Patent that he should identify a characteristic of the class
`of access devices to which he wants to send power that distinguishes these from devices
`that should not receive power. A person of ordinary skill in March 1999 would
`understand that this distinguishing characteristic can be an existing feature of the access
`device circuits (a naturally occurring characteristic of the device electronics), or a
`characteristic added to the front end circuits of the access device to facilitate detection.
`
`This does not involve modifying the conventional operating functions of the access
`
`122904.00601/35776962v.1
`
`

`

`CONFIDENTIAL — ATTORNEYS EYES ONLY
`
`device, so again, no description of the conventional part of the access device circuitry is
`needed.
`
`43. I have reviewed the testimony of Mr. Katzenberg and Mr. Caceres with regard
`to hysteresis functions of a DC-DC switching power supply. Mr. Ennis’ citation of this
`testimony takes testimony out of context and omits important details that make his
`conclusions incorrect.
`
`44. Mr. Caceres testified that he did not participate in the development of the
`sensing functions or the selection of the power supply, and could not remember for
`certain if or how the power supply and its hysteresis functions were implemented. His
`testimony thus did not support D-Link’s claim that a special hysteresis circuit was
`implemented in their prototype. See Caceres Tr. at 266:15 to 267:17, 283:16~287:10.
`45. Mr. Katzenberg, who was responsible for the sensing function design,
`explained that hysteresis is needed to provide reliable startup operation in a DC-DC
`switching power supply and is therefore a feature of standard commercially available
`DC-DC switching power supplies. Katzenberg Tr. at 230:15-24. He testified that at least
`one commercially available power supply provided enough hysteresis, and therefore no
`special circuits were required. Katzenberg Tr. at 232214-233zl3.
`46. Mr. Katzenberg’s testimony that the DC-DC switching power supply naturally
`exhibited hysteresis is correct from a technical perspective. Commercially available DC-
`DC switching power supplies inherently provide a hysteresis response, unless specific
`circuits are added to suppress this response. The typical inherent hysteresis response is
`enough to produce a detectable sawtooth waveform as described in the Katzenberg
`Patent. This was true in March 1999 as it is today.
`47. As evidence of this fact, the D—Link PoE access devices I tested all
`incorporate DC—DC switching power supplies that have a hysteresis function, producing a
`sawtooth waveform as described in the Katzenberg Patent. Figure 1 below shows power
`being applied to the DWL-2200 access device. Power was deliberately held to a low
`level — enough to override the lockout, but low enough to prevent the access device from
`sustaining operation. The access device switching power supply clearly shows the classic
`sawtooth waveform as described in the preferred embodiment.
`
`CH1 DC ZUUUm'V' 500m: NORMAL
`
`15.24 V
`
`
`
`. .wmy—m—«um
`......
`,_WW"MM”WM W””””
`
`WM
`
`I
`
`Figure 1: DWL-2200AP Displaying Sawtooth Waveform
`
`122904.00601/35776962v.l
`
`

`

`CONFIDENTIAL — ATTORNEYS EYES ONLY
`
`48. In summary, Mr. Ennis’ conclusion that the ‘930 patent is defective for failing
`to disclose the so-called “hysteresis circuitry” is without basis. No special circuitry is
`required. A person of ordinary skill in the art in March 1999 would have been able to
`select a DC-DC switching power supply and implement detection of a sawtooth
`waveform produced by that power supply without undue experimentation, based on the
`Katzenberg disclosure.
`49. In my opinion, a person of ordinary skill in the art as of March 1999, upon
`reviewing the Katzenberg patent, would understand that the inventors were in possession
`of each element of the claimed invention, including the access device and circuits for
`delivering power to the access device as claimed.
`
`VOLTAGE SENSING
`
`50. Mr. Ennis and Mr. Carlson both argue that there is insufficient information for
`a person of ordinary skill in the art in March 1999 to determine how to “sense the
`resulting voltage” across resistors 26 and 30.
`I most strongly disagree.
`51. The two key figures for implementation of this feature in the Katzenberg
`Patent are Figure 1 and Figure 2. Figure 1 provides the basic design, which is a
`straightforward sourcing of power through the access device, with the ability to limit the
`sourced current by a single switch paralleling a resistor 26. The relationship between the
`resistors (26 and 28) and the current through them and the voltage across them is defined
`by Ohm’s Law, quite literally the first equation taught to EE students in the first electrical
`engineering class they take. It is unthinkable that any graduate EE would have any
`difficulty assigning resistor values to limit the maximum applied current to the access
`device.
`
`52. In Figure 1 the voltage level across these resistors is sensed by a
`microprocessor and A/D (Analog-to-Digital) converter. Microprocessors and A/D
`converters were known within the art much longer than Ethernet itself, certainly back into
`the early 1970’s. Motorola MC68HC11 microprocessors with 8 channels or more of A/D
`converters date from the late 80’s. The usage of the converter (how it is driven by the
`microprocessor to determine the voltage on an A/D input pin) is well documented in the
`Motorola M68HC1 1 Reference manual (M68HC1 lRM/AD). This manual includes
`detailed information, including register values.
`53. Other microprocessor and A/D converter manufacturers provide similar
`information for their IC’s, many including detailed code examples. [Example:
`Microchip Application Note, AN546: Using the Analog-to-Digital Converter; 1997. See
`Exhibit 3] Any of these processors and converters (both separate and integrated) could
`be chosen to implement the invention of the Katzenberg patent. Since such information
`was readily available to those of ordinary skill in the art in March 1999, it is my opinion
`that the lack of any vendor-specific software disclosure within the ‘930 patent in no way
`prevents one of ordinary skill in the art from successfully implementing the disclosed
`invention without undue experimentation. As of March 1999, once a person of ordinary
`skill in the art learned from the Katzenberg disclosure what operating characteristics were
`. desired, such a person would know how to implement the desired characteristics by
`programming a standard microprocessor and A/D converter.
`
`l 22904.00601/35776962v. l
`
`

`

`CONFIDENTIAL — ATTORNEYS EYES ONLY
`
`54. In his discussion of sensing methods, Mr. Carlson’s report discusses at length
`the differences between the information provided in the IEEE standard and the
`Katzenberg patent and indicates that his conclusions rely on this analysis. As noted
`before, this methodology of comparing the patent to a later-issued industry standard to
`determine enablement is improper, and leads to scientifically inaccurate conclusions.
`Therefore, Mr. Carlson’s analysis cannot be considered reliable.
`55. In my opinion, a person of ordinary skill in the art as of March 1999, upon
`reviewing the Katzenberg patent, would understand that the inventors were in possession
`of each element of the claimed invention, including methods and circuits for voltage
`sens1ng.
`
`CONTROL MEANS
`
`56. In Section 4.5 of his report, Mr. Ennis argues that the description of the
`control means is insufficient or non-enabling. Mr. Carlson makes similar arguments in
`Paragraph 58-63 of his report.
`I do not agree with their conclusions.
`57. It would have been clear to anyone skilled in the art as of March 1999 that as
`the only switch shown in the Katzenberg Patent, switch 28 (referenced sometimes as $1)
`performs the described switch functions. Microprocessors that include A/D converters
`and methods for using these devices were also within the knowledge of a person of
`ordinary skill in the relevant art in March 1999 as noted previously.
`58. Microprocessor control of either internal or external switches is a fairly basic
`EE skill, well within the ability of a person of ordinary skill in this field in March 1999.
`In a typical implementation, switching between the low level detection current and
`operating power is as simple as setting or clearing a control bit associated with switch 28
`(Figure 1) or an output port of the processor (for an external switch). As those skilled in
`the art knew in March 1999, a switch can often be controlled with a single line of code.
`For example, activating or deactivating a switch in the previously mentioned Motorola
`MC68HC11 is implemented by the single instruction “BSET Power__Control, On”.
`59. Mr. Ennis and Mr. Carlson suggest that they would expect to see an algorithm
`description, pseudocode, flowcharts, and software listings for this simple function.
`I
`disagree, the detail provided is more than adequate for a person skilled in the art to know
`how to implement the invention without undue experimentation as of March 1999.
`60. Microprocessors that include controllable switching functions, and full
`documentation for their use, were commercially available, and a person of ordinary skill
`in the art in March 1999 would have been familiar with their operation and use through
`both coursework and industry experience. Such a person would not need any details or
`description beyond that provided in the Katzenberg Patent to implement this function in
`both hardware and software in March 1999.
`
`61. Mr. Ennis also argues that Figure 1 does not show a circuit that provides
`'
`remote power. Mr. Ennis asserts that a summary description of the drawings for Figure 1
`(col. 2, lines 21-28) does not specifically mention a power feed circuit, and concludes
`that no power feed circuit is shown in Figure 1. However, a one-sentence summary
`obviously does not describe everything shown in Figure 1. The detailed description of
`Figure 1, beginning at col. 2 line 35 and continuing to col. 3 line 27, and continuing at
`
`122904.00601/35776962v.1
`
`10
`
`

`

`CONFIDENTIAL — ATTORNEYS EYES ONLY
`
`col. 3 lines 49-58, d9_e_s_ show the access device and power connections, and explains how
`the circuit in Figure 1 provides power to the access device.
`62. Mr. Ennis’ argument about switch 28 is contrary to the Court’s ruling. The
`Court clearly considered D-Link’s argument that the discussion in the specification of
`how S1 controls power to the access device does not relate to switch 28. Yet, the Court
`determined in its Markman ruling that the structure used in the Katzenberg Patent “to
`control power supplied by said secondary power source to said access device” uses
`switch 28 to perform this function.
`63. With regard to switch 28, I have previously testified that a person of ordinary
`skill in this art in March 1999 would know that the reference to switch S1 in the
`
`specification was intended to refer to switch 28. There is only one switch in the patent,
`and the description of switch S1 is completely consistent with switch 28 as shown.
`64. Mr. Ennis’ claim that the patent does not explicitly show or describe a remote
`power feed circuit is also incorrect. Figure 2 of the Katzenberg Patent shows an example
`of how a remote power source can be connected through center-tapped transformers to
`provide power via data signaling pairs to an access device.
`65. Mr. Ennis’ argument that the patent does not disclose two power sources is
`incorrect. If the Court is correct that the claims require physically separate power
`sources, Figure 3 shows an embodiment that meets this requirement. One power source
`is shown as an AC source that feeds main power supply 70 to power Ethernet switch
`cards 68. As described at col. 3 lines 28-31, this Ethernet switch card may include a
`secondary power source, such as remote power supply 34 (shown in Figure 2).
`In this
`embodiment, the AC connection of main power supply 70 and remote power supply 34
`are physically separate.
`'
`66. Mr. Carlson claims, among other things, that the ‘930 patent fails to explain
`the software which controls switch 28, or the “operator 32”. He further argues that the
`specification does not show a “logic circuit” in Figure 1. Mr. Carlson argues that the
`specification of the ‘930 patent does not disclose, suggest, or enable any “control means”
`or a way of “controlling power.” These assertions seem to be taken from D-Link’s
`infr