`U.S. Patent No. 9,098,685
`
`Filed on behalf of EMC Corporation
`Docket No.: 0100157.00269US1
`
`By: Peter M. Dichiara, Reg. No. 38,005
`Arthur Shum, Reg. No. 74,973
`WILMER CUTLER PICKERING HALE AND DORR LLP
`peter.dichiara@wilmerhale.com
`arthur.shum@wilmerhale.com
`Tel.: 617-526-6466
`Fax: 617-526-5000
`
`
`
`
`
`
`
`
`UNITED STATES PATENT AND TRADEMARK OFFICE
`
`___________________________________________
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`____________________________________________
`
`EMC CORPORATION
`Petitioner
`
`v.
`
`ACTIVIDENTITY, INC.
`Patent Owner
`
`Case IPR2017-00338
`U.S. Patent No. 9,098,685
`
`
`PETITIONER’S REPLY
`
`
`
`IPR2017-00338
`U.S. Patent No. 9,098,685
`
`TABLE OF CONTENTS
`
`
`Introduction .................................................................................................... 1
`I.
`II. Wood Anticipates the Challenged Claims ................................................... 2
`A. Wood discloses “determining a security policy . . . based on
`previously stored policy data and the received indication of
`[computing conditions]” ........................................................................ 2
`B. Wood discloses “registering the user identification data against stored
`user data” ............................................................................................. 13
`1. IV’s interpretation of “registering” is inconsistent with the claim
`and specification .................................................................................. 13
`2. Wood discloses the “registering” limitation under IV’s proposed
`construction. ........................................................................................ 16
`C. Wood discloses the limitations of claims 5 and 13 ............................. 19
`III. The Challenged Claims are Obvious over Wood in view of Neuman .... 22
`A. Neuman teaches “determining a security policy . . . based on
`previously stored policy data and the received indication” of
`computing conditions .......................................................................... 22
`It was Obvious to Combine Neuman with Wood ............................... 27
`
`B.
`
`i
`
`
`
`IPR2017-00338
`U.S. Patent No. 9,098,685
`
`I.
`
`INTRODUCTION
`IV’s Patent Owner Response (“POR”) repeats the same flawed arguments
`
`raised in its Preliminary Response (“POPR”), which the Board correctly rejected
`
`and should do so again now. IV repeats its contention that Wood (Ex-1011) and
`
`Neuman (Ex-1005) use a “fundamentally different” approach to determining a
`
`security policy than the ’685 patent (Ex-1001) — according to IV, one based solely
`
`on the information resource to be accessed, and not based on computing
`
`conditions.
`
`There is no dispute that Wood and Neuman consider the resource to be
`
`accessed in determining a security policy. But as the Board correctly explained in
`
`its Decision on Institution (“DI”), the determining step is “not exclusionary, i.e., it
`
`does not expressly preclude consideration of other parameters in addition to
`
`[computing conditions].” DI, 17; see also id., 26-27. The relevant question is
`
`whether Wood or Neuman consider the recited computing conditions claimed by
`
`the ’685 patent when determining a security policy. They plainly do. In fact,
`
`Wood and Neuman disclose identical security policies as those disclosed in the
`
`’685 patent, and they each “determine” a security policy based on the same inputs
`
`including the same recited computing conditions.
`
`IV’s arguments regarding other challenged limitations are similarly meritless
`
`and should be rejected.
`
`1
`
`
`
`IPR2017-00338
`U.S. Patent No. 9,098,685
`II. WOOD ANTICIPATES THE CHALLENGED CLAIMS
`As the Petition explains, Wood anticipates the ’685 patent. See, e.g., Pet.
`
`20-29, 33-54. IV focuses its challenges to Wood on three limitations. The Board
`
`correctly rejected IV’s arguments regarding these limitations in its institution
`
`decision, and should do so again. DI, 16-23.
`
`A. Wood discloses “determining a security policy . . . based on
`previously stored policy data and the received indication of
`[computing conditions]”
`IV attempts to paint Wood as using a “fundamentally different approach
`
`from the ’685 patent.” POR, 1. Specifically, IV claims that Wood “always
`
`determines a security policy based on the information resource being accessed.”
`
`Id., 18.1 And that “[i]t never determines a security policy based on ‘previously
`
`stored policy data’” and “the received indication of the type of communication link
`
`between the workstation and the security server, the geographic location of the
`
`workstation, or the time of access of the workstation” (id.) (hereinafter “computing
`
`conditions”).
`
`As an initial matter, the fact that Wood determines a security policy based in
`
`part on the information resource being accessed is irrelevant. As the Board
`
`recognized, the challenged claims are not exclusionary, and do not preclude the
`
`
`
` 1
`
` All emphasis added unless otherwise noted.
`
`2
`
`
`
`IPR2017-00338
`U.S. Patent No. 9,098,685
`consideration of other parameters. DI, 17. Indeed, the ’685 patent also discloses
`
`determining a security policy based on both computing conditions and the resource
`
`being accessed. See, e.g., ’685 patent, 7:23-30 (parameters include “the data being
`
`accessed…, the type of secured data being requested from the data server 19.…”);
`
`cl. 14 (security data identifies “data being accessed”); see also Goldschlag Dep.
`
`Tr., 66:8-15 (agreeing ’685 contains embodiments in which general accesses
`
`“secured data”); POPR, 8 (proposing a construction of “security policy” that
`
`includes a “request to access a resource”). Any criticism that Wood also
`
`determines a security policy based in part on the resource being accessed is thus
`
`misplaced.
`
`IV’s arguments that Wood uses a fundamentally different approach than the
`
`’685 patent, one which never determines a policy based on stored policy data and
`
`the recited computing conditions, are flatly inconsistent with Wood’s disclosures.
`
`As the Petition explains, Wood discloses the same examples of security policies,
`
`determined based on the same examples of stored policy data and the same
`
`computing conditions. Pet. 7, 15, 29, 35, 37. For example:
`
`3
`
`
`
`IPR2017-00338
`U.S. Patent No. 9,098,685
`Time: The ’685 patent discusses a security policy2 where attempts to access
`
`a resource between midnight and 6am are automatically denied. ’685 patent, 7:55-
`
`58 (see Pet. 7, 15, 29, 37). In this example, the time of a request (i.e., a recited
`
`computing condition) is compared against the policy-specified period of midnight
`
`to 6 am (i.e., as expressed by “previously stored policy data” 3). If the request falls
`
`within that period, a predetermined security policy applies which denies the
`
`request for access. If the request occurs at other times, a different predetermined
`
`security policy will apply. Ex-1040 (Goldschlag Dep. Tr.), 62:17-63:15.
`
`As the Petition explains, Wood discloses a substantively identical approach
`
`(Pet. 29, 37):
`
`
` 2
`
` In its POPR, IV argued “security policy” should mean “one or more rules to
`
`apply to a security relevant event such as a request to access a resource.” POPR,
`
`8-9. Dr. Goldschlag agreed IV’s proposed construction was within the broadest
`
`reasonable interpretation (BRI) of this term, Goldschlag Dep. Tr., 248:10-250:18,
`
`and agreed this example was properly within the BRI of “security policy.” Id.,
`
`35:14-36:16.
`
`3 As the Board recognized, “[p]olicy data may include multiple parameters” such
`
`as “the time, the day of the week.” DI, 18. See also Pet. 38-39; Neuman Decl.,
`
`¶88; POR, 4.
`
`4
`
`
`
`’685 Patent
`
`IPR2017-00338
`U.S. Patent No. 9,098,685
`Wood
`
`“For example, in accordance with the
`
`“For example, a given security policy
`
`security policy no access is to be
`
`and associated trust level mappings
`
`provided between midnight and 6:00
`
`may dictate a REFUSE response in
`
`am, the user requesting an access
`
`response to an access request to a
`
`during this period of time is
`
`sensitive resource (such as financial
`
`automatically denied access.”
`
`data) by any client entity…if the
`
`’685 patent, 7:55-58.
`
`
`
`access request is received over a dial-
`
`up line and originates from an
`
`unknown number or is received outside
`
`of working hours.” Wood, 9:59-67.
`
`Wood, like the ’685 patent, expressly discloses comparing the time of an
`
`access request (i.e., a recited computing condition) against a specified time period
`
`5
`
`
`
`IPR2017-00338
`U.S. Patent No. 9,098,685
`(i.e., “outside of working hours,” which is “previously stored policy data”4). If the
`
`request falls within that period, Wood, like the ’685 patent, applies a
`
`predetermined security policy which denies (REFUSE) the request for access. If
`
`the request occurs at other times, Wood, like the ’685 patent, will apply another
`
`predetermined security policy to the request.
`
`For example, Wood discloses a security policy that authorizes access based
`
`on trust levels. Wood explains that, if the current trust level of an access request
`
`meets or exceeds the required trust level for accessing the requested resource, then
`
`access is “ALLOWED” without further authentication needed. Wood, 9:50-52.
`
`Conversely, if the current trust level is too low, then the request may be
`
`“REDIRECTED” in which case Wood determines (e.g., selects) authentication
`
`schemes based on the user’s computing conditions or environment (e.g., time of
`
`access, location, and network) to permit the user to escalate its current trust level to
`
`
`
` 4
`
` A POSITA would understand that Wood’s security architecture includes pre-
`
`stored data, such as algorithms, parameters, and other data for implementing its
`
`process for determining a security policy. Pet. 38-39; Ex-1002 (Neuman Decl.),
`
`¶88; DI, 17-19.
`
`6
`
`
`
`IPR2017-00338
`U.S. Patent No. 9,098,685
`meet the required trust level.5 See, e.g., Wood, 10:13-21; Neuman Decl., ¶¶ 59-60;
`
`POR, 10-14; Ex-2007 (Goldschlag Decl.), ¶¶ 44-48; Goldschlag Dep. Tr., 144:17-
`
`145:13.
`
` IV argues that Wood determines a security policy based only on the
`
`information resource being accessed. POR, 18-21. According to IV, if the
`
`resource being accessed is not very sensitive, like archived data, Wood assigns a
`
`security policy that imposes only a low required trust level. See id., 19. If the
`
`resource is sensitive, like financial data, Wood applies a different security policy
`
`that requires a more substantial trust level. See id.
`
`
`
` 5
`
` This functionality in Wood is not disputed by IV. See, e.g., POR, 26-27. During
`
`his cross-examination, Dr. Goldschlag repeatedly characterized these passages in
`
`Wood as discussing a “process” for governing access to a resource, e.g., ALLOW
`
`or REFUSE access. See, e.g., Goldschlag Dep. Tr., 153:16-20; 155:7-10; 157:21-
`
`158:2; 162:3-14; 181:17-182:4. Ordinary dictionary definitions confirm that a
`
`“policy” is a “method of action”, i.e., process. See, e.g., Ex-1042 (Merriam-
`
`Webster Dictionary) (defining “policy” as, among other things, “a definite course
`
`or method of action selected from among alternatives and in light of given
`
`conditions to guide and determine present and future decisions.”).
`
`7
`
`
`
`IPR2017-00338
`U.S. Patent No. 9,098,685
`IV’s argument that the policy is determined solely on the resource is flatly
`
`inconsistent with both Wood’s disclosures and the ‘685 embodiments. Both the
`
`’685 patent and Wood determine a security policy based on the resource and on
`
`the computing condition of “time of access,” as well as on “previously stored
`
`policy data” that specifies the relevant time period for the policy. As Dr. Neuman
`
`explains, IV is wrong to suggest that Wood uses a fundamentally different
`
`approach than the ’685 patent; instead, he reiterates that Wood teaches
`
`“remarkably similar” policies, as its description of denying access “outside of
`
`working hours” functions the same way as the ’685 patent’s example of denying
`
`access between midnight and 6am. Neuman Decl. ¶67; Ex-1040 (Neuman Reply
`
`Decl.), ¶12.
`
`Location: The ’685 patent also describes security policies that are
`
`determined based on the location from which a request is made. See, e.g., ’685
`
`patent, 9:16-37 (discussed at Pet. 29); 9:55-62 (discussed at Pet. 7). Dr.
`
`Goldschlag focused on these location-based policies in his testimony, and created
`
`an example of a “security policy,” reproduced below, which is virtually identical to
`
`another Wood example (Pet. 29, 35, 37):
`
`
`
`8
`
`
`
`Goldschlag Dep., Ex-1039
`Goldschlag Example
`
`Wood
`
`IPR2017-00338
`U.S. Patent No. 9,098,685
`
`“Access is allowed from work by
`
`“In an exemplary enterprise tool
`
`employees”
`
`
`
`configuration, a desired security policy
`
`may dictate that a salary tool is
`
`Goldschlag Dep., Ex-1039; see also id.
`
`accessible only from with[in] a
`
`Tr., 29:3-22.
`
`
`
`
`
`company’s internal network. No level
`
`of authenticated trust may be sufficient
`
`to access such a tool from outside
`
`company network. To facilitate
`
`implementation of such a security
`
`policy, authorization component 140
`
`could refuse access based on
`
`environment parameters indicating a
`
`session originating outside the
`
`company’s internal network.” Wood,
`
`6:20-28.
`
`Wood expressly discloses a predetermined security policy determined by
`
`comparing the location of an access request (i.e., a computing condition) against
`
`9
`
`
`
`IPR2017-00338
`U.S. Patent No. 9,098,685
`specified location criteria (i.e., “within a company’s internal network,” expressed
`
`with “previously stored policy data”). If the request falls within that specified
`
`location, the policy permits access. This first security policy may impose
`
`additional requirements before access is permitted, e.g., by comparing a current
`
`trust level of the access request against a required trust level, as explained above.
`
`See Wood, 2:28-48; 9:26-46; 11:30-33; 14:8-25; POR, 10-14; Goldschlag Decl.,
`
`¶¶44-48. If the request occurs from outside the company network, a different
`
`security policy governs the request, which refuses access and in which “no level of
`
`authenticated trust may be sufficient.”
`
` Faced with this clear evidence about Wood, IV attempts to contort Wood’s
`
`use of trust levels. See generally POR, 18-27. IV contends that Wood’s security
`
`policies are based only on the minimum required trust level of the requested
`
`resource, id., 18, and that Wood uses “environment information” (i.e., computing
`
`conditions such as time, location, or type of communication link) only to determine
`
`the current trust level and not the required trust level. Id., 22-24, 26. Both
`
`arguments are demonstrably false.
`
`First, Wood’s location-based policy (like its time-based and other policies)
`
`is based on the information resource being accessed (i.e., a salary tool) and the
`
`computing condition of location (i.e., within a company’s internal network). Wood
`
`further explicitly states that “authorization component 140 may be programmed to
`
`10
`
`
`
`IPR2017-00338
`U.S. Patent No. 9,098,685
`perform some stringent testing beyond a trust level requirement” and “[n]o level
`
`of authenticated trust may be sufficient to access such a [salary] tool from outside
`
`company network.” Wood, 6:18-28. Accordingly, Wood’s location-based policy
`
`expressly refuses access regardless of trust levels, contradicting IV’s contention
`
`that Wood’s security policies are based only on the minimum required trust level
`
`of the requested resource. Neuman Reply Decl., ¶18-20.
`
`Second, Wood expressly teaches that environment information can be used
`
`to determine both the current trust level and the required trust level. Neuman
`
`Decl., ¶¶ 59-60, 87, 89, 92; Neuman Reply Decl., ¶21. Wood teaches that the
`
`required trust level (in addition to the current trust level) may vary based on
`
`computing conditions such as time, location, and type of communication link.6 As
`
`the Petition explains, Wood discloses that the required trust level for a resource
`
`may be determined “in the context of the requesting session environment.” Wood,
`
`
` 6
`
` Dr. Goldschlag agrees that Wood’s required trust levels are associated with
`
`security policies. See, e.g., Goldschlag Dep. Tr., 165:10-12 (“the required trust
`
`level…is the policy governing access to the resource”); see also 149:19-20;
`
`151:15-17; 159:10-11. This is consistent with the Petition’s analysis that changing
`
`the required trust level is another instance of determining security policy based on
`
`computing conditions. Pet. 37-39.
`
`11
`
`
`
`IPR2017-00338
`U.S. Patent No. 9,098,685
`14:48-51; Pet. 37-38; see also Wood, Abstract (“In some configurations, changing
`
`environmental parameters may cause a previously sufficient credential to become
`
`insufficient.”). Thus, Wood teaches that an information resource may have one
`
`required trust level under one set of computing conditions (i.e., environment) and a
`
`different required trust level under another set of computing conditions (i.e., a
`
`different environment). Therefore, when Wood teaches that a required trust level
`
`for accessing a resource may vary based on a “requesting session environment”,
`
`Wood again teaches determining a security policy based on computing conditions.
`
`Faced with this, IV makes the untenable claim that the term “environment”
`
`as used in column 14, lines 48-51 of Wood means something entirely different than
`
`what that term means throughout the rest of Wood (and the ’685 patent, for that
`
`matter). IV argues “environment” refers “to whether a session token is included in
`
`an information resource access request.” POR, 22-23. This strained interpretation
`
`of Wood should be rejected. To make this argument, IV attempts to rewrite
`
`Wood’s reference of “session environment” to “session token.” But Wood
`
`repeatedly uses the word “environment” to refer to the same things that the ’685
`
`patent calls “computing conditions,” and which the ’685 patent also calls
`
`“environment.” See, e.g., Wood, Abstract (“environmental parameters (e.g.,
`
`connection type or location)”); 2:66-67 (same); 6:32-34 (“environment
`
`information such as time of request, source of request, connection speed…”); 7:38-
`
`12
`
`
`
`IPR2017-00338
`U.S. Patent No. 9,098,685
`8:22 (listing “connect location”, “temporal information such as request time/date”,
`
`“origination point (e.g., inside or outside of a corporate network)” as examples of a
`
`“requesting client’s environment information.”); Goldschlag Dep. Tr., 109:14-
`
`110:14 (Dr. Goldschlag agreeing the ’685 patent also uses the word “environment”
`
`to refer to a received indication of a type of communication link, time of access,
`
`and geographical location.)
`
`In short, Wood and the ’685 patent disclose examples of “security policies”
`
`that are not only substantively identical, but are also determined based on the same
`
`“computing conditions” and “previously stored policy data.”
`
`B. Wood discloses “registering the user identification data against
`stored user data”
`IV premises its arguments that Wood fails to disclose the “registering”
`
`limitation on a strained, unsupportable reading of this limitation. The Board
`
`considered IV’s argument (see, e.g., POPR, 27-31), but did not construe the term
`
`(see, e.g., DI, 9; 19-20), consistent with the notion that “registering” does not carry
`
`any special meaning. Even so, the Board correctly recognized that Wood discloses
`
`this limitation even under IV’s proposed construction. DI, 20.
`
`1.
`
`IV’s interpretation of “registering” is inconsistent with the
`claim and specification
`IV does not dispute that Wood discloses comparing received user
`
`identification data (i.e. login credentials) against stored user data (i.e. pre-stored
`
`13
`
`
`
`IPR2017-00338
`U.S. Patent No. 9,098,685
`login credentials). POR, 27-37. Likewise, IV cannot seriously dispute that, during
`
`such authentication, Wood stores (i.e., registers) in its computers the received user
`
`identification (e.g., the received fingerprint or password) so that it may be
`
`compared against the pre-stored login credentials (i.e., “authorized” fingerprint
`
`data or password).7 See Goldschlag Dep. Tr., 125:6-14 (“[Q]: …with
`
`authorization, …the person presents some information, some claim, a fingerprint, a
`
`password, or whatever. And it’s compared against something to prove whether the
`
`person is who he or she claims to be, right? …[A]: Yes, that’s right…”); see also
`
`id., 128:11-16.
`
`There accordingly should be no dispute that Wood satisfies the registering
`
`step. For example, the fact that the recited registering operation is done “against
`
`previously stored user data” indicates that the BRI of the recited registering
`
`encompasses storing and comparing of “user data” against the “previously stored
`
`user data” that occurs in Wood’s authentication steps. As the Petition and Dr.
`
`
` 7
`
` IV and Dr. Goldschlag are wrong to suggest that some “authentication systems,
`
`do not generate a record of the received user identification data.” POR, 30;
`
`Goldschlag Decl., ¶88. Instead, as Dr. Neuman explains, the process of
`
`authenticating received login credentials would require “making a record” of the
`
`received login credentials so they may be verified. Neuman Reply Decl., ¶¶25-28.
`
`14
`
`
`
`IPR2017-00338
`U.S. Patent No. 9,098,685
`Neuman explain, Wood teaches that once a login credential has been supplied for
`
`authentication, that login credential is stored and compared against pre-stored login
`
`credentials. See, e.g., Pet. 44-45; Neuman Decl., ¶¶ 96-97; Wood, 12:13-15;
`
`12:32-36; Neuman Reply Decl., ¶29. If the received login credential matches with
`
`a pre-stored login credential, the former is authenticated, authorizing the access
`
`request.
`
`IV attempts to distort this plain meaning of registering by arguing that it
`
`means “make a record of the received user identification data,” and by demanding
`
`that this “registering” step must be distinct and separate from an “authentication”
`
`step and must precede authentication. POR, 28-31. There is no basis for this
`
`reading.
`
`First, IV’s interpretation blatantly attempts to read limitations into the claim.
`
`Nothing in the claim term “registering…against” demands an exclusion of the
`
`registering performed during authentication. Likewise, nothing in the claim term
`
`requires that this step must not only be separate and distinct from authentication
`
`but must precede it.
`
`IV cites a passage in the ’685 patent stating that “the security server
`
`performs an operation of registering the user data against previously stored user
`
`data in accordance with the determined at least an authorization method.
`
`Thereafter, the security server identifies the user and optionally authorizes the user
`
`15
`
`
`
`IPR2017-00338
`U.S. Patent No. 9,098,685
`to access secured data….” POR, 30-31 (citing 5:58-64). This passage simply
`
`parrots the claim language, falls far short of lexicography, and in no way demands
`
`the restrictions that IV injects into its interpretation. IV also points to the word
`
`“[t]hereafter,” POR, 31, but read in context, the passage supports Petitioner. It
`
`simply confirms that, as a consequence of the registering/authentication operation,
`
`the system can now properly identify the user (i.e., the purpose of authentication),
`
`and authorize the access if appropriate (i.e., optionally). In fact, this operation is
`
`depicted as a single step in FIG. 4 of the patent. This passage thus does not
`
`support IV’s argument.
`
`Second, IV does not even attempt to explain how one can “store” something
`
`“against” something else, further confirming its interpretation is wrong.
`
`Questioned about this point, Dr. Goldschlag testified that “registering user
`
`identification data against stored user data” means storing user identification data
`
`in a “database”—rewriting “against” to “in” and demanding “a database” even
`
`though he was unable to specifically identify such a database in the ’685 patent.
`
`Goldschlag Dep. Tr., 131:7-133:19; 136:10-15.
`
`Accordingly, IV’s interpretation of “registering” is unsupported, and
`
`improperly reads in limitations where none exist.
`
`2. Wood discloses the “registering” limitation under IV’s
`proposed construction.
`
`16
`
`
`
`IPR2017-00338
`U.S. Patent No. 9,098,685
`Moreover, Wood discloses this limitation even under IV’s proposed
`
`interpretation through its session credential structure and session token. See, e.g.,
`
`DI, 20.
`
`IV argues that Wood does not make such a record because Wood’s “session
`
`credential structure 420” and “session token 430” (illustrated in FIG. 4 of Wood,
`
`below) do not include the specific login credentials supplied by the user, such as
`
`received biometric data. POR, 33-35. However, the very figure that IV cites at
`
`page 34 of its POR (Wood at FIG. 4) shows that a “SessionCookie”, which IV
`
`does not dispute is “stored” by Wood’s system, includes a “principalID”.
`
`17
`
`
`
`IPR2017-00338
`U.S. Patent No. 9,098,685
`
`
`
`Wood, FIG. 4 (Annotated)
`As the Board previously recognized, there can be no dispute that
`
`“principalID” comprises “user identification data.” See, e.g., Wood, 19:51-53
`
`(“[p]rincipal id encodes an identifier for a principal to which the security
`
`architecture has resolved login credentials.”); (Ex. 2006) Neuman Dep. Tr., 77:20-
`
`25 (“Principal, that’s the identity, the entity that’s been authenticated….”); DI,
`
`20 (“Wood also explains that a principal identification (ID) may be encoded
`
`(stored) in a secured session credential.”). The principalID field includes data that
`
`18
`
`
`
`IPR2017-00338
`U.S. Patent No. 9,098,685
`is “received” as part of a user’s login credentials. See Wood, 18:40-42; 19:53-58.
`
`Principal ID is therefore an instance of “received user identification data,” which
`
`IV admits is stored as part of a record.
`
`Accordingly, as the Board recognized, Wood discloses the “registering”
`
`limitation even under IV’s flawed reading. See DI, 19-20.
`
`C. Wood discloses the limitations of claims 5 and 13
`Claims 5 and 13 require that “the security data further comprises data
`
`relating to available user data input devices for providing input to the workstation.”
`
`As the Petition explains, Wood discloses these limitations when it states that
`
`“capabilities of a particular client entity (e.g., … availability of a fingerprint
`
`scanner or card reader) may affect the availability or sufficiency of particular
`
`authentication schemes to achieve a desired trust level.” Pet. 52-53; Wood,
`
`Abstract; see also id., 2:67-3:4. Here, Wood plainly states it considers the
`
`“availability” of “user input devices” (e.g., fingerprint scanner or card reader) to
`
`determine whether certain authentication schemes are available and “sufficient” to
`
`satisfy a required trust level for a given access. Clearly, this information needs to
`
`be received as security data for Wood’s server to determine if a scheme is available
`
`or sufficient. Neuman Reply Decl., ¶35.
`
`IV complains that Wood’s disclosures are “vague” and “do not positively
`
`recite the usage of data relating to available user data input devices to ‘determine’
`
`19
`
`
`
`IPR2017-00338
`U.S. Patent No. 9,098,685
`an authorization method.” POR, 38-39. However, Wood need not “positively
`
`recite” a feature to disclose it. Instead, the issue is whether “a person of skill in the
`
`art, reading the reference, would ‘at once envisage’ the claimed arrangement or
`
`combination.” Microsoft Corp. v. Biscotti Inc., 878 F.3d 1052, 1068 (Fed. Cir.
`
`2017). See also In re Preda, 401 F.2d 825, 826 (CCPA 1968) (“[I]n considering the
`
`disclosure of a reference, it is proper to take into account not only specific
`
`teachings of the reference but also the inferences which one skilled in the art would
`
`reasonably be expected to draw therefrom.”).
`
`IV also attempts to distort Wood’s plain disclosure by arguing that “[i]f the
`
`client entity is not equipped with a biometric input device…, Wood’s system does
`
`not alter and/or select a different authentication scheme.” POR, 39. This is a clear
`
`misreading of Wood. Wood is explicit that it considers the availability of user
`
`devices to determine the sufficiency of an authentication scheme; in other words, it
`
`will only choose an authentication scheme that is available to the user and which
`
`would attain a “sufficient” current trust level to be permitted access to the resource.
`
`Neuman Reply Decl., ¶36.
`
`IV and Dr. Goldschlag further argue that “Wood offers the usage of multiple
`
`authentication schemes in the hopes that one or more of the schemes will be
`
`sufficient.” Id. Wood does not discuss such a (mere-hope) scheme anywhere in its
`
`specification. And as Dr. Goldschlag admitted in his deposition, it would have
`
`20
`
`
`
`IPR2017-00338
`U.S. Patent No. 9,098,685
`been unreasonable to interpret Wood to request a specific form of authentication
`
`(e.g., a password, biometric data, etc.) if it did not have stored credential
`
`information for the corresponding scheme. Goldschlag Dep. Tr., 227:4-228:10. A
`
`POSITA would therefore not have interpreted Wood to offer an authentication
`
`scheme merely in the “hope” that this scheme would be available and sufficient.
`
`Dr. Neuman explains that, contrary to IV’s unreasonable interpretation,
`
`Wood’s Abstract expressly teaches that its security server receives data relating to
`
`available user data input devices. First, Wood’s Abstract states that information
`
`regarding the availability of a fingerprint scanner or card reader affects the
`
`“availability or sufficiency of particular authentication schemes to achieve a
`
`desired trust level.” Wood’s security server thus decides what authentication
`
`schemes would be “available” and/or “sufficient” based on the “availability” of a
`
`fingerprint scanner or card reader. To make decisions based on this information,
`
`Wood’s security server must receive this information as data. And as the Petition
`
`explains for related independent limitation [1b], this information is received by the
`
`“entry handler functionality” in Wood’s “gatekeeper / entry handler component
`
`110,” which is responsible for receiving “environment information” including
`
`“client type and/or capabilities” and “system type.” Wood 7:47-49, 8:10-15, Pet.
`
`36. This corresponds to the “receiv[ed] security data” recited in claims 1, 5, and
`
`13. Neuman Reply Decl., ¶38. Second, as a matter of logic, the fact that Wood’s
`
`21
`
`
`
`IPR2017-00338
`U.S. Patent No. 9,098,685
`security server is offering a type of authentication scheme (e.g., fingerprints)
`
`means that Wood’s security server has a copy of data related to the acceptable
`
`fingerprint on file. And if Wood’s security server has an “official” fingerprint on
`
`file, this means the user had a fingerprint scanner available.
`
`III. THE CHALLENGED CLAIMS ARE OBVIOUS OVER WOOD IN
`VIEW OF NEUMAN
`IV raises similar complaints with Neuman as with Wood, arguing that
`
`Neuman selects a security policy based on the information resource accessed, not
`
`based on previously stored policy data and computing conditions. IV also claims
`
`that Petitioner has not shown sufficient motivation to combine Neuman and Wood.
`
`Again, the Board correctly rejected these arguments in its institution decision, and
`
`should do so again. DI, 26-27. Indeed, Neuman is abundantly explicit about
`
`determining security policies based on computing conditions, both disclosing a
`
`computer language and framework for doing so, and detailed examples.
`
`A. Neuman teaches “determining a security policy . . . based on
`previously stored policy data and the received indication” of
`computing conditions
`As the Petition explains and the Board recognized, Neuman is expressly
`
`directed at an access control framework capable of implementing multiple security
`
`policies. Pet. 31-33, 57; DI, 11. Neuman further explicitly discloses time-,
`
`location-, and connection-based security policies governing access to a resource.
`
`22
`
`
`
`IPR2017-00338
`U.S. Patent No. 9,098,685
`That Neuman also considers the information resource in determining a security
`
`policy is irrelevant. See DI, 17. All of IV’s arguments accordingly