VMWARE, INC.

Patent Trials and Appeals Board

Nov 10, 2021

2020004830 (P.T.A.B. Nov. 10, 2021)

UNITED STATES PATENT AND TRADEMARK OFFICE
UNITED STATES DEPARTMENT OF COMMERCE
United States Patent and Trademark Office
Address: COMMISSIONER FOR PATENTS
P.O. Box 1450
Alexandria, Virginia 22313-1450
www.uspto.gov
APPLICATION NO. FILING DATE FIRST NAMED INVENTOR ATTORNEY DOCKET NO. CONFIRMATION NO.
13/870,878 04/25/2013 Jonathan Clark B017 2489
151795 7590 11/10/2021
FISH & RICHARDSON P.C. (VMware)
P.O. BOX 1022
MINNEAPOLIS, MN 55440-1022
EXAMINER
HESS, MICHAEL J
ART UNIT PAPER NUMBER
2481
NOTIFICATION DATE DELIVERY MODE
11/10/2021 ELECTRONIC
Please find below and/or attached an Office communication concerning this application or proceeding.
The time period for reply, if any, is set in the attached communication.
Notice of the Office communication was sent electronically on above-indicated "Notification Date" to the
following e-mail address(es):
PATDOCTC@fr.com
ipadmin@vmware.com
PTOL-90A (Rev. 04/07)
UNITED STATES PATENT AND TRADEMARK OFFICE
BEFORE THE PATENT TRIAL AND APPEAL BOARD
Ex parte JONATHAN CLARK

Appeal 2020-004830
Application 13/870,878
Technology Center 2400

Before JEAN R. HOMERE, TREVOR M. JEFFERSON, and
MICHAEL J. STRAUSS, Administrative Patent Judges.
HOMERE, Administrative Patent Judge.

DECISION ON APPEAL
I. STATEMENT OF THE CASE1
Pursuant to 35 U.S.C. § 134(a), Appellant2 appeals from the
Examiner’s decision to reject claims 1–3, 5, 7–10, 12, 14–17, 19, 21, and
22.3 Appeal Br. 10. We have jurisdiction under 35 U.S.C. § 6(b).
We reverse.

1 We refer to the Specification, filed Apr. 25, 2013 (“Spec.”); Final Office
Action, mailed July 16, 2019 (“Final Act.”); Appeal Brief, filed Feb. 12,
2020 (“Appeal Br.”); Examiner’s Answer, mailed Apr. 16, 2020 (“Ans.”);
and Reply Brief, filed June 11, 2020 (“Reply Br.”).
2 “Appellant” refers to “Applicant” as defined in 37 C.F.R. § 1.42(a).
Appellant identifies the real party in interest as VMware, Inc. Appeal Br. 1.
3 Claims 21 and 22 have been withdrawn from consideration as being
directed to a non-elected invention. Final Act. 17–18; Ans. 12.
Appeal 2020-004830
Application 13/870,878
2
II. CLAIMED SUBJECT MATTER
The claimed subject matter relates to a method and system for
compressing video data using an image block matching technique in virtual
desktop infrastructure (VDI) deployments by pre-calculating hash values for
a reference frame prior to other frames becoming available based on changes
to a graphical user interface (GUI) thereby reducing latency in compressing
videos. Spec. ¶¶ 5, 10. Appellant describes the claimed subject matter as
follows:
Systems and methods described herein compress video
data by using image block matching that utilizes a full screen
exact block search and pre-calculated block hashes in a latest
frame of a video to reduce latency. A server for remoting a
display to a remote client is provided, the server includes a
processor that is configured to access a previous frame of an
image in a video and to obtain hash values for each pixel in the
previous frame. The processor is also configured to create a
hash map that stores each of the hash values for each pixel in
the previous frame. The processor is configured to receive a
current frame of the image in the video and to separate the
current frame into a plurality of current blocks of pixels. The
processor is also configured to calculate, using a hash function,
a hash value for each of the plurality of current blocks of pixels.
The processor is further configured to compare the hash values
in the hash map with the hash values associated with the current
frame and to identify a hash value in the hash map that matches
a hash value in the current frame. Moreover, the processor is
configured to compress the current frame of the video for
transmission to the remote client using the identified matching
hash values and to pre-calculate a new hash map based on the
current frame for use in compressing a next frame of the video.
Id. ¶ 5.
Figure 1, reproduced below, is useful for understanding the claimed
subject matter:
Appeal 2020-004830
Application 13/870,878
3

Figure 1 depicts GUI 119 for desktop 117 generated by a virtual machine
hosted on server computer 110, which generates, updates and compresses a
video from the GUI contents based on image block matching and
subsequently transmits them to client computer 164. Id. ¶¶ 10, 11.

Appeal 2020-004830
Application 13/870,878
4
Independent claim 1, with disputed limitations emphasized, is
illustrative:
1. A server for remoting a display to a remote client, the
server comprising a processor configured to:
pre-calculate a first hash map for a previous frame of a
video that represents a graphical user interface of a desktop
prior to a current frame of the video becoming available,
wherein the current frame becomes available based on an
update to the graphical user interface of the desktop, the first
hash map comprising pre-calculated hash values stored for each
of a plurality of pixel blocks at each of a plurality of pixel
offsets in the previous frame, each pixel block including a
plurality of pixels relative to the corresponding pixel offset, and
wherein the previous frame depicts a view of the graphical user
interface;
receive a current frame of the video, wherein the current
frame is a sequentially next frame in the video from the
previous frame and depicts an updated view of the graphical
user interface that is updated from the previous frame based on
the update to the graphical user interface;
separate the current frame into a plurality of current
blocks of pixels at each of the plurality of pixel offsets of the
current frame;
calculate, using a hash function, a hash value for each of
the plurality of current blocks of pixels;
compare the previous frame to current frame of the video
including comparing the pre-calculated hash values in the first
hash map for the previous frame with the calculated hash values
for the current blocks of pixels at each corresponding pixel
offset of the current frame;
identify a hash value in the first hash map at a particular
pixel offset that matches a hash value in the current frame at a
corresponding pixel offset;
compress the current frame of the video for transmission
to the remote client using the identified matching hash values;
pre-calculate a second hash map for the current frame of
the video prior to a next current frame of the video that depicts
a next updated view of the graphical user interface becoming
Appeal 2020-004830
Application 13/870,878
5
available based on another update to the graphical user
interface, the second hash map comprising the calculated hash
values for each of the plurality of current blocks of pixels; and
replacing the first hash map with the second hash map
such that hash values calculated for the next current frame are
compared to the hash values calculated for the current blocks
of pixels of the current frame and such that hash values
calculated for the next current frame are not compared to hash
values of previous frames of the video when comparing the next
current frame to the current frame and compressing the next
current frame based on the comparing, the replacing the first
hash map with the second hash map including replacing the
pre-calculated hash values of the first hash map at each pixel
offset with the hash values for each block of pixels in the
current frame concurrently with comparing the pre-calculated
hash value at that pixel offset with the calculated hash values
such that the pre-calculated hash values are replaced for each
pixel offset after each respective comparison.

Appeal Br. 12–13 (emphasis added).
III. REFERENCES
The Examiner relies upon the following references as evidence.4
Name Reference Date
Burgess US 2004/0131115 A1 July 8, 2004
Carlson US 2012/0286979 A1 Nov. 15, 2012
IV. REJECTION5
The Examiner rejects claims 1–3, 5, 7–10, 12, 14–17, and 19 under 35
U.S.C. § 103 as unpatentable over the combined teachings of Burgess and
Carlson. Final Act. 8–17.

4 All reference citations are to the first named inventor only.
5 The Examiner withdrew the 112 rejections previously entered against
claims 1, 8, and 15. Advisory Act.
Appeal 2020-004830
Application 13/870,878
6
V. ANALYSIS
We consider Appellant’s arguments, as they are presented in the
Appeal Brief, pages 5–10 and the Reply Brief, pages 2–5.6
a. Examiner’s Election/Restriction
As a preliminary matter, Appellant argues that the Examiner’s
withdrawal of claims 21 and 22 from consideration is premised upon an
improper election/restriction under 37 C.F.R. § 1.145 (MPEP § 821.03).
Appeal Br. 5–6. Therefore, Appellant requests reversal of the restriction to
claims 21 and 22. Id. at 6. Appellant is reminded that the issue of claim
restriction is a petitionable, not an appealable matter and is, therefore, not
properly raised before us. See MPEP § 1201 (“The Board will not ordinarily
hear a question that should be decided by the Director on petition . . . .”). See
37 C.F.R. § 1.181. Accordingly, we do not reach the issue regarding the
Examiner’s election/restriction of claims 21 and 22.
b. Obviousness Rejection
Appellant argues that the Examiner errs in finding that the
combination of Burgess and Carlson renders claim 1 obvious. Appeal Br. 8–
9. In particular, Appellant argues that the cited combination of references
does not teach or suggest the following limitation:
[R]eplacing the first hash map with the second hash map
such that hash values calculated for the next current frame are
compared to the hash values calculated for the current blocks of
pixels of the current frame and such that hash values calculated
for the next current frame are not compared to hash values of

6 We have considered in this Decision only those arguments Appellant
actually raised in the Briefs. Arguments not made are forfeited. See
37 C.F.R. § 41.37(c)(1)(iv) (2017).
Appeal 2020-004830
Application 13/870,878
7
previous frames of the video when comparing the next current
frame to the current frame

as recited in independent claim 1. Id. According to Appellant, Burgess
discloses comparing “block of pixels” to all of the least oldest “block of
pixels” that can be stored in a finite-sized cache, as opposed to comparing
hash values between just two adjacent frames. Id. (citing Burgess ¶ 97).
More particularly, Appellant states the following:
The claimed replacing and comparing ensures that each hash
value of the next current frame is compared to the hash values
of the current frame, without either wasting time and processing
power comparing the hash values of the next current frame to
various other frames or missing some of the hash values of the
current frame due to a limited cache size, which would be the
case using Burgess’s technique of overwriting of older data
based merely on age.
Id. at 9.
In response, the Examiner finds the following:
Appellant contends, “the claimed comparison is a comparison
between hash values of two frames only,” rather than
comparing to blocks of pixels from older frames as taught by
Burgess. Examiner finds Appellant’s focus on just two frames
obstructs the issue because there will be a third frame, and a
fourth frame, and so on. Using Appellant’s argument as the
premise, when the third frame is the current frame, the third
frame compares back to the second frame. Now, pause here to
consider what that second frame actually represents. The
second frame is an amalgam of the comparison between the
first and second frames. As taught by Burgess’s “cache miss”,
the second frame (second hash map) replaces the first frame
(first hash map) only to the extent the second frame has blocks
differing from the first frame. Those blocks from the first frame
that are also found in the second frame, and subsequent frames
(Burgess’s “cache hits”), persist into the next current frame
(and on and on). So, when considering a comparison between a
Appeal 2020-004830
Application 13/870,878
8
second frame and a third frame and when considering the third
frame will replace the second frame, one has to realize that the
second frame contains relics from the first frame so that the
third frame and subsequent frames effectively compare their
blocks to blocks owing back to the first frame. If one does not
agree with this explanation, then they do not understand what
makes video compression possible. Therefore, Examiner does
not agree that concentrating the discussion on just two frames,
as Appellant insists, is really a worthwhile approach to
understanding the prior art or the claimed invention. It creates a
semantic disagreement where there is no technological
distinction. Thus, because Appellant’s argument, focusing on
just two frames in succession, fails to address how a larger
string of frames would be handled, Appellant’s argument does
not particularly address the true nature of the claimed invention.
Perhaps what Appellant is arguing is that the claimed invention
only keeps the most recent previous frame in the reference
picture buffer rather than a plurality of previous frames.
Single-frame reference picture buffer implementations are
known in the art, so had Appellant clearly argued this point
during prosecution, Examiner would have put more in the
record on how this feature would have been obvious to the
skilled artisan. For now, though, a quick review of Burgess’s
paragraph [0095] demonstrates Burgess teaches, or at least
suggests, single-frame reference picture buffer
implementations. Burgess explains that one could make the
buffer huge, effectively keeping every previous frame and
creating an ever-growing dictionary of image blocks. However,
Burgess also goes the other way, teaching smaller cache sizes
would be beneficial to reduce memory and hardware
requirements as well as reduce the number of comparisons
needed to be performed for each received frame. Burgess uses
terms like “may” to demonstrate the invention does not require
any particular number of cached frames or blocks to be
effective. Instead, Burgess, like the skilled artisan, recognizes
the tradeoffs implicated as one considers the spectrum of
possible cache sizes. Accordingly, Appellant’s averred
distinction over the prior art is not persuasive.
Ans. 14–16 (emphasis added).
Appeal 2020-004830
Application 13/870,878
9
Appellant’s arguments are persuasive of reversible Examiner error.
Burgess discloses a server sharing real-time video signals displayed on a
local GUI with a remote computer, wherein the video signal is digitized,
compressed and transmitted to the remote computer. Burgess ¶¶ 2, 8, 18, 25.
In particular, Burgess discloses applying a smoothing sub-algorithm to the
digital video image data including block of pixels by computing therefor a
CRC similar to a hash code for the block, which is stored in the cache as a
reference hash code. Id. ¶¶ 94–98, 133. Further, Burgess discloses
comparing the hash code for each current block of pixels to hash codes of
corresponding blocks of pixels stored in the cache to determine a cache hit
or a cache miss. Id. ¶¶ 95, 96. Furthermore, Burgess discloses overwriting
oldest block of pixels within the cache by flushing therefrom the least
recently-transmitted block. Id. ¶ 97. We agree with the Examiner that
Burgess’s disclosure comparing a current hash code with other hash codes
(of adjacent and non-adjacent frames) stored in the cache teaches comparing
hash values between at least two adjacent frames. However, we are
persuaded by Appellant’s argument that because such a comparison also
includes that of the current hash code with non-adjacent frames, it vitiates
the claim language, which precludes comparing hash values calculated for
the next current frame to hash values of previous frames of the video when
comparing the next current frame to the current frame as a way to reduce
latency.7 Reply Br. 2.

7 Although we agree the disputed negative limitation is not specifically
disclosed by Burgess, we leave it to the Examiner to determine in the event
of continued prosecution whether the required omission of comparisons of
calculated hash values to hash values for previous frames is the obvious
elimination of a step (additional comparisons) and its function (i.e.,
Appeal 2020-004830
Application 13/870,878
10
Because Appellant shows at least one reversible error in the
Examiner’s obviousness rejection of independent claim 1, we do not reach
Appellant’s remaining arguments. Accordingly, we do not sustain the
Examiner’s obviousness rejection of independent claim 1 over the
combination of Burgess and Carlson. Similarly, we do not sustain the
rejection of claims 2, 3, 5, 7–10, 12, 14–17, and 19, which also recite the
disputed limitations.
VI. CONCLUSION
On this record, we reverse the Examiner’s rejection of claims 1–3, 5,
7–10, 12, 14–17, and 19.

identification of better matches). Although the Examiner finds in passing
that keeping only the most recent previous frame in the reference picture
rather than a plurality of previous frames is an obvious variation that would
have been within the skill of the ordinary artisan, the Examiner fails to fully
develop such an obviousness approach to thereby change the thrust of the
rejection. Ans. 15–16. See In re Thompson, 545 F.2d 1290, 1294 (CCPA
1976) (obvious to eliminate wax pretreatment and its function); In re Kuhle,
526 F.2d 553, 555 (CCPA 1975) (deletion of switch member and its function
was an obvious expedient); In re Marzocchi, 456 F.2d 790, 793 (CCPA
1972) (omission of paraffin and its function would not be an unobvious
expedient); In re Keegan, 331 F.2d 315, 319 (CCPA 1964) (omission of
suture holder and its function of guarding against accidental displacement
was obvious); and In re Porter, 68 F.2d 971, 973 (CCPA 1934) (no
invention where omission of an element is attended by a corresponding
omission of the function performed by that element); see also MPEP §
2144.04 II.
Appeal 2020-004830
Application 13/870,878
11
VII. DECISION SUMMARY
In summary:
Claim(s)
Rejected
35
U.S.C. §
Reference(s)/Basis Affirmed Reversed
1–3, 5, 7–10,
12, 14–17, 19
103 Burgess, Carlson 1–3, 5, 7–
10, 12, 14–
17, 19

REVERSED