USOO8468426B2
`
`(12) United States Patent
`Bims
`
`(10) Patent No.:
`(45) Date of Patent:
`
`US 8,468.426 B2
`Jun. 18, 2013
`
`(54) MULTIMEDIA-AWARE
`QUALITY-OF-SERVICE AND ERROR
`CORRECTION PROVISIONING
`
`(75) Inventor: Harry Bims, Menlo Park, CA (US)
`(73) Assignee: Apple Inc., Cupertino, CA (US)
`(*) Notice:
`Subject to any disclaimer, the term of this
`patent is extended or adjusted under 35
`U.S.C. 154(b) by 984 days.
`
`(21) Appl. No.: 12/217,308
`
`(22) Filed:
`(65)
`
`Jul. 2, 2008
`Prior Publication Data
`
`US 201O/OOO2692 A1
`
`Jan. 7, 2010
`
`(2006.01)
`
`(51) Int. Cl.
`H03M, 3/00
`(52) U.S. Cl.
`USPC ............................................. 714/774; 714/52
`(58) Fist of Classification Search
`71.4/52, 752, 774
`See application file for complete search history. s
`
`(56)
`
`References Cited
`
`U.S. PATENT DOCUMENTS
`5,446,747 A
`8, 1995 Berrou
`5.488,609 A
`1/1996 Hluchy et al.
`5,581,544 A 12/1996 Hamada et al.
`6,026,164 A
`2/2000 Sakamoto et al. ............ 380,217
`6,134,243 A * 10/2000 Jones et al. ................... 370/465
`6,453,355 B1
`9, 2002 Jones
`6,512,778 B1
`1/2003 Jones
`6,714.984 B2
`3, 2004 Jones
`6,717,952 B2
`4/2004 Jones
`6,744,763 B1
`6, 2004 Jones
`6,829,648 B1
`12/2004 Jones
`7,016,366 B2
`3/2006 Kawarai et al.
`7,068,645 B1* 6/2006 Phadnis et al. ................ 370,352
`7,075,927 B2
`7/2006 Mo et al.
`
`82006 Zhang et al.
`7,085,291 B2
`2:SS R: 858. as: al
`7,142,934 B2 11/2006 Janik
`7,142,935 B2 11/2006 Janik
`7,167,765 B2
`1/2007 EE
`7,180,860 B2
`2/2007 Fonden et al.
`7.263,064 B2
`8/2007 Yoshimura et al.
`7,280,562 B2 10/2007 Sindhushayana et al.
`23.9 R: 1939,
`g
`al.
`7.301928 B2 11/2007 Nakabayashi et al.
`2001/0033581 A1* 10, 2001 Kawarai et al. ............... 370/468
`2001.0053149 A1* 12/2001 MO et al. ........
`370,389
`2002/0024944 A1* 2/2002 Zhang et al. ...
`370,349
`2002/0093.913 A1* 7/2002 Brown et al. ................. 370,232
`(Continued)
`FOREIGN PATENT DOCUMENTS
`EP
`1739 900 A1
`1, 2007
`Primary Examiner — Joshua Lohn
`(74) Attorney, Agent, or Firm — GaZdzinski & Associates,
`PC
`
`ABSTRACT
`(57)
`Methods and apparatus for associating each data packet in a
`media stream with logic corresponding to a particular quality
`of-service (QoS) and/or error correction requirement. In an
`exemplary embodiment, each packet in the media stream is
`assigned a frame tag which designates a particular quality
`of-service and/or error correction scheme for the correspond
`ing packet. At least a portion of each packet is encoded
`according to the packet's designated quality-of-service as
`indicated by the frame tag. A receiver accesses the frame tags
`from within the transmitted media stream in order to deter
`mine the appropriate means for processing or decoding the
`encoded portion of each packet. In this manner, each packet
`within the media stream can have its own quality-of-service
`and/or error correction requirements and processing, thereby
`enhancing link efficiency and better enforcing QoS policy
`across the system.
`
`53 Claims, 18 Drawing Sheets
`
`
`
`
`
`
`
`
`
`Tags 216
`
`Unencoded
`Data Stream
`302
`
`To Receiving PHY 400
`
`Transmitting PHY 300
`
`
`
`
`
`202 Media
`
`Physical Layer FEC
`EncoderEntity 304
`
`
`
`
`
`306(2)
`
`306(1)
`Encoder 1
`
`
`
`
`
`Page 1 of 36
`
`GOOGLE EXHIBIT 1015
`
`

`

`US 8,468.426 B2
`Page 2
`
`U.S. PATENT DOCUMENTS
`2002/O126675 A1* 9, 2002 Yoshimura et al. ...... 370,395.21
`2003/0012372 A1* 1/2003 Cheng ............................. 380/28
`2004/0202156 A1* 10, 2004 Fonden et al.
`370,389
`2004/0223611 A1* 11/2004 Yan et al. ........................ 380,37
`
`2005/02200 11 A1* 10/2005 Parker et al. .................. 370,229
`ck
`S.S. A. 1939. SNS S. it... :
`ayashi et al. ........
`
`
`
`* cited by examiner
`
`Page 2 of 36
`
`

`

`U.S. Patent
`U.S. Patent
`
`Jun. 18, 2013
`Jun. 18, 2013
`
`Sheet 1 of 18
`Sheet 1 of 18
`
`US 8,468.426 B2
`US 8,468,426 B2
`
`
`
`s
`100
`
`wwoO
`-
`did
`ie]_
`re
`O
`oO
`ma
`S.
`5
`C
`g
`E
`she,
`3
`d.
`fa
`
`2Qa
`
`o Eo
`
`wn
`2
`2
`yo
`5ws
`s
`5
`5
`o
`A
`Go
`a
`S
`&
`g
`he
`t
`
`Mm
`n
`
`3
`g
`O
`
`
`
`rd
`3
`vo
`d
`S
`£
`s
`oa
`l
`
`-
`A4
`S
`mg
`
`r- - ps
`|- A.
`2s
`mePa
`
`aR
`
`Page 3 of 36
`
`Page 3 of 36
`
`€
`

`

`-.|@(e)
`——pwonels
`
`JOAIag
`
`uddbeneue
`€dd
`
`Sheet 2 of 18
`
`ooz|r002
`aseg
`
`‘Iddy
`
`—-|(¥)
`(¢)
`
`uddSeenee
`¢dd
`
`00c|"
`002
`
`US 8,468.426 B2
`US8,468,426 B2
`
`INV
`YONd
`
`sjoyoeg7adh
`(Z)p0z
`
`syayped|
`9dX]
`(p07
`sjoyoed1904y
`(1)P07
`sjoyoudJ9dAJ,(1)p0Z
`syjoyord7odAI,(Z)p07
`syayoeg1904y,
`(1)p0z
`
`U.S. Patent
`U.S. Patent
`
`Jun. 18, 2013
`Jun. 18, 2013
`
`507
`
`OYA
`
`BIPsW207
`
`wreey¢
`
`PIP]207
`
`wesns
`
`Z ‘
`
`SI
`
`
`
`OIZ
`o1z
`
`STIQoW
`
`uoneys
`
`4 of 36
`
`Page
`
`Page 4 of 36
`
`
`
`

`

`U.S. Patent
`
`Jun. 18, 2013
`
`Sheet 3 of 18
`
`US 8,468.426 B2
`
`
`
`e??CI ZIZ
`
`Xogu I,
`
`uaq | ~~~~
`
`0 IZ
`
`Page 5 of 36
`
`

`

`U.S. Patent
`
`Jun. 18, 2013
`
`Sheet 4 of 18
`
`US 8,468.426 B2
`
`Fig.2b
`
`Hint Track 214
`
`
`
`
`
`
`
`
`
`Frame
`
`216(1-n)
`
`Packet ID 218
`
`OOS Class 220
`13
`
`FEC Mode 222
`
`13
`
`
`
`
`
`
`
`
`
`
`
`
`
`13
`
`13
`
`13
`
`
`
`
`
`
`
`
`
`
`
`
`
`Page 6 of 36
`
`

`

`U.S. Patent
`
`Jun. 18, 2013
`
`Sheet 5 of 18
`
`US 8,468.426 B2
`
`
`
`Packet Assignment Table 224
`
`Packet
`Type
`204
`
`QoS FEC
`Class Mode
`
`Fig. 2c
`
`Hint Track 214
`
`Packet
`Packet 2
`Packets
`Packet 4
`Packets
`Packet 6
`
`Packet 9
`
`Packet 10
`
`Packet 11
`
`Packet 12
`
`Packet 13
`
`Packet 15
`
`Packet 16
`Packet 17
`
`-Om
`
`-O
`
`W
`
`1
`
`5
`
`1.
`
`s
`s
`3
`3
`3
`3
`
`5
`
`3
`
`3
`
`3
`
`3
`
`5
`
`s
`
`1.
`
`Frame
`Tags
`216(1-n)
`
`Page 7 of 36
`
`

`

`Sheet 6 of 18
`
`SpoWOFA
`
`8st“
`
`9577
`
`A
`
`C]
`ye
`ourwsjoq
`Bey
`owely
`asieg
`
`US 8,468,426 B2
`
`OSC
`
`PZ
`‘Stz
`
`pueSod
`sulULisjeg
`
`U.S. Patent
`
`Jun. 18, 2013
`
`097%
`
`Z9OT~
`
`PeALlag
`a[quiessy
`
`yoelyJU
`
`weaysqns
`
`B
`
`yoyoeg
`
`&PeLLWIA
`
`oq}S]
`
`UISSIIOI
`JOYIeg
`YIOMJANSSIDV
`
`CSC
`
`PoAlsooy
`osieg
`
`sjoyoeg
`
`8 of 36
`
`Page
`
`Page 8 of 36
`
`
`
`
`
`
`

`

`OO€AHd
`suniiwusueLyOOFAHd
`uxpostg
`4“
`BUIAIN90YOT,
`(€)90€
`fo
`
`US 8,468,426 B2
`
`| J
`
`opoougq
`
`Z J
`
`aposuq
`
`(Z)90€€
`
`Iopooug
`
`(u)90¢
`
`U.S. Patent
`
`Jun. 18, 2013
`
`Sheet 7 of 18
`
`OddJoAvyq
`yeorshyg
`
`SF
`
`Wess
`ejeq
`
`cOE
`
`poposusun,
`
`917
`s8BL
`
`BIpsIlCOT
`
`
`
`9 of 36
`
`Page
`
`Page 9 of 36
`
`
`
`
`
`

`

`(4)90r
`bopAMUY
`Japooaq
`
`u
`
`13p0d9(q
`
`&
`
`eona-
`(€)90¢
`
`JadJ9Kke'7T
`Jeotshyg
`
`Jun. 18, 2013
`Jun. 18, 2013
`
`eed
`popoousuy)
`
`youlyBILTIC
`
`
`00¢AHd
`SUIAIOIOY
`
`weaTsqns
`your],
`
`NS
`
`WIH17
`
`weransqns
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`U.S. Patent
`U.S. Patent
`
`
`
`
`
`JoyZIE{OL
`
`JOAR’T
`YIOMJON,
`
`BIPSWCOC
`
`weeds
`
`10 of 36
`
`Page
`
`Sheet 8 of 18
`Sheet 8 of 18
`
`US 8,468.426 B2
`US 8,468,426 B2
`
`J[esa
`
` ] t
`
`(1)90¢
`
`opoooqd
`
`(7)90r
`
`Page 10 of 36
`
`
`
`
`

`

`U.S. Patent
`
`Jun. 18, 2013
`
`Sheet 9 of 18
`
`US 8,468.426 B2
`
`
`
`possoboudun
`
`Z09
`
`
`
`uueòJ?S e?eG
`
`I º ImpoW
`
`
`
`
`
`
`
`
`
`
`
`
`
`Page 11 of 36
`
`

`

`U.S. Patent
`
`Jun. 18, 2013
`
`Sheet 10 of 18
`
`US 8,468.426 B2
`
`00
`
`(£)909
`
`I JOSS90OJAI
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`Page 12 of 36
`
`

`

`U.S. Patent
`
`Jun. 18, 2013
`
`Sheet 11 of 18
`
`US 8,468.426 B2
`
`Fig. 7
`
`Assign Packet
`Type to Each
`Packet 700
`
`
`
`Provide
`Appropriate QoS
`Profile/Processing
`for Each Packet
`702
`
`Provide
`Appropriate FEC
`Encoding for Each
`Packet 704
`
`Arrange Packets
`into Media Stream;
`Transfer Media
`Stream
`706
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`Process according
`to QoS Profile for
`Each Packet
`712
`
`Decode FEC for
`Each Packet
`710
`
`Receive Media
`Stream at
`Recipient Node
`708
`
`
`
`Page 13 of 36
`
`

`

`U.S. Patent
`
`Jun. 18, 2013
`
`Sheet 12 of 18
`
`US 8,468.426 B2
`
`Fig. 8
`
`702 QoS Processing
`
`
`
`Any more
`Packets to
`analyze?
`800
`
`Route Packets to
`Appropriate QoS
`Module(s) Based
`upon QoS Class 806
`
`Determine Packet
`Type 802
`
`Process Packets
`808
`
`Assign QoS Class to Packet
`Based upon Packet Type,
`Indicate in Hint Track 804
`
`Page 14 of 36
`
`

`

`U.S. Patent
`
`Jun. 18, 2013
`
`Sheet 13 of 18
`
`US 8,468.426 B2
`
`Fig. 9
`
`704 FEC Encoding
`
`
`
`Any more
`Packets to
`analyze?
`900
`--
`
`Y
`
`Route Packets to
`Appropriate FEC
`Encoder(s) Based
`upon FEC Mode
`906
`
`Determine Packet
`Type 902
`
`Encode Packets
`908
`
`Assign FEC Mode to Packet
`Based upon Packet Type;
`Indicate in Hint Track
`904
`
`Page 15 of 36
`
`

`

`U.S. Patent
`
`Jun. 18, 2013
`
`Sheet 14 of 18
`
`US 8,468.426 B2
`
`Fig. 10
`
`
`
`710 FEC Decoding
`
`Assemble Hint Track
`from Hint Track
`Substream 1000
`
`FEC Decode Packet
`1008
`
`Any more
`Packets to
`analyze?
`1002
`
`Route Packet to
`Appropriate FEC
`Decoder Based upon
`FEC Mode 1006
`
`Determine FEC Mode
`of Next Packet from
`Hint Track 1004
`
`Page 16 of 36
`
`

`

`U.S. Patent
`
`Jun. 18, 2013
`
`Sheet 15 of 18
`
`US 8,468.426 B2
`
`Fig. 11
`
`Any more
`Packets to
`analyze?
`1 102
`
`712 QoS Processing
`
`
`
`QoS
`Processing of
`Packet 1108
`
`Route Packet to
`Appropriate QoS
`Module Based Upon
`QoS Class 1106
`
`Determine QoS Class
`of Next Packet from
`Hint Track 1104
`
`Page 17 of 36
`
`

`

`U.S. Patent
`
`Jun. 18, 2013
`
`Sheet 16 of 18
`
`US 8,468.426 B2
`
`
`
`2OO
`
`O2
`
`2O4.
`
`Page 18 of 36
`
`

`

`U.S. Patent
`
`Jun. 18, 2013
`
`Sheet 17 of 18
`
`US 8,468.426 B2
`
`1300
`
`MCROPROCESSOR
`
`1302
`
`MEMORY UNIT
`
`
`
`1304
`
`
`
`DEVICE
`INTERFACE
`
`
`
`DISPLAY
`INTERFACE
`
`WIRELESS
`NTERFACE
`
`Fig. 13
`
`Page 19 of 36
`
`

`

`bOrl
`asequied
`
`U.S. Patent
`U.S. Patent
`
`zovl
`
`Jun. 18, 2013
`Jun. 18, 2013
`
`Sheet 18 of 18
`Sheet 18 of 18
`
`901
`Jossad01g/|
`Asoursjl
`
`wajysXsqng
`waysAsqng
`
`US 8,468.426 B2
`US 8,468,426 B2
`
`ZIrl
`
`V9OPT
`/
`
`V80rl
`
`
`
`[eqquag
`
`PL
`S17
`
`80bT Olvl
`
`} w
`
`ayshsqng
`
`Jossos01J/]
`
`Aroursy
`WAMOd
`
`[etjuad
`
`aseqeieq
`
`aoeyiayu
`
`Page 20 of 36
`
`Page 20 of 36
`
`

`

`US 8,468.426 B2
`
`1.
`MULTIMEDIA-AWARE
`QUALITY-OF-SERVICE AND ERROR
`CORRECTION PROVISIONING
`
`COPYRIGHT
`
`A portion of the disclosure of this patent document con
`tains material that is subject to copyright protection. The
`copyright owner has no objection to the facsimile reproduc
`tion by anyone of the patent document or the patent disclo
`sure, as it appears in the Patent and Trademark Office patent
`files or records, but otherwise reserves all copyright rights
`whatsoever.
`
`10
`
`BACKGROUND OF THE INVENTION
`
`15
`
`2
`correction requirements may change dynamically over the
`lifetime of the multimedia stream on the downlink. For
`example, FIG. 1 (prior art MPEG decoding) illustrates an
`MPEG video encoding 100 comprising three frame types,
`Intra frames 102 (I-frames), Predictive frames 104
`(P-frames), and Bi-directional frames 106 (B-frames).
`I-frames are generally more important to an MPEG video
`codec than P-frames, and P-frames are generally more impor
`tant to an MPEG video codec than B-frames. This is because
`P-frames are dependent on previous I-frames and P-frames,
`and B-frames are dependent on previous and future I-frames
`and P-frames. As a result, the loss of a B-frame will not affect
`I-frame and P-frame processing, yet the loss of a P-frame,
`though not affecting I-frame processing, may affect B-frame
`processing. Finally, the loss of an I-frame may affect both
`P-frame and B-frame processing.
`One of the primary goals of video compression is to reduce
`the size (in bytes) of any particular video stream: MPEG
`compression thus aims to accomplish this reduction by
`removing spatial redundancy from within each video frame,
`and temporal redundancy between video frames. Such redun
`dancy is known to occur frequently in video streams. For
`example, in a scene where a person moves in front of a
`stationary background, only the moving regions need to be
`represented. The parts of the scene that are not changing are
`redundant and therefore do not need to be transmitted repeat
`edly.
`A variety of approaches are often used to minimize this
`redundancy, and therefore also minimize the information nec
`essary for displaying the entire video stream. In some cases,
`data representing an entire image is required to be transmit
`ted, while in other cases, the only data that is necessary is the
`data representing the differences between the current image
`and the prior one. Other approaches can also be utilized. Such
`as techniques in global motion compensation, block motion
`compensation, variable block-size motion compensation,
`motion estimation, etc. Typically, the approach which
`requires the fewest bits for image representation is the
`approach which is deemed optimal for a particular instant in
`time. Since the optimal approach usually varies with time
`(and in many instances, may also depend upon what image
`was sent in the prior instant in time), then in these cases, a
`plurality of frame types is used in order to accomplish opti
`mization of the video stream (for example, in an MPEG
`stream, the I-frame, B-frame, and P-frame formats discussed
`above).
`FIG. 2 is a block diagram illustrating a generalized three
`node transmission stream sequence as known in the prior art.
`A media stream 202 is transmitted from an application server
`206 to a base station 208; the base station transmits the media
`stream to the mobile station 210. As shown by the figure, the
`media stream comprises in data packets 200, 200(1)-200(n),
`where each data packet is assigned to one of i packet types
`204, 204(1)-204(i). The sequencing of the transmitted pack
`ets need not necessarily correspond with any sequencing of
`possible packet types. For example, the first and second pack
`ets 200(1) 200(2) in the stream 202 might be a Type 1 packet
`204(1), while the third packet 200(3) might be a Type 2 packet
`204(2).
`Irrespective of packet type 204, the access network or base
`station 208 selects an appropriate transmission mode for all
`data packets 200 in the media stream 202. This mode is
`selected based upon what is necessary for the transmitter 208
`to accommodate the highest QoS necessary for any individual
`packet 200(1)-200(n) within the stream 202, to ensure timely
`delivery at the mobile station 210. The mode selected may
`potentially also select the most expansive error correction
`
`25
`
`30
`
`35
`
`40
`
`1. Field of Invention
`The present invention relates generally to the field of qual
`ity-of-service (QoS) maintenance or enhancement overa data
`network. More specifically, the present invention is in one
`exemplary aspect directed to providing packets in a media
`stream with their own QoS and forward error correction
`(FEC) mechanism.
`2. Description of Related Technology
`So-called “Quality-of-service' or QoS refers generally to
`resource reservation and allocation control mechanisms. For
`example, QoS when implemented can provide different pri
`ority to different data flows, or guarantee a certain level of
`performance to a data flow. The goal of QoS is to improve the
`user experience and a networks ability to deliver predictable
`results for sensitive applications such as audio, video, and
`voice applications. Elements of network performance within
`the scope of QoS often include bandwidth (throughput),
`latency (delay), and error rate. There are generally two broad
`classes of QoS: data reliability and temporal reliability. Each
`makes different demands on network technologies.
`QoS utilization and allocation of resources have long been
`implemented and described in the prior art for various other
`types of networks including, interalia, packet Switched net
`works. For example, U.S. Pat. No.5.488,609 to Hluchy, et al.
`issued Jan. 30, 1996 and entitled “Dynamic rate adjustment
`for overload control in communication networks’ is exem
`plary of one such implementation. It discloses a device (500)
`and method (300) which provide for management of resource
`allocation on selected links in a connection-oriented commu
`45
`nication network Such that existing connections may share the
`burden of freeing up resources for accommodating new con
`nections. The rate of certain connections is dynamically
`adjusted for the entire connection using information on the
`status of each link selected or marked for reallocation. Links
`are marked based on control information in the link state, and
`the in-call rate adjustment is based on negotiable Quality-of
`service (QoS) parameters.
`The current generation protocol standards for broadband
`access networks provide for a Quality-of-service (QoS) pro
`55
`visioning and forward error correction (FEC) provisioning of
`downlink data streams that is uniformly applied to all packets
`in a data stream. Examples of such networks include so-called
`“3G” (Third Generation) cellular and IEEE 802.11e net
`works. However, many multimedia streams, such as MPEG
`60
`streams, contain a heterogeneous mix of packet types. Each
`type of packet within the stream can potentially be encoded
`differently according to media type (e.g. a different audio or
`video codec), and have its own QoS and/or error correction
`requirements. In addition, each type of packet can potentially
`be further differentiated according to frame type for a given
`codec. In some applications, QoS requirements and/or error
`
`50
`
`65
`
`Page 21 of 36
`
`

`

`US 8,468.426 B2
`
`10
`
`15
`
`3
`protocol necessary to reasonably accommodate the transmis
`sion of any individual packet 200(1)-200(n) within the media
`stream 202. In this manner, the mode selected may be chosen
`based upon any of the following aspects of a connection (or
`combination thereof): time to provide service, audio quality,
`Video quality, echo, loss, reliability, and/or error correction.
`The mobile receiver 210 comprises any logic necessary for
`accommodating the transmission mode selected.
`Traditionally, access networks 208 parse the “Type of Ser
`vice' field in an IPv4 header, or the “Traffic Class field' in an
`IPv6 header to determine data handling processes. These field
`values are mapped by the access network to an internal QoS
`policy and/or error correction policy for delivery. For certain
`applications (e.g. MPEG data streams), various packets
`within a data stream require different QoS levels. Referring
`back to FIG. 1, a media stream comprised of I-frames 102.
`B-frames 106, and P-frames 104 is being transmitted by an
`access network. As previously mentioned, the base station
`selects an appropriate transmission mode for all data packets
`within the media stream. The transmitter operates on “physi
`cal and “medium access control layers, and must treat the
`media stream as a single unified data stream. The transmitter
`does not have Sufficient computing resources to further clas
`Sify packets by a combination of the application and its indi
`vidual application-layer packet types (the transmitter does
`not have “application layer” processing capabilities). Even
`with Sufficient resources to perform application layer classi
`fication, the transmitter cannot duplicate the decision logic
`within every application that dynamically chooses the correct
`QoS requirements for individual packet types. Due to the
`media stream's mixed composition of frames, the transmit
`ting device must accommodate the highest level of both QoS
`and error correction (the I-frame is the most critical frame
`type for MPEG encoding/decoding, and will necessitate high
`QoS and error correction for the entire media stream).
`Current and next generation access networks can perform
`deeper packet inspection than the aforementioned “Type of
`Service' fields in order to perform the packet classification
`required for QoS policy enforcement. Deeper packet inspec
`tion can allow packet classification of higher protocol layers
`in application streams with Such classification mapped to
`specific QoS policies. However, as these solutions do not
`perform application-aware functionality (for aforementioned
`reasons), and require the access network to be programmed
`ahead of time with the proper QoS policy of every class of
`45
`application stream that is Supported by the access network.
`Supported application stream classes must be determined
`a-priori by a management system component of the access
`network and communicated to every base station in the access
`network. In this way, all application streams that are mapped
`to a particular stream class by a base station will have the
`same QoS policy enforced.
`Since the access network cannot anticipate every possible
`application that might use the network to stream multimedia
`packets, it cannot rely upon packet classification to identify
`any and all future applications whose streams are transported
`over the access network. Instead, all possible classes of
`streams that are supported by the QoS policies are identified.
`Each application stream is thereby mapped to a particular
`stream class by the packet classification algorithm. As a
`result, a plurality of application streams might result in the
`same classification without the awareness of the base station.
`As evidenced by this arrangement, the prior art relies upon
`a common logic in order to process each data packet 200(1)-
`200(n) arriving at the mobile station 210. The prior art is
`inefficient in the sense that certain packet types need not
`require the same error correction/quality-of-service provi
`
`55
`
`4
`Sioning necessary for other packets types. More specifically,
`the error correction/quality-of-service provisioning cannot
`change dynamically during the media stream session. Static
`provisioning for packet types of changing requirements often
`unnecessarily wastes high priority network resources on
`lowerpriority packet types, and increases the number of bytes
`necessary for transmitting the entire media stream 202 (and
`likewise, increases the media stream's 202 expected transfer
`time). At the other extreme, selecting a less stringent trans
`mission mode for the media stream 202 in order to increase
`transfer speed results in lower priority network resources
`being used for high priority packet types, and unacceptable
`packet error rates. In the same vein, selecting an error correc
`tion protocol with less redundancy often requires retransmis
`sion of all or part of the media stream 202 in the event that
`errors ultimately occur. In some cases, the efficiency gains
`associated with the selection of an error correction protocol
`with less redundancy are more than offset by the efficiency
`costs of retransmission.
`For example, an iTunes movie that is viewed on a client
`device as it is being streamed (let's say a user is watching a
`preview) may have a very different QoS requirement set for
`P-frames, B-frames, and I-frames, than an iTunes movie that
`is downloaded to the client device in the background for later
`viewing offline. For every variation in downlink QoS require
`ments demanded by the iTunes application, the access net
`work would have to maintain a separate stream classification.
`This can become impractical for large numbers of applica
`tions with large numbers of stream types per application.
`U.S. Pat. No. 7,085,291 to Zhang et al. filed on Aug. 1,
`2006 discloses an enhanced radio link protocol (RLP) in a
`wireless access network that is network aware. The RLP
`attempts to increase radio link quality by various Automatic
`RepeatreCuest (ARQ) mechanisms. The RLP framing struc
`ture included Supports and enables at least network layer
`packet boundary detection, dynamic and adaptive ARQ
`schemes for QoS Support on a per-packet basis, and an RLP
`frame structure for fast adaptation to physical layer channel
`rate/RLP frame sizes. Optional uses include Supporting nega
`tive acknowledgment (NAK) based ARQ.
`U.S. Pat. No. 7.301,928 to Nakabayashi et al. filed on Nov.
`27, 2007 discloses an error correction encoding rate selection
`table provided in an error correction processing unit of a
`packet transfer apparatus. The table stores an error correction
`encoding rate preset to maintain a desired QoS in correspon
`dence with a protocol type and an application type. When a
`transmission packet is transferred to a wireless transmission
`path, an encoding control unit judges the protocol type and
`application type of a transmission packet from a header of the
`transmission packet, and in accordance with a judgment
`result and the error correction encoding rate selection table,
`an error correction encoding rate is selected, and the trans
`mission packet is subjected to error correction encoding and
`transferred.
`U.S. Pat. No. 7,075,927 to Mo et al. filed on Jul. 11, 2006
`discloses a method and system for transporting traffic having
`disparate qualities of service classes across a packet-switched
`network that includes receiving at an ingress node of a net
`work a plurality of packets, some of which comprise a trans
`port label, where the transport label has an associated QoS
`class that is defined externally to the network. Packets having
`a QoS class comprising delay bound guarantees and a low
`drop priority are combined into a first internal QoS class.
`Packets having a QoS class comprising a flexible drop prior
`ity and no delay bound guarantees are combined into a second
`internal QoS class. Packets having a QoS class including no
`
`25
`
`30
`
`35
`
`40
`
`50
`
`60
`
`65
`
`Page 22 of 36
`
`

`

`5
`delivery guarantees are combined into a third internal QoS
`class. The packets are transmitted in the network based on
`their internal QoS class.
`U.S. Pat. No. 7,016,366 to Kawarai et al. filed on Mar. 21,
`2006 discloses a method to achieve QoS control, drop control 5
`and multicast control of a variable-length packet at high speed
`in small-scale hardware. A packet divider divides a variable
`length packet into fixed-length packets, and an input buffer
`section stores the divided fixed-length packets into queues by
`output lines and by QoS classes. A large number of QoS 10
`classes are mapped into only two kinds of classes including a
`guaranteedbandwidth class for which an assigned bandwidth
`is guaranteed and a best effort class for which a surplus
`bandwidth is allocated, thereby to achieve scheduling at the
`inputside by an inter-line scheduler. An output buffer section 15
`assembles a variable-length packet from fixed-length packets
`that have been obtained by Switching at a Switch section in an
`output buffer section. A QoS control is performed based on a
`packet length.
`U.S. Pat. No. 7.263,064 to Yoshimura et al. filed on Aug. 20
`28, 2007 discloses a method for transmitting packets classi
`fied according to QoS requirement from a transmitting node
`to a receiving node. The transmitting node is configured to
`select sequentially a QoS class, to divide a queued packet to
`be transmitted into a plurality of predetermined data units, to 25
`transmit one of the predetermined data units, and to apply a
`transmitter-side retransmission control process to the data
`unit to be transmitted when the selected class is a QoS class
`specified for data type packets. The receiving node is config
`ured to receive sequentially the data units transmitted from 30
`the transmitting node, to assemble a plurality of received data
`units to restorean original packet for each QoS class, and to
`apply a receiver-side retransmission control process to the
`received data units when the received data unit belongs to one
`of the data type QoS classes.
`U.S. Pat. No. 7,292,591 to Parker et al. filed on Nov. 6,
`2007 discloses a packet processing system architecture and
`method. According to a first aspect of the invention, packet
`parser functions are distributed throughout a packet process
`ing system comprising a packet classification system and a 40
`packet modification system. According to a second aspect of
`the invention, an egress mirroring function is provided to the
`system. According to a third aspect of this invention, a multi
`dimensional quality-of-service indicator for a packet is pro
`vided. According to a fourth aspect of this invention, a cas- 45
`caded combination of multiple, replicated packet processing
`systems is used to process a packet. A fifth aspect of this
`invention involves any combination of one or more of the
`foregoing.
`U.S. Pat. No. 7,280,562 to Sindhushayana et al. filed on 50
`Oct. 9, 2007 discloses a method and apparatus for variable
`length Physical Layer packet generation. Multiple Security
`Layer packets may be multiplexed into a single Physical
`Layer packet to increase efficiency, wherein the Multiple
`Security Layer packets may have variable lengths. In one 55
`embodiment, different format Multiple Security Layer pack
`ets for different users are combined into capsules that form
`the Physical Layer packet. Shorter packets are for users in
`poor channel conditions or requiring Smaller amounts of data
`due to the applications and the accompanying QoS require- 60
`ments. In one embodiment, a modified Preamble structure
`provides for Unicast or multi-user packets. Alternate embodi
`ment provides modified Rate Sets, a mechanism for identify
`ing ACK from a single-user packet or a multiplexed packet
`(delayed ACK), ON/OFF keying for ACK channel v/s bi- 65
`polar keying used in IS-856, and/or multi-valued interpreta
`tion of DRC.
`
`35
`
`US 8,468.426 B2
`
`6
`U.S. Pat. No. 7, 180,860 to Fonden et al. filed on Feb. 20,
`2007 discloses a method for the provision of a defined qual
`ity-of-service in a packet Switched communication system
`with interconnected nodes for the forwarding of data packets
`is described. The system comprises at least one edge node for
`the connection to user equipment or a further communication
`system and for processing data packets. The packets comprise
`a data field specifying the handling of the packets and the
`nodes perform a differentiated handling of packets according
`to said data field. The communication system comprises or is
`connectable to a database which contains a record for a user
`specifying a quality-of-service. An edge node which pro
`cesses a packet for said user is provided with quality param
`eters from the database. The edge node sets the data field
`specifying the handling of the packet according to the record.
`Devices and programs to perform the method are also
`described.
`European Patent Publication No. EP1739900 to Lin
`entitled “A METHOD FOR ACQUIRING THE QOS OF
`THE MULTIMEDIA STREAM PERIODICALLY dis
`closes a method for acquiring media stream Quality of Ser
`vice (QoS) periodically. The MGC sets an inspect duration
`during which the MGC periodically acquires the QoS infor
`mation of the media stream, thereby acquiring the QoS infor
`mation of the media stream and controlling the media stream.
`Two ways for the MGC periodically acquiring the QoS infor
`mation are provided. The MG actively submits the QoS infor
`mation to the MGC and the MG submits the QoS information
`to the MGC according to the periodic request of the MGC..
`One embodiment provides the capability of adjusting the call
`microscopic QoS., enhances the real-time quality of evaluat
`ing the QoS and that of dynamically adjusting call control
`strategy, inherits the ability of the original protocol for evalu
`ating the microscopic QoS of the network during a certain
`period and ostensibly provides more accurate evaluation.
`U.S. Pat. No. 7,068,645 to Phadnis et al. filed on Jun. 27,
`2006 discloses a network device (for example, a network
`access server or home gateway) providing different quality
`of-services to different layer-3 datagrams when transporting
`on tunnels. A tunnel may be implemented to provide different
`QoS to different packets depending on the packet header. The
`network device examines the header of each datagram to
`determine the specific QoS to be provided. At least the data
`portion in the datagram is encapsulated for transportation on
`the tunnel. The encapsulated data portion in turn is encapsu
`lated in the form of one or more packets, with the packet
`format to reflect the QoS determined for the datagram. When
`the tunnel is implemented on UDP/IP and the datagram is an
`IP (Internet protocol) datagram, the TOS/Precedence bits of
`the IP datagram may be copied into the precedence/TOS bits
`of the UDP/IP packet(s).
`U.S. Pat. No. 5,581,544 to Hamada, et al. issued Dec. 3,
`1996 and entitled “Method and apparatus for evaluating in
`ATM multiplexing apparatus in which priority control is per
`formed and for controlling call admissions and optimizing
`priority control on the basis of the evaluation discloses a
`probability transition matrix S, that expresses a multiplexing
`process which includes nested threshold priority control and
`classified priority control. A state equation using the matrix S,
`is solved by substituting therein upper and lower bounds of a
`probability distribution of a cell arrival count at in an average
`time series, to calculate a probability distribution of a cell
`length in a buffer. From the thus calculated cell length prob
`ability distribution, the QoS is evaluated on a priority class
`basis. Based on the QoS evaluation, optimization of call
`admission control and priority control is accomplished.
`
`Page 23 of 36
`
`

`

`1