Davis Social Links

1
Davis Social Links
S. Felix Wu Computer Science Department University
of California, Davis wu_at_cs.ucdavis.edu http//ww
w.cs.ucdavis.edu/wu/
2
Internet

• Basic datagram service between one IP address and
  another
• Routing exchanging the information regarding the
  address space and how to reach them.
• Allow by Default
• Applications built on top of the services
• MySpace, Facebook, Orkut, SecondLife,
  LiveJournal, LinkedIn, Wikipedia, Skype, Google,
  Bittorrent, LimeWire, Youtube, IM,

3
Easy to Send Receive

• Many, if not most, in our society can quickly
  learn how to use Email Internet.
• You dont need to understand the low level
  details (e.g., how SMTP, BGP, OLSR or AODV really
  works or what my current IP address is) before
  you can communicate with your friends as long as
  you know their identity!

4
Urgent! Please contact me!
FROMMR.CHEUNG PUI Hang Seng Bank Ltd Sai Wan Ho
Branch 171 Shaukiwan Road Hong Kong. Please
contact me on my personal box puicheungcheungpui_at_
yahoo.com Let me start by introducing myself. I
am Mr. Cheung Pui,director of operations of the
Hang Seng Bank Ltd,Sai Wan Ho Branch. I have a
obscured business suggestion for you. Before the
U.S and Iraqi war our client Major Fadi Basem who
was with the Iraqi forces and also business man
made a numbered fixed deposit for 18
calendar months, with a value of Twenty Four
millions Five Hundred Thousand United State
Dollars only in my branch. Upon maturity several
notice was sent to him,
5
ITRI considers this as a spam!
6
Easy to Send Receive

• Easy for both the good users and the spammers.
  (fair simplicity)
• The spammers abuse the sending right, while the
  good users have very limited options to counter
  back.
• how easy can we change our email address?
• how often do we need to do that?
• A receiver or the owner of the identity
  should have some control.
• But, that means also burden to the users.

7
Easy to Send Receive

• Easy for both the good users and the spammers.
  (fair simplicity)
• The spammers abuse the sending right, while the
  good users have very limited options to counter
  back.
• how easy can we change our email address?
• how often do we need to do that?
• A receiver or the owner of the identity
  should have some control.
• But, that means also burden to the users.

8
Possible Ideas of Future Internet

• Allow by Default versus Deny by Default
• Receiver versus Sender controllability
• Reconsider the role of Service Providers versus
  P2P

9
DSL (Davis Social Links)
A
B

• Principle
• Effective human communication reflects the
  social/trust relationship and reputation within
  our society.
• Design
• Can we integrate and enhance our cyber
  communication system with the human social/trust
  model?

F
A
B
F
F
10
(No Transcript)
11
Social Networking

• Social Network services
• Friendster, MySpace, Facebook, Orkut, LinkedIn..
• Social Links, Interest Keywords, Search
  Community Services

12
Social Networks of Users
http//en.wikipedia.org/wiki/List_of_social_networ
king_sites

• FaceBook 34M
• Friendster 47M
• Hi5 50M
• LinkedIn 12M
• MySpace 200M
• Orkut 47M
• Yahoo! 360o 4M
• among many others.

13
(No Transcript)
14
Social Network Access Interfaces

• FBML - Facebook Markup Language
• FQL - Facebook Query Language
• REST (REpresentational State Transfer) Web API
• FBJS - Implementation of Javascript on the
  Facebook platform

15
DSL (Davis Social Links)
A
B

• Principle
• Effective human communication reflects the
  social/trust relationship and reputation within
  our society.
• Design
• Can we integrate and enhance our cyber
  communication system with the human social/trust
  model?

F
A
B
F
F
16
The same message content

• M from Cheung Pui
• M from Cheung Pui via IETF mailing list
• M from Cheung Pui via Karl Levitt

17
Social Link Identification

• M from Cheung Pui
• ? Probably a spam
• M from Cheung Pui via IETF mailing list
• ? Probably not interesting
• M from Cheung Pui via Karl Levitt
• ? Better be more serious

18
Social Link Identification

• M from Cheung Pui
• ? Probably a spam
• M from Cheung Pui via IETF mailing list
• ? Probably not interesting
• M from Cheung Pui via Karl Levitt
• ? Better be more serious

Either M is important, or Karls machine has
been subverted!
19
A Social Link
Trust Attribute
representing a trust relationship between two
directly communicating entities
20
A Direct Social Link
Trust Attribute
Karl is a very well respected individual to me! I
suspect that Karl is too busy to install the
latest patches!
21
Message Value Prioritization
Link Ranks Reputation Incentives Other Trust
Metrics
Application IDS
22
DSL (Davis Social Links)
A
B

• One important goal
• Can we build a P2P social network service?
• Allows Search/Crawl and Control
• Scalable and Secure/Robust
• Control implies relatively more dynamic social
  network!

F
A
B
F
F
23
A Path of Social Links
24
Receiving an email via a Path of Social Links
25
A Couple of Issues first

• How to discover the social route path, if no
  direct social link exists?
• How to prevent spammers from easily discovering a
  large number of valid social paths?

26
Social Keywords
Soccer, BGP, Davis, California, Intrusion
Detection,
27
Social Keywords
Soccer, BGP, Davis, California, Intrusion
Detection,
Ownership
Social keywords represents my interests and the
semantic/social interpretation of myself.
28
Cultural Natural Language
???, BGP, ??,
Students, Committee, Soccer
29
IP

• IP address is all we need to communicate and
  everything else is on top of IP.
• IP is global routable identity
• However, too simple to be perfect?!
• Can we remove the need for global routable
  identities?

30
Identifier Keyword

• Under DSL, each peer has its own decision or
  policy regarding what social keywords to
  announce, and it can change the announcement
  itself any time.
• NO global unique network-wide identifier
• No unique IP address or email address in the
  network layer
• Identity only in the application layer
• Relative Peering relationship only
• Local link identity

31
DSL

• Lets decide some of the everything else
• Global Flooding (w/o GRI)
• Then, we decide the possible candidates for us to
  communicate
• Scope-limiting
• Finally, we rely on the Application layer to nail
  down the one or ones.

32
Multiple Social Links
BGP, Davis, California, Intrusion Detection,
Soccer, Wine, Bike, Davis, California,
33
Policy-Driven Forwarding
BGP, Davis, California, Intrusion Detection,
Soccer, Wine, Bike, Davis, California,
34
Propagation Aggregation
35
Aggregation
Ownership
Similar to Scope-Limited Flooding in
Gnutella/Freenet, but more scalable with
aggregation.
Ownership
36
Scalability Locality
???
soccer
Ownership
football
Ownership
37
Scalability Locality
Davis
soccer
Ownership
football
Ownership
38
How to find content X?
Owner of X
Other Peers
Remember that under DSL there is no such thing
called global identifier.
39
Use the right keywords for X!
soccer
Owner of X
Other Peers
Query
Davis
soccer
40
Layering
Application
Application
TCP/IP
DSL Social Layer
Link
Social Link
Layer-2
IP-like
Traditional layer-2
41
A Social/Overlay Link
42
Davis - Washington DC - Sacramento

• Social path discovery
• Sending the real messages along the path

43
Route Discovery
Keywords and aggregated keywords
. . . .
content addressable emails
44
Function F(incoming) outgoing
Policy driven Maybe itself a service
Northern California
Davis CA
001
002
Potential Senders
Potential Receivers
incoming
outgoing
DSL ID
Davis California
Japan
001
Japan
Northern California
002
003
45
Forwarding Source Routing
Sender Data packet SR 003, 001
46
Forwarding Source Routing
Sender Data packet SR 003, 001
001
003
47
Privacy and Accountability

• No single intermediate node will have the
  knowledge, at least in the network layer, about
  who is talking to who on what.
• In DSL, no global identifier.
• However, the messages themselves contain
  sufficient information to trace back toward the
  social path, if necessary and nodes are
  collaborative.
• BTW, this (tracing on a per packet/email basis)
  is impossible for todays Internet, even for
  route path.

48
(No Transcript)
49
Problems with Overlay

• Triangle Overlay Routing
• Davis, Washington DC, Sacramento
• Traffic following the Social/Overlay path
• You can still bed DDoSed or Spammed!
• IP/email addresses are still there.

50
MANET in Airport, Train, Flight
Application
DSL Social Layer
Social Link
Layer-2
Who has the 1307th piece of media X? Promotion on
new Pizza toppings A,B, C! Who has seen this
lost baby girl in the past 3 minutes? Who wants
to play OK-bridge or Warcraft-III?
51
DSL-wifi

• Link Established
• representing a neighbor relationship in the
  application layer

52
MANET

• MANET (Mobile Ad Hoc Network)
• Network Layer Identify

53
Infrastructure-less Content Sharing

• Conference/Meeting
• File Transfer
• Infrastructure ?Infrastructure-less
• Identification

54
DSL Routing
55
DSL Keyword Propagation
56
A Relative Identity under DSL

• Key words plus one DSL path
• Why will this be sufficient for communication?
• Especially, when even the DSL information itself
  is decentralized? (I.e., in general, no one
  single DSL router knows the complete DSL path.)

57
(No Transcript)
58
DSL in a Social Network
Owner of X
Other Peers
Query AppTest
59
Examples

• I want to find all the pizza places in Davis.
• Can I do it without yet another service provider?
• How about unpopular topics?
• Can I find the contact information for Felix Wu
  in Davis (who is interested in soccer)?

60
DSL vs. Google
61
Google

• Its about the content
• Data-centric networking.
• Input to the Engine
• A set of key words characterizing the target
  document.
• Output
• A set of documents/links matching the keywords

62
DSL

• Its also about the content
• Application will decide the mechanism to further
  the communication.
• Input to the Decentralized Engine
• A set of key words characterizing the target
  document (plus the aggregation keywords).
• Output
• A set of DSL entities with the DSP (Davis Social
  Path pointer) matching the keywords

63
DSL Search Engine
Receiver or Content
Sender or Reader
DSL Social World
We are not just connecting the IP addresses! We
are connecting all the contents that can be
interpreted!
64
Google vs. DSL

• Google is essentially a routing framework
  between the contents and their potential
  consumers.
• Google decides how to extract the key words
  from your (the owner) web page or document.

65
Google vs. DSL

• Google is essentially a routing framework
  between the contents and their potential
  consumers.
• Google decides how to extract the key words
  from your (the owner) web page or document.
• A DSL owner/receiver to be has the complete
  control over that. A balance between
• How I would like others to know about me?
• And, I might want different folks to know me in
  different ways!
• How I can differentiate myself from other Felix
  Wu?

66
Applications

• 911-like scenarios (power/infrastructure)
• Commercial and non-Commercial Advertisements for
  the local community.
• Ad Hoc Peer-to-Peer Sharing without the Internet!

67
Unstructured P2P
Where is this piece of music? What is the
location of this content? What is the current IP
address of this skype user?
68
Structured P2P
69
DHT-based BT Tracker
Sharing/Downloading FreeBSD 5.4 CD images
The owner of the identifier matching this hash
entry will be the tracker for the corresponding
index key!
Index key (Content-Dependent)
PUT GET
70
DHT-based P2P
PUT GET
Index key
Routing from a key to an identifier
71
DHT vs. DSL

• DHT
• Contents, Keys, Identifiers
• Conceptually, we have a global hash table!
• DSL
• Contents, Keys, Relative Identifiers
• Social/trust relationships along the route path

72
No Unique Identifier!!!

• Two peers might announced exactly the same set of
  keywords!
• Both of them will receive the query, and this
  might not be a problem.
• Usually, end2end applications can tell which one
  (or ones) is really wanted.

73
Trackerless BitTorrent
Owner of X
Other Peers
Query AppTest
74
(No Transcript)
75
Tracker Volunteer
Content/Key Tracker
Volunteer
Other Peers
Query AppTest
76
Tracker Volunteers
Content/Key Tracker
Volunteers
Other Peers
Query AppTest
77
DSL-Torrent

• How to seed your torrent?
• Use DSL to search for all the volunteers
• You can screen out unqualified trackers by not
  contacting them. (owner controllability)
• How to contact/select the tracker(s)?
• Use the keywords, of course
• But, use the right aggregation keywords to reach
  the trackers closer to the UCDavis social
  circle!

78
(No Transcript)
79
Tracking the BT clients

• Seed
• Tracker
• With Trackerless, the inspector might become the
  tracker itself to record all the IP addresses of
  the peers
• Traceability vs. Privacy
• How about DSL-Torrent?

80
Applications

• 911-like scenarios (power/infrastructure)
• Commercial and non-Commercial Advertisements for
  the local community.
• Ad Hoc Peer-to-Peer Sharing without the Internet!

81
Applications

• 911-like scenarios (power/infrastructure)
• Commercial and non-Commercial Advertisements for
  the local community.
• What is a community?
• Ad Hoc Peer-to-Peer Sharing without the Internet!

82
Community-Oriented Social Network

• Peers sharing some common interests
• Communities intersect
• Community as a boundary for information flow and
  policy

83
Community-Oriented Social Network

• Peers sharing some common interests
• Communities intersect
• Community as a boundary for information flow and
  policy
• How to establish a community?
• How to maintain the membership?
• How to set the inter- intra- community policy?

84
Forming a Community

• UCDavis as a community
• When should we have a community for Aggie?
• Davis, California, Aggie

85
Scope-Limiting Keyword Propagation
86
Scope-Limiting Keyword Propagation/Flooding
87
Scope-Limiting Keyword Propagation/Flooding
88
Penalization on Spams

• If Karl had forwarded lots of junk mails to
  Felix, then Felix will lower the priority of the
  social link between them.
• Karl will be informed about the downgrade.
• Karl needs to investigate which messages (from
  which incoming DSLs) were causing the downgrade
  and penalize toward the sources.

89
DSL Email
90
(No Transcript)
91
Parallel/Overlay Communication
92
DSL is an old idea!
A
B

• We, as human, have been using similar
  communication principles. Maybe it is a good
  opportunity to re-think about our cyber
  communication system.
• Identity is a per-application, context-oriented,
  and sometime relative issue.
• Forming cyber communities of interests for
  application.

F
A
B
F
F
93
DSL is still an old idea!
A
B

• Many applications already have social network
  like structure to enable P2P sharing across
  Internet.
• e.g., media sharing, on-line game, restaurant
  recommendation,
• Should we push these into a generic Social
  Network layer-3 to support all the applications?

F
A
B
F
F
94
A Different Internet?!

• Current Internet every IP address will be able
  to communicate with every other IP address!
• Allow by Default
• DSL-based Internet we have a large number of
  pairs (two entities and their corresponding
  direct social link)
• Deny by Default

95
Comparison

• IP/email
• Convergence to an absolute consistent state
• IP/email addresses are all you need, but the
  controllability is biased toward the sender
• DSL
• Convergence to a relative consistent state
• No global network identity. Every DSL entity
  defines its own relative identity based on origin
  keywords.
• Controllability is more balanced with other
  application challenges.

96
Spam on DSL

• Still needs much more evaluation on how the
  spammer can exploit the DSL framework.
• But, maybe some spammers will no longer spam
  economically!
• Social network will give the merchants much
  better channels to send their advertisements!
• Using spam to deliver product information over
  DSL might seriously hurt the sales on the social
  channels.
• Still there will still be some global spam
  activities.

97
(No Transcript)
98
Acknowledgement
A
B

• Lerone Banks
• Dimitri DeFigueiredo
• Felix Huang
• Juan Lang
• Xiaoming Lu
• Matt Spear
• Shih-Ming Tsang
• Shawn Ye

more info wu_at_cs.cudavis.edu
F
A
B
F
F
99
Questions?
wu_at_cs.ucdavis.edu or DSL to me