Efficient Dissemination of Personalized Information Using Content-Based Multicast (CBM)

1
Efficient Dissemination of Personalized
Information Using Content-Based Multicast (CBM)
An SAIC Company
Rahul Shah Ravi Jain Farooq Anjum Dept.
Computer Science Autonomous Comm. Lab Applied
Research Rutgers University NTT DoCoMo USA
Labs Telcordia sharahul_at_paul.rutgers.edu jain_at_doco
molabs-usa.com fanjum_at_telcordia.com
Work performed while at Applied Research,
Telcordia
2
Outline

• Motivation and background
• Problem definition
• Simulation results
• Concluding remarks

3
Mobile Filters for Efficient Personalized
Information Delivery

• Users want targeted, personalized information,
  particularly
• as the amount and diversity of information
  increases,
• the capabilities of end devices are limited and
  resources are scarce
• Applications like personalized information
  delivery to large numbers of users rely on
  multicast to conserve resources
• Traditional network multicast (e.g. IP multicast)
  
• does not consider the content or semantics of the
  information sent
• Management difficult as number of groups increase
• Content-Based Multicast (CBM) filters the
  information being sent down the multicast tree in
  accordance with the interests of the recipients
• Problem how to place software information
  filters in response to
• the location and interests of the users, and how
  these change
• the additional cost and complexity of the filters

4
Related work

• Multicast
• Application layer multicast
• Assumes only unicast at the IP layer, while CBM
  assumes a multicast tree (either at the IP or the
  application layer)
• Examples Francis, Yoid, 2000 Chu et al., End
  System Multicast, Sigmetrics 2000 Chawathe et
  al., Scattercast, 2000
• Publish-subscribe systems
• Many-many distribution with matching done by
  brokers in the network
• In CBM the brokers form the underlying multicast
  tree
• Examples Aguilera, 1998 Banavar, 1998
  Carzaniga, 1998
• Modifications to IP multicast
• Opyrchal, Minimizing number of multicast groups,
  Middleware 2000
• Wen et al., Use active network approaches,
  OpenArch 2001
• Theoretical work
• Classical k-median and facility location problems

5
Multicast filtering example

• Without filters, all 8 items are sent on all 15
  links 120 traffic units
• With filters at all internal nodes, traffic
  47 units
• With filters at 3 internal nodes, traffic
  63 units

6
Mobile code problem definition

• Problem 1 Bandwidth optimization problem
• Criterion Find optimal placement to minimize
  total bandwidth
• Cost model k-Filters Allow at most k filters to
  be used
• Problem 2 Delay optimization problem
• Criterion Find optimal placement to minimize
  mean delivery delay
• Cost model Delay
• Each filter adds a delay D for processing
• The reduction in link utilization also results in
  reduction in link delay
• Optimal placement changes as users move or change
  interests
• the filtering code should or could be mobile and
• the placement algorithm should be fast
• Results
• optimal centralized off-line algorithm for
  bandwidth optimization. Time O(k n2)
• optimal centralized off-line algorithm for delay
  optimization. Time O(n2)
• Two centralized O(n) heuristics that restrict
  filter moves
• Evaluation using simulations

7
Filtering algorithm framework

• For simplicity, we assume the following framework
• 1 The multicast tree has previously been
  constructed and is known
• 2 Filters can be placed at all internal nodes of
  the multicast tree
• If not, simply consider the subtree where filters
  are permitted
• 3 Subscriptions propagate from the users to the
  source
• There is a simple list of information items that
  users can request
• Subscription changes are batched at the source
• At every batch (time slice) x of the users
  change subscription
• 4. The source calculates filter placements
• 5 The source dispatches filters to the (new)
  placement
• Currently we ignore signaling costs of
  subscriptions and filter movement because
  negligible for the applications considered (news
  clips, video clips, music, etc)
• Alternatively could consider that filters are
  available at all nodes and are only
  activated/deactivated by signaling messages

8
Bandwidth minimization problemOptimal
centralized algorithm
f(p)
Child of Lowest filtering ancestor, p
Model of multicast tree at source
f(p)

• f(x) Traffic required at node x
• Execution time O(k n2)
• n number of nodes in tree
• Time complexity calculated
• using Tamir (1996)

T(v, i, p)
Node v
i filters, max
f(l)
f(r)
j filters
i - (j -1) filters

• Dynamic programming recurrence relations
• Traffic in the subtree rooted at v, with a filter
  at v
• T(v, i, p) f(l) f(r) min j 0 ? j ? i
  T(l, j, l) T(r, i - j - 1, r)
• Traffic with no filter at v
• S(v, i, p) 2 f(p) min j 0 ? j ? i T(l, j,
  p) T(r, i - j, p)
• Traffic at a leaf node v T(v, i, p) S(v,
  i, p) 0
• Minimum traffic is min T(v, k,
  p), S(v, k, p)

9
Simulation results Filters can be very
effective

• Seven-level complete binary tree (n 127), with
  64 leaves
• m 64 messages
• Uniform subscription p(i, j) Prob User i
  subscribes to message j p

10
Interest Locality increases filtering benefits
Locality model P(i, j) 1/N if i j
qr /N else,
where r LCA(i, j) q is a skew parameter
inversely proportional to locality
11
Bandwidth minimization problemHeuristic
centralized algorithm

• Node importance, I
• amount by which total
• traffic changes by
• placing a filter there
• Execution time O(n)

• Importance of node v
• I(v) (f(v) - f(l)) z(l) (f(v) - f(r))
  z(r), where
• z(x) 1, if x has a filter
• 1 z(left-child of x) z(right-child of
  x), otherwise
• z(x) is number of edges in the subtree rooted at
  x affected by a filter at x

12
Centralized heuristic

• Subscriptions propagate up to the source, which
• calculates the required flow amount at each edge
  and the Importance value of each node
• tries the Importance Flip
• Imax(v) max v v does not have a filter I(v)
  
• Imin(u) min u u has a filter, I(u)
• If Imax(v) gt Imin(u), move the filter from u to v
• If the most Important non-filtering node is more
  important than the least Important filtering
  node, swap the filter location
• otherwise, tries the Parent-child flip
• is allowed to make at most one filter move
• The source dispatches one new filter, or a move
  instruction to one existing filter

13
Code mobility is not useful with uniform
subscriptions and static users

• opt optimal placement at each trial
• heu heuristic re-run at each trial
• Init initial placement, kept unchanged

14
Mobility model

• User mobility Users gradually move from the left
  subtree to the right subtree
• Subscription skew, q
• At t 0, users in left subtree have
• p 0.3 q, users in right p 0.3 - q
• At t i, swap probabilities of user i in left
• subtree with user i in right subtree

15
User mobility motivates filter mobility

16
Further work

• Theoretical improvements
• More efficient algorithms
• Achieves O(n logn) time complexity
• Prototype and obtain actual bandwidth costs and
  delays for filter movement using Aglets
  technology
• A distributed filtering algorithm, where the
  filters are agents that coordinate with minimal
  involvement of the source
• How to avoid thrashing and loops
• How to ensure semi-autonomous agent movements do
  not degrade performance
• Investigate different application domains