Title: Impact%20of%20Structure%20on%20Complexity%20%20Carla%20Gomes%20gomes@cs.cornell.edu%20Bart%20Selman%20selman@cs.cornell.edu%20Cornell%20University%20Intelligent%20Information%20Systems%20Institute%20Kickoff%20Meeting%20AFOSR%20MURI%20May%202001
1Impact of Structure on Complexity Carla
Gomesgomes_at_cs.cornell.eduBart
Selmanselman_at_cs.cornell.eduCornell
UniversityIntelligent Information Systems
InstituteKickoff MeetingAFOSR MURIMay 2001
2Outline
- I - Overview of our approach
- II - Structure vs. complexity -
- results on a abstract domain
- III - Examples of Application Domains
- IV - Conclusions
3Overview of Approach
- Overall theme --- exploit impact of structure on
computational complexity - Identification of domain structural features
- tractable vs. intractable subclasses
- phase transition phenomena
- backbone
- balancedness
-
- Goal
- Use findings in both the design and operation of
distributed platform - Principled controlled hardness aware systems
4- Part I
- Structure vs. Complexity
5Quasigroup Completion Problem (QCP)
Given a matrix with a partial assignment of
colors (32colors in this case), can it be
completed so that each color occurs exactly once
in each row / column (latin square or
quasigroup)? Example
32 preassignment
6- Structural features of instances provide
insights into their hardness namely - Phase transition phenomena
- Backbone
- Inherent structure and balance
7 Are all the Quasigroup Instances (of same size)
Equally Difficult?
What is the fundamental difference between
instances?
8 Are all the Quasigroup Instances Equally
Difficult?
1820
165
50
40
9Complexity of Quasigroup Completion
Median Runtime (log scale)
Fraction of pre-assignment
10Phase Transition
Fraction of unsolvable cases
Fraction of pre-assignment
11Quasigroup Patterns and Problems Hardness
12Bandwidth
Bandwidth permute rows and columns of QCP to
minimize the width of the diagonal band that
covers all the holes. Fact can solve QCP in time
exponential in bandwidth
swap
13Random vs Balanced
Balanced
Random
14After Permuting
Balanced bandwidth 4
Random bandwidth 2
15Structure vs. Computational Cost
Balanced QCP
Computational cost
QCP
Aligned/ Rectangular QCP
of holes
Balancing makes the instances very hard - it
increases bandwith!
16Backbone
Backbone is the shared structure of all the
solutions to a given instance.
This instance has 4 solutions
17Phase Transition in the Backbone (only
satisfiable instances)
- We have observed a transition in the backbone
from a phase where the size of the backbone is
around 0 to a phase with backbone of size close
to 100. - The phase transition in the backbone is sudden
and it coincides with the hardest problem
instances.
(Achlioptas, Gomes, Kautz, Selman 00, Monasson et
al. 99)
18New Phase Transition in Backbone
Backbone
of Backbone
Computational cost
Fraction of preassigned cells
19Why correlation between backbone and problem
hardness?
- Small backbone is associated with lots of
solutions, widely distributed in the search
space, therefore it is easy for the algorithm to
find a solution - Backbone close to 1 - the solutions are tightly
clustered, all the constraints vote to push the
search into that direction - Partial Backbone - may be an indication that
solutions are in different clusters that are
widely distributed, with different clauses
pushing the search in different directions. -
20Structural Features
- The understanding of the structural properties
that characterize problem instances such as
phase transitions, backbone, balance, and
bandwith provides new insights into the
practical complexity of computational tasks.
21- Examples of Application Domains
22Fiber Optic Networks
- Wavelength Division Multiplexing (WDM) is the
most promising technology for the next
generation of wide-area backbone networks. - WDM networks use the large bandwidth available in
optical fibers by partitioning it into several
channels, each at a different wavelength.
23Fiber Optic Networks
Nodes connect point to point fiber optic links
24Fiber Optic Networks
Nodes connect point to point fiber optic links
25Routing in Fiber Optic Networks
Input Ports
Output Ports
1
1
2
2
3
3
4
4
Routing Node
How can we achieve conflict-free routing in each
node of the network?
Dynamic wavelength routing is a NP-hard problem.
26QCP Example Use Routers in Fiber Optic Networks
Dynamic wavelength routing in Fiber Optic
Networks can be directly mapped into the
Quasigroup Completion Problem.
(Barry and Humblet 93, Cheung et al. 90, Green
92, Kumar et al. 99)
27ANTs Challenge Problem
IISI, Cornell University
- Multiple doppler radar sensors track moving
targets - Energy limited sensors
- Communication
- constraints
- Distributed
- environment
- Dynamic problem
28Domain Models
IISI, Cornell University
- Start with a simple graph model
- Successively refine the model in stages to
approximate the real situation - Static weakly-constrained model
- Static constraint satisfaction model with
communication constraints - Static distributed constraint satisfaction model
- Dynamic distributed constraint satisfaction model
- Goal Identify and isolate the sources of
combinatorial complexity
29Initial Assumptions
IISI, Cornell University
- Each sensor can only track one target at a time
- 3 sensors are required to track a target
30Initial Graph Model
IISI, Cornell University
- Bipartite graph G (S U T, E)
- S is the set of sensor nodes, T the set of
target nodes, E the edges indicating which
targets are visible to a given sensor - Decision Problem Can each target be tracked by
three sensors?
31IISI, Cornell University
Initial Graph Model
32IISI, Cornell University
Initial Graph Model
- The initial model presented is a bipartite
graph, and this problem can be solved using a
maximum flow algorithm in polynomial time
33IISI, Cornell University
Sensor Communication Constraints
- In the graph model, we now have additional edges
between sensor nodes
34IISI, Cornell University
Constrained Graph Model
sensors
targets
communication edges
possible solution
35- Complexity and Phase Transition Phenomena of
- Sensor Domain
36Complexity
IISI, Cornell University
- Decision Problem Can each target be tracked by
three sensors which can communicate together ? - We have shown that this constraint satisfaction
problem (CSP) is NP-complete, by reduction from
the problem of partitioning a graph into
isomorphic subgraphs
37- Average Case complexity and Phase Transition
Phenomena
38Phase Transition w.r.t. Communication Level
IISI, Cornell University
Experiments with a random configuration of 9
sensors and 3 targets such that there is a
communication channel between two sensors with
probability p
Insights into the design and operation of sensor
networks w.r.t. communication level
Probability( all targets tracked )
Communication edge probability p
39Phase Transition w.r.t. Radar Detection Range
IISI, Cornell University
Experiments with a random configuration of 9
sensors and 3 targets such that each sensor is
able to detect targets within a range R
Insights into the design and operation of sensor
networks w.r.t. radar detection range
Probability( all targets tracked )
Normalized Radar Range R
40 41Distributed CSP Model
IISI, Cornell University
- In a distributed CSP (DCSP) variables and
constraints are distributed among multiple
agents. It consists of - A set of agents 1, 2, n
- A set of CSPs P1, P2, Pn , one for each agent
- There are intra-agent constraints and
inter-agent constraints
42DCSP Model
IISI, Cornell University
- We can represent the sensor tracking problem as
DCSP using dual representations - One with each sensor as a distinct agent
- One with a distinct tracker agent for each target
43Sensor Agents
- Binary variables associated with each target
- Intra-agent constraints
- Sensor must track at most 1 visible target
- Inter-agent constraints
- 3 communicating sensors should track each target
44Target Tracker Agents
- Binary variables associated with each sensor
- Intra-agent constraints
- Each target must be tracked by 3 communicating
sensors to which it is visible - Inter-agent constraints
- A sensor can only track one target
45Implicit versus Explicit Constraints
- Explicit constraint (correspond to the
explicit domain constraints) - no two targets can be tracked by same sensor
(e.g. t2, t3 cannot share s4 and t1, t3 cannot
share s9) - three sensors are required to track a target
(e.g. s1,s3,s9 for t1) - Implicit constraint (due to a chain of
explicit constraints (e.g. implicit constraint
between s4 for t2 and s9 for t1 )
s1
s2
s3
s4
s5
s6
s7
s8
s9
t1
1
1
x
x
1
0
x
x
x
x
x
1
x
x
x
1
x
1
t2
x
x
x
1
0
x
x
1
1
t3
46Communication Costs for Implicit Constraints
- Explicit constraints can be resolved by direct
communication between agents - Resolving Implicit constraints may require long
communication paths through multiple agents ?
scalability problems
47- Conclusions and Research Directions
48Future directions
- Study structural issues and inpact on
complexity, as they occur in the distributed
environments e.g. - effect of balancing
- backbone (insights into critical resources)
- refinement of phase transition notions
considering additional parameters
49DCSP Model
- With the DCSP model, we plan to study both
per-node computational costs as well as
inter-node communication costs -
- We are in the process of applying known DCSP
algorithms to study issues concerning complexity
and scalability
50Summary
- We have made considerable progress in our
understanding of the nature of hard
computational problems - structure matters! - We have harnessed a variety of mechanisms with
proven impact on time-critical problem solving. - A rich spectrum of applications taking advantage
of these new methods are on the horizon in
planning, scheduling and many other areas. - Future focus on Dynamic Distributed models
51The End