single sink edge installation kunal talwar uc berkeley

46
Single Sink Edge Installation Kunal Talwar UC Berkeley

Post on 20-Dec-2015

215 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Single Sink Edge Installation Kunal Talwar UC Berkeley

Single Sink Edge Installation

Kunal TalwarUC Berkeley

Page 2: Single Sink Edge Installation Kunal Talwar UC Berkeley

Problem Definition

Given: A graph G=(V,E) , sink ‘t Sources s1,s2,…, sm

k Discount types: “building” cost per unit length q

“routing” cost per unit length q

Find cheapest installation to route a unit of demand from each source si

Page 3: Single Sink Edge Installation Kunal Talwar UC Berkeley

Example

Given sources and the underlying graph

t

Find subgraph and cable type for each edge to minimize total cost

Page 4: Single Sink Edge Installation Kunal Talwar UC Berkeley

t

Example

Given sources and the underlying graphFind subgraph and cable type for each edge to minimize total cost

t

Page 5: Single Sink Edge Installation Kunal Talwar UC Berkeley

Related Work

Salman, et.al. ’97 - A constant factor for single cable type case (LASTs)Awerbuch, Azar ’98 - O(log n loglog n)- approximation (tree embedding)Meyerson, et.al. ’00 - O(log n) (comb.) Garg et.al. ’01 - O(k) (LP rounding)Guha et.al. ’01 - O(1) (you just heard !)This work – LP rounding - O(1).Goel, Estrin, ’03 – O(log n) oblivious to q,q

Page 6: Single Sink Edge Installation Kunal Talwar UC Berkeley

Special case

Only one cable typeExtraspecial subcases: If ) build steiner tree. If ) use shortest path tree.

In general, want a Light Approximate Shortestpath Tree

KRY ’94 – LASTs – tree of cost at most 2 times a given connecting tree, with dT(root,v) at most 3 times dG(root,v).

Page 7: Single Sink Edge Installation Kunal Talwar UC Berkeley

Single cable type: Algorithm

Build a steiner tree.Convert to LAST !

Analysis -OPT ¸ (optimal steiner tree).OPT ¸ v dG(s,v).

Cost = 2(steiner tree) + 3v dG(s,v) · 7OPT.

Page 8: Single Sink Edge Installation Kunal Talwar UC Berkeley

About OPT…

It’s a tree !As we travel from a source to sink Total traffic only increases… .. So thicker and thicker cables…

We let the LP know the above…

Page 9: Single Sink Edge Installation Kunal Talwar UC Berkeley

Integer Program

Variable zqe is 1 if edge e has

discount type q installedFlow variable fj

e;q is 1 if flow from j uses discount type q on edge eObjective function – cost of building the network –

Page 10: Single Sink Edge Installation Kunal Talwar UC Berkeley

Integer Program

Variable zqe is 1 if edge e has

discount type q installedFlow variable fj

e;q is 1 if flow from j uses discount type q on edge eObjective function – cost of building the network – cost of routing the demands –

Page 11: Single Sink Edge Installation Kunal Talwar UC Berkeley

Integer program

Subject to : Flow conservation Flow monotonicity Outflow =1 Route on edge e only if edge built Integrality contraints

Linear

Page 12: Single Sink Edge Installation Kunal Talwar UC Berkeley

Rounding the linear program

Top downUse the linear program solution to guide the algorithmUse the linear program cost as the lower bound

Page 13: Single Sink Edge Installation Kunal Talwar UC Berkeley

Algorithm outline

Tk = {t}

For each discount type q (from highest to lowest) Identify what to connect in this stage Connect it to Tq+1 with discount type q

to get Tq

Page 14: Single Sink Edge Installation Kunal Talwar UC Berkeley

Identifying what to connect

For a fractional solution f, flow from vj travels some average number

of edges on low discount types.

Page 15: Single Sink Edge Installation Kunal Talwar UC Berkeley

Identifying what to connect

For a fractional solution f, flow from vj

travels some average number of edges on low discount types.Beyond that radius, fractional solution uses high discount types

Page 16: Single Sink Edge Installation Kunal Talwar UC Berkeley

Identifying what to connect

For a fractional solution f, flow from vj

travels some average number of edges on low discount types.Beyond that radius, fractional solution uses high discount typesForm balls of radius Rq

j around node vj

Page 17: Single Sink Edge Installation Kunal Talwar UC Berkeley

Identifying what to connect

For a fractional solution f, flow from vj

travels some average number of edges on low discount types.Beyond that radius, fractional solution uses high discount typesForm balls of radius Rq

j around node vj

Select a set of non intersecting balls in increasing order of radii.

Page 18: Single Sink Edge Installation Kunal Talwar UC Berkeley

Identifying what to connect

For a fractional solution f, flow from vj travels some average number of edges on low discount types.Beyond that radius, fractional solution uses high discount typesForm balls of radius Rq

j around node vj

Select a set of non intersecting balls in increasing order of radii.Each node vl has a buddy within distance 4Rq

l

Page 19: Single Sink Edge Installation Kunal Talwar UC Berkeley

Identifying what to connect….

Page 20: Single Sink Edge Installation Kunal Talwar UC Berkeley

Identifying what to connect….

LessThan4Rq

j

vj

Page 21: Single Sink Edge Installation Kunal Talwar UC Berkeley

Connecting it…

Contract all selected ballsShrink Tq+1

Build a Steiner tree on the contracted nodesConvert to LASTEach selected vertex has a proxy in its ball at distance at most Rq

j

Page 22: Single Sink Edge Installation Kunal Talwar UC Berkeley

Connecting it ….

j

i

Page 23: Single Sink Edge Installation Kunal Talwar UC Berkeley

Connecting it ….

j

i

Page 24: Single Sink Edge Installation Kunal Talwar UC Berkeley

Connecting it ….

Less than Rqj

j

i

Page 25: Single Sink Edge Installation Kunal Talwar UC Berkeley

Connecting it ….

Less ThanRq

j

j

i

Page 26: Single Sink Edge Installation Kunal Talwar UC Berkeley

Connecting it ….

LessThanRq

jj

i

Each node (using its buddy) has someone in the tree Tq within Rq

j

Page 27: Single Sink Edge Installation Kunal Talwar UC Berkeley

Analysis: Building Cost

Key Lemma : Let (S) be the set of edges on the boundary of S ¶ B(vj,Rq

j), r 2 Sc. Let z,f be any feasible solution to the LP. Then

Proof :

Page 28: Single Sink Edge Installation Kunal Talwar UC Berkeley

Lemma proof:

t

Flow f

Flow (1-f)

vj

Page 29: Single Sink Edge Installation Kunal Talwar UC Berkeley

Lemma proof:

t

Flow f

Flow (1-f)

vj

S

A total flow of 1 leaves S

Flow crossing the boundary on low discounts reaches there on low discounts. (monotonicity)

Suppose the high discounts built at (S) < 1/2

Then > half the flow travels at least 2Rq

j on low discounts

This flow itself contributes > Rqj

to Rqj. Contradiction.

Page 30: Single Sink Edge Installation Kunal Talwar UC Berkeley

Lemma proved…

Lemma : Let (S) be the set of edges on the boundary of S ¶ B(vj,Rq

j), r 2 Sc. Let z,f be any feasible solution to the LP. Then

Page 31: Single Sink Edge Installation Kunal Talwar UC Berkeley

Lemma proved…

Lemma : Let (S) be the set of edges on the boundary of S ¶ B(vj,Rq

j), r 2 Sc. Let z,f be any feasible solution to the LP. Then

Recall that Steiner tree LP is:

Page 32: Single Sink Edge Installation Kunal Talwar UC Berkeley

Lemma proved…

Lemma : Let (S) be the set of edges on the boundary of S ¶ B(vj,Rq

j), r 2 Sc. Let z,f be any feasible solution to the LP. Then

Recall that Steiner tree LP is:

i.e. 2p¸ q zqe is a feasible fractional Steiner

tree.

Page 33: Single Sink Edge Installation Kunal Talwar UC Berkeley

Building Cost (contd…)

Steiner tree LP has gap · 2Hence our steiner tree cost is no more than 2q times e p¸ q zp

e.

Thus the LASTq is no more than twice this.Let OPTb

q = OPT’s building cost for discount type q.Then LASTq cost is 8p¸ q (q/p)OPTb

p

Page 34: Single Sink Edge Installation Kunal Talwar UC Berkeley

Scaling…

We prune the discount types in the beginning to be

sure that they are all different enough !More formally, ensure q+1 ¸ 2 q

q+1 · (1/2) q

Can be done with a factor 2 change in cost

Page 35: Single Sink Edge Installation Kunal Talwar UC Berkeley

Building Cost (contd…)

p ¸ 2p-qq

LASTq cost is 8p¸ q (q/p)OPTbp

Now things work out fine ! We get :

Total building cost · 16 OPTb

Page 36: Single Sink Edge Installation Kunal Talwar UC Berkeley

Routing costs

Path from v to t uses increasingly higher discount type. Let the path be v=u0,u1,…uk=t uq- uq+1 uses discount type q.

v’s routing cost = q q dT(uq,uq+1)

Page 37: Single Sink Edge Installation Kunal Talwar UC Berkeley

Routing costs…

u3= t

u0=vu1

u2 Proxy(v)2 Length of this= 2 d(u1,u2)

Routing cost on

discount type 2

Page 38: Single Sink Edge Installation Kunal Talwar UC Berkeley

Routing costs…

u3= t

u0=vu1

u2 Proxy(v)2 Length of this= 2 d(u1,u2)· 3 2 d(u1,Proxy(v))

coz we built a LAST !

Page 39: Single Sink Edge Installation Kunal Talwar UC Berkeley

Routing costs…

u3= t

u0=vu1

u2 Proxy(v)2 Length of this= 2 d(u1,u2)· 3 2 d(u1,Proxy(v))· 3 2(d(u1,v) + d(v,Proxy(v)))

Page 40: Single Sink Edge Installation Kunal Talwar UC Berkeley

Routing costs…

u3= t

u0=vu1

u2 Proxy(v)2 Length of this= 2 d(u1,u2)· 3 2 d(u1,Proxy(v))· 3 2(d(u1,v) + d(v,Proxy(v)))

bounded by 6R2v

~ what fractional sol. pays

Page 41: Single Sink Edge Installation Kunal Talwar UC Berkeley

Routing costs…

u3= t

u0=vu1

u2 Proxy(v)

already paid 1 ¸ 2 2 for this

2 Length of this= 2 d(u1,u2)· 3 2 d(u1,Proxy(v))· 3 2(d(u1,v) + d(v,Proxy(v)))

Page 42: Single Sink Edge Installation Kunal Talwar UC Berkeley

Routing costs…

u3= t

u0=vu1

u2 Proxy(v)2 Length of this= 2 d(u1,u2)· 3 2 d(u1,Proxy(v))· 3 2(d(u1,v) + d(v,Proxy(v)))Total routing cost v pays in sol is O(what v pays in

LP)

Page 43: Single Sink Edge Installation Kunal Talwar UC Berkeley

Hence….

Theorem: Algorithm described has cost within a constant factor of the LP optimum.

Recap:LP tells us how far from vj to go before LP can pay for building high discount typesLAST + selection of balls ensures routing cost is not too highScaling crucial in both cases !

Page 44: Single Sink Edge Installation Kunal Talwar UC Berkeley

Conclusions

Get a constant factor approximation algorithmThe natural LP has a constant integrality gap.

Page 45: Single Sink Edge Installation Kunal Talwar UC Berkeley

Conclusions

Get a constant factor approximation algorithmThe natural LP has a constant integrality gap.

Open problems:More reasonableh constantsThe general buy-at-bulk problem Combinatorial lower bounds off by log LP might be the right approach

Page 46: Single Sink Edge Installation Kunal Talwar UC Berkeley

Paper available at

http://www.cs.berkeley.edu/~kunal/acads/bb.ps

Questions?