Information Theory for Mobile Ad-Hoc Networks (ITMANET): The FLoWS Project

FLoWS Progress and Next Steps

Andrea Goldsmith

Phase 3 KickoffSept. 14-15, 2009

FLoWS Challenge, Progress, and Goals

• Develop and exploit a more powerful information theory for mobile wireless networks.

• The development of this theory has progressed along three main thrust areas, with breakthrough progress and new theory in each area.

• Synergies between thrust areas have emerged, which are blurring the lines between thrusts.

• In Phases 3-4 our goal is to identify and attack the largest outstanding ITMANET challenges

Capacity Delay

Power

Upper Bound

Lower Bound

Capacity and Fundamental Limits

Application Metrics

Capacity

Delay

Power

Utility=U(C,D,E)

Application andNetwork Optimization

(C*,D*,E*)

Constraints

Degrees ofFreedom

Models andDynamics

Application Metrics and

Network Performance

LayerlessDynamic Networks

New Paradigmsfor Upper

Bounds

Models

New MANET Theory

MANET Metrics

Metrics

Fundamental Limitsof Wireless Systems

Thrust Objectives and Rationale• Models and Metrics (Leads: Effros,Goldsmith,Medard):

– Objective: Develop a set of metrics for dynamic networks that capture requirements of current and future applications

– Rationale: Models for MANETs are needed that are tractable yet lead to general design and performance insights

• New Paradigms for Upper Bounds (Leads: Effros,Medard)– Objective: Obtain bounds on a diversity of objectively-defined metrics

for complex interconnected systems.– Rationale: A comprehensive theory for upper bounding the

performance limits of MANETs will help guide design

• Layerless Dynamic Networks (Lead: Zheng, Coleman)– Objective: Design of networking strategies as a single dynamic

probabilistic mapping, without pre-assigned layered structure – Rationale Remove layering and statics from MANET theory.

• End-to-End Metrics and Performance (Leads: Ozdaglar,Shah)– Objective: Provide an interface between application metrics and

network performance– Rationale: A theory of generalized rate distortion, separation, and

network optimization will improve application performance

Two New PIs Added to FLoWS

• Cover and El Gamal have been added for the last two phases to complement the existing team

• Cover’s work will focus on – Coordinated capacity: How much dependence can be set up

with a given set of communication constraints– Applications include distributed game theory, task assignment

and rate distortion theory.

• El Gamal’s work will focus on – Network information theory to develop new coding schemes

for the canonical channel models with many users– Computing/decision making over a network with distributed

sources: lossy distributed averaging

Metrics and ModelsLead: Goldsmith and Effros

All PIs Contribute

Project Thrusts and Organization

New Paradigmsfor Upper Bounds

Co-Leads: Effros and Medard

— Cover— El Gamal— Koetter— Goldsmith

LayerlessDynamic Networks

Co-Leads: Zheng and Coleman

— Cover— Effros— El Gamal— Goldsmith— Koetter— Medard— Moulin— Shah

App. Metrics andNetwork Performance

Co-Leads: Ozdaglar and Shah

— Coleman— Effros— Goldsmith— Johari — Medard

StructuredCoding

Thrust Synergies and New Intellectual Tools

EquivalenceClasses

Optimization

Code Construction

Combinatorial Tools

DynamicNetwork IT

Thrust 1

Thrust 2

Game Theory

Thrust 3

Optimization

StochasticNetworkAnalysis

CSI, Feedback,and Robustness

Open Questions circa 2006

• Capacity of large dynamic networks

• Capacity of basic network building blocks

Xi

Yi-1

p(yi,si|xi,si-1)

Si-1 Si D

YiTx Rx

• Capacity of time-varying links (with/without feedback)

Progress on these questions

Xi

Yi-1

p(yi,zi,si|xi,si-1)

Si-1 Si D

YiTx Rx

• Capacity of time-varying links (with/without feedback)

• Capacity of finite-state Markov channels with feedback• Converses under unequal error protection• Multiplexing-diversity-delay-distortion tradeoffs in MIMO • Generalized capacity and separation

• Capacity of basic network building blocks

• Capacity region/bounds for Z channel and interference channels• Capacity bounds for cognitive interference/MIMO channels• Upper bounds and converses for interference channels with a

relay (via interference and message forwarding)

• Capacity of dynamic networks

• Network equivalence• Scaling laws for arbitrary node placement and demand• Multicast capacity• Effect of feedback and side information• Dynamic/multiperiod network utility maximization• Generalized Max-Weight policies• Game-theoretic approaches• Mobility for interference mitigation• Delay or energy minimization• Distributed optimization

New Theory• Thrust 1

– Equivalence classes

• Thrust 2– Layered and structured codes– Control principles for feedback channels– Generalized capacity and separation

• Thrust 3– Stochastic Multi-period Network Utility Maximization– Relaxation and distributed techniques for network optimization– Stochastic games

• Interthrust– Relaying, cooperation and cognition– Network coding– Capacity regions for more than 3 users– Coordination capacity

Thrust 0 Recent Achievements

Metrics

Effros, Goldsmith: Expectation and Outage in Capacity and Distortion

Models

Goldsmith: Diversity/multiplexing/delay tradeoffs

Effros: networks with side information

Shah: multicast capacity

Moulin: Mobility

Medard: delay/energy minimization

Zheng: UEP

Medard, Zheng: Distortion-Outage tradeoff

Coleman, Effros, Goldsmith, Medard, Zheng: Channels and Networks with Feedback

Goldsmith: Cognitive Nodes

Cover: Coordinated Networks

Medard: Stability Regions

El Gamal: More than 3 users

New bounding techniques

Thrust 1 Recent Achievements

Code constructionNetwork information theory

Networkingand optimization

Combinatorial Tools

MetricsKoetter, Medard: On the stability region of networks with instantaneous decoding

Goldsmith: multicasting with a relay

Effros: linear code construction

Effros: continuity of network coding regions

Zheng, Medard : distortion-outage tradeoff

Medard: effect of coding versus routing

Goldsmith: capacity and interference rates for the interference channel

Goldsmith: multi-way relay channel

Goldsmith: joint source-channel coding with limited feedback

Cover : Capacity of coordinated actions

El Gamal: more than 3 users

Dynamic Network Information Theory

CSI, feedback, and robustness Structured coding

Thrust 2 Recent Achievements

Effros: two stage polar codes

Coleman: Control principle for feedback channels

Zheng: tilted matching for feedback channels

Medard, Zheng: Diversity-distortion tradeoff

Goldsmith: Joint source channel coding / outage

Effros: linear representation of network coding

Cover: coordination capacity

El Gamal: BC with 3+ receivers

Effros: distributed network coding with coded side information

Goldsmith: Multicast with relay; BC with cognitive relay

Moulin: exploiting mobility of relay networks

Thrust 3 Recent Achievements

Shah: Distributed MAC using queue based feedback

Johari: Large network games

Meyn: Q-learning for network optimization

Boyd, Goldsmith: Wireless network utility maximization as a stochastic optimal control problem

OptimizationDistributed and dynamic

algorithms for resource allocation

Stochastic Network AnalysisFlow-based models and

queuing dynamics

Game TheoryNew resource allocation paradigm that focuses on

hetereogeneity and competition

Ozdaglar: Distributed second order methods for network optimization

Ozdaglar: Noncooperative power control using potential games

Effros: Noncooperative network coding

Medard: Decoding and network scheduling for increased capacity

Ozdaglar: Near potential games for network analysis

Johari: Supermodular games

El Gamal: Overhead in distributed algorithms

FLoWS progress since March• New breakthroughs in upper bounds, feedback and CSI,

cognitive techniques, interference forwarding, multicast traffic, and dynamic/distributed network optimization,

• New synergies within and between our thrust areas

• New/ongoing collaborations among PIs within FLoWS and with Nequit PIs; integration of new PIs Cover and El Gamal

• Overview paper for Scientific American to appear– Co-authors: Effros, Goldsmith, Medard

• Comm. Magazine paper with overview of FLoWS– Submitted and reviewed; likely to be accepted after revision

• Discussion of Phase 3 and 4 progress criteria– Identification of main challenges

• Website updated with March PI meeting slides, recent publications, and recent results.

Focus Talks and Posters

• Thrusts 1 and 2:– El-Gamal: More than Three Users – Cover: Coordination Capacity

• Thrust 2:– Zheng: Tilted Matching for Feedback Channels

• Thrust 3:– Ozdaglar: Near-Optimal Power Control in Wireless

Networks: A Potential Game Approach

• Posters on all recent achievements

Progress Criteria: Phase 31. Revolutionize upper bounding techniques through new and different

approaches that go beyond the classical MIN-CUT bounds and Fano's inequality that have dominated capacity bounds for the last several decades.

2. Determine the optimal channel/network “coding” that achieves these capacity upper bounds when possible, and characterize for which classes of networks gaps still exist between achievability & upper bounds, & why.

3. Develop a generalized theory of rate distortion and network utilization as an optimal and adaptive interface between networks and applications that results in maximum performance regions

4. Demonstrate the consummated union between information theory, networks, and control; and why all three are necessary ingredients in this union

• Progress towards meeting each criteria (more details in Thrust talks)• Identifying “grand challenges” remaining to develop an IT for MANETS

• First pass will be presented in the thrust talks• Will focus on these challenges during Phases 3 and 4• Team meeting Tuesday dedicated to this topic

Project Impact To Date• Recent Plenary Talks

– Boyd: Stevun Lec.’08, CNLS’08, ETH’08, ISACCP’09, ISMP’09, ICOCA’09, CCCSP’09– Goldsmith: Gomachtech’08, ISWPC’08, Infocom’08, RAWC’09, WCNC’09, ICCCN’09– Medard: IT Winter School’08, UIUC Student Conference’08, Wireless Network

Coding’08, ITC.09, ITW’09– Meyn: Erlang Centennial’09, Yale Workshop’09, Diaconis Symp.’09– Ozdaglar: ACC 2009, NecSys'09 , ASMD’08– Johari: World Congress of the Game Theory Society’08– El-Gamal: Allerton’09, Padovani Lecture’09, Brice Lecture’09 – Shah: Net Coop’09, Winedale’09

• Conference Session/Program Chairs/Panels– CTW’09, ITW’09, ISMP’09, INFORMS’09, ITW’10, CTW’10

• Recent Tutorials– Meyn: Mathematics of OR’09, – Shah: CDC’09,

• Invited/award winning journal papers– “Breaking spectrum gridlock through cognitive radios: an information-theoretic

approach”, Goldsmith, Jafar, Maric, Srinivasa, IEEE Proc’09.– “A Random Linear Network Coding Approach to Multicast”, Ho , Medard , Koetter,

Karger, Effros, Shi, and Leong, Joint IT/Comsoc Paper Award 2009.– "XORs in the Air: Practical Wireless Network Coding“, Katti, Rahul, Hu, Katabi, Medard,

and Crowcroft. Bennett Prize in Communications Networking 2009.

Publications to date

• 22 accepted journal papers, 16 more submitted• 127 conference papers (published or to appear)• SciAM paper to appear• Comm. Magazine paper submitted and reviewed• Book on FLoWS vision and results under development

– Alternative to NoW Foundations and Trends article

• Publications website:– http://www.stanford.edu/~adlakha/ITMANET/flows_publications.htm

Summary

• Significant progress in and across all thrust areas• Ongoing and fruitful collaborations between PIs• Powerful new theory has been developed that goes

beyond traditional Information Theory and Networking• Addition of El Gamal and Cover adds new perspective

and experience to our team• Significant impact of FLoWS research on the broader

research community (IT, communications, networking, and control/optimization)

• Want to maximize research impact in the final two phases of the project by identifying key challenges within and beyond the progress criteria