1
The Power of Non-Uniform Wireless Power
ETH Zurich – Distributed Computing Group
Magnus M. HalldorssonReykjavik University
Stephan HolzerETH Zürich
Pradipta MitraReykjavik University
Roger WattenhoferETH Zürich
Presented by Klaus-Tycho Förster Slides by Stephan Holzer and Roger WattenhoferETH Zürich
ETH Zurich – Distributed Computing Group Stephan Holzer SODA 2013
Wireless Communication
ETH Zurich – Distributed Computing Group Stephan Holzer SODA 2013
Wireless Communication
EE, PhysicsMaxwell EquationsSimulation, Testing‘Scaling Laws’
Network Algorithms
CS, Applied Math[Geometric] GraphsWorst-Case Analysis
Any-Case Analysis
ETH Zurich – Distributed Computing Group Stephan Holzer SODA 2013
InterferenceRange
CS Models: e.g. Disk Model (Protocol Model)
ReceptionRange
ETH Zurich – Distributed Computing Group Stephan Holzer SODA 2013
5
ETH Zurich – Distributed Computing Group Stephan Holzer SODA 2013
EE Models: e.g. SINR Model (Physical Model)
ETH Zurich – Distributed Computing Group Stephan Holzer SODA 2013
ETH Zurich – Distributed Computing Group Stephan Holzer SODA 2013
Signal-To-Interference-Plus-Noise Ratio (SINR) Formula
Minimum signal-to-interference
ratio
Power level of sender u Path-loss exponent
Noise
Distance betweentwo nodes
Received signal power from sender
Received signal power from all other nodes (=interference)
ETH Zurich – Distributed Computing Group Stephan Holzer SODA 2013
The Capacity of a Network(How many concurrent wireless transmissions can you have)
ETH Zurich – Distributed Computing Group Stephan Holzer SODA 2013
… is a well-studied problem in Wireless Communication
The Capacity of Wireless NetworksGupta, Kumar, 2000
[Toumpis, TWC’03]
[Li et al, MOBICOM’01]
[Gastpar et al, INFOCOM’02]
[Gamal et al, INFOCOM’04][Liu et al, INFOCOM’03]
[Bansal et al, INFOCOM’03]
[Yi et al, MOBIHOC’03]
[Mitra et al, IPSN’04]
[Arpacioglu et al, IPSN’04]
[Giridhar et al, JSAC’05]
[Barrenechea et al, IPSN’04][Grossglauser et al, INFOCOM’01]
[Kyasanur et al, MOBICOM’05][Kodialam et al, MOBICOM’05]
[Perevalov et al, INFOCOM’03]
[Dousse et al, INFOCOM’04]
[Zhang et al, INFOCOM’05]
etc…
ETH Zurich – Distributed Computing Group Stephan Holzer SODA 2013
The Capacity of a Network(How many concurrent wireless transmissions can you have)
• Power control helps: arbitrarily better than uniform power (worst case)
• Arbitrary power: O(1)-Approximation Complex optimization problem
• Simpler ways?
ETH Zurich – Distributed Computing Group Stephan Holzer SODA 2013
Oblivious Power
• Power only depends on length of link• Mean power has “star status” O( + )-approximation [SODA’11] [ESA’09]
Δ=𝑚𝑎𝑥𝑖𝑚𝑎𝑙 h𝑙𝑒𝑛𝑔𝑡𝑚𝑖𝑛𝑖𝑚𝑎𝑙 h𝑙𝑒𝑛𝑔𝑡
(usually: small constant)Essentially:O( ) vs.
This Paper𝒑 𝜶 :𝒑∈(𝟎 ,𝟏)
ETH Zurich – Distributed Computing Group Stephan Holzer SODA 2013
Old Definitions
• Affectance
• SINR-condition is now just
• p-power: assigns power to link l
v w
ETH Zurich – Distributed Computing Group Stephan Holzer SODA 2013
New Crucial Definitions
• Length-ordered version of symmetric affectance:
• Interference measure:
𝐼𝑄𝑃 (𝐿)
ETH Zurich – Distributed Computing Group Stephan Holzer SODA 2013
Structural Property
Let be a p-power power-assignment and be an arbitrary power-assignment, then *
*= for non-weak links
ETH Zurich – Distributed Computing Group Stephan Holzer SODA 2013
Algorithm
Links increase by length For i=1 to n do
If then
Yields -approximation for capacity.Analysis uses
ETH Zurich – Distributed Computing Group Stephan Holzer SODA 2013
Applications
Connectivity: given a set of nodes, connect them in an interference aware manner.
Strongly connected in + time slotsusing mean power.
Centralized and distributed algorithms
any p-power
ETH Zurich – Distributed Computing Group Stephan Holzer SODA 2013
Applications
Distributed Scheduling: schedule a given set of links in a minimal number of time slots.
There is randomized distributed + -approximationto scheduling using mean-power. any p-power
ETH Zurich – Distributed Computing Group Stephan Holzer SODA 2013
Applications
Spectrum Sharing Auctions: channels, users, each user has valuation of each subset of channels.Find: allocation of users to channels such that each channel is assigned a feasible set and the social welfare is maximized.
-approximation
ETH Zurich – Distributed Computing Group Stephan Holzer SODA 2013
Summary
SINR-model capacity max.Length-oblivious
power-approximation
3 (out of >5) applications
ETH Zurich – Distributed Computing Group Stephan Holzer SODA 2013
Thanks!