BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:Fundamental Rate-Reliability-Complexity Limits in Outage Limited M
 IMO Communications  
DTSTART:20110615T141500
DTSTAMP:20260407T051134Z
UID:e0bca880cff7e19cb47f71a01b81d8a996087b34fbfa05e079cf3976
CATEGORIES:Conferences - Seminars
DESCRIPTION:Prof. Petros Elia\, EURECOM\nThe work establishes fundamental 
 limits with respect to rate\, reliability and computational complexity\, f
 or encoding-decoding in a general setting of outage-limited MIMO communica
 tions. In the high-SNR regime\, the limits are optimized over all encoders
 \, all decoders\, and all complexity regulating policies. The work then pr
 oceeds to explicitly identify encoder-decoder designs and policies\, that 
 meet this optimal tradeoff. In practice\, the limits aim to meaningfully q
 uantify different pertinent measures\, such as the optimal rate-reliabilit
 y capabilities per unit complexity and power\, the optimal diversity gains
  per complexity costs\, or the optimal number of numerical operations (i.e
 .\, flops) per bit. Finally the tradeoff's simple nature\, renders it usef
 ul for insightful comparison of the rate-reliability-complexity capabiliti
 es for different encoders-decoders. In the same setting we identify the fi
 rst ever lattice decoding solution that achieves both a vanishing gap to t
 he error-performance of the exact solution of lattice decoding\, as well a
 s a computational complexity that is subexponential in the rate and in the
  problem dimensionality. This complexity behavior is rigorously demonstrat
 ed here for the first time\, where it is also proven that for most common 
 codes\, the complexity cost for asymptotically optimal regularized lattice
  (sphere) decoding is exponential in the rate\, and in many cases it in fa
 ct matches the complexity cost of ML (bounded) sphere decoding. In light o
 f the fact that\, prior to this work\, a vanishing gap was generally attri
 buted only to full lattice searches of exponential complexity\, whereas su
 bexponential complexity was generally attributed to early-terminated (line
 ar) solutions having a performance gap that can be up to exponential in di
 mension and/or rate\, the work constitutes the first proof that subexponen
 tial complexity need not come at the cost of exponential reductions in lat
 tice decoding error performance. Finally\, the performance and complexity 
 guarantees in this work hold for the most general MIMO setting\, for all r
 easonable fading statistics\, all channel dimensions\, and all lattice cod
 es. Prof. Elia's homepage
LOCATION:BC 01 https://plan.epfl.ch/?room==BC%2001
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR