SURFACE Self Configurable Air Interface

1
SURFACESelf Configurable Air Interface
http//www.ist-surface.org
2
The SURFACE vision

• To develop and exploit the latest advances in
  Communication and Information Theory in the field
  of air-interface self configuration
• applying concepts like Trace-orthogonal
  Cyclotomic Space-Time Coding, Dirty Paper Coding,
  Nested Lattice Coding, Cognitive Radio
• and including a proof of concept in 3.9G and 4G
  systems.

3
The SURFACE Road Map
WP 1 Project Management
LONG TERM TECHNOLOGICAL DEVELOPMENT ACTIVITIES
LONG TERM RESEARCH ACTIVITIES
WP 6 Transmit and Receive Architecture Developme
nt
WP 3 Single-User Optimal Transmit and Receive
WP 7 Performance Evaluation
Task 7.1 System Performance Evaluation
Task 7.2 Air Interface Emulator Development
WP 2 System Requirements
WP 4 Multi-User Optimal Transmit and Receive
WP 8 Implementation and Deployment Evaluation
Task 8.1 HW Implementation Optimization
WP 5 Network Optimal Transmit and Receive
Task 8.2 Feasibility and Deployment Study
WP 9 Dissemination and Assessmen t
4
SURFACE Key Issues

• Strategies for channel state information
  estimation, prediction and distribution.
• Air interface self configuration for a MIMO
  uplink (multiple-access) and downlink (broadcast)
  channels for arbitrary channel state information
  knowledge.
• Optimization at network level including Advanced
  RRM.
• Air Interface emulator showing the impact on the
  end-to-end QoS.
• Hardware Optimization and Complexity Evaluation
  of the proposed solutions anticipating hardware
  implementation.

5
(No Transcript)
6
ICASSP 2007 Special Session Proposal

• Non-Linear Precoding for the Broadcast Channel

Resource allocation strategies for the Broadcast
Channel are usually based on assigning a disjoint
set of resources to each of the users to be
served. Using this approach, the per-user
information bearing signals are orthogonal (in
frequency, time or space) and the complexity of
the receivers is kept low by using single user
detectors without interference rejection
capabilities. Traditionally, this
orthogonalization is accomplished by means of
resource division techniques as the basic
strategy to avoid interference but which is often
power inefficient. The Dirty Paper framework
indicates however that for those channels
corrupted by additive interference non-causally
known at the transmitter, the same capacity
results as in the interference free case hold
regardless of the transmit power constraint.
Lattice codes are a means of implementing Dirty
Paper coding techniques whose scalar (simplest)
version is the well-known Tomlinson-Harashima
precoder, one of the best known non-linear
precoding scheme. In this session we plan to
collect some of the most significant recent
contributions in the area of non-linear precoding
for multi-user systems, including issues like
multiple transmit antennas (MIMO), limited
channel state information at the transmit side
and power efficient resource allocation. This
session intends to present results from the
research projects SURFACE (www.ist-surface.org),
MASCOT (www.ist-mascot.org), as well as from
other relevant research groups worldwide.
7
ICASSP 2007 Special Session Proposal

• Invited Authors

Frank A. Dietrich and Wolfgang UtschickPrecoding
with Partial CSI for the Broadcast Channel -- A
General FrameworkEduard Calvo and Javier R.
FonollosaOptimized Resource Allocation in
Downlink MIMO OFDMADominik Seethaler, Gerald
Matz, and Franz HlawatschLow-Complexity
Broadcast Precoding using Seysen's Lattice
Reduction AlgorithmMartin Schubert, Shuying
Shi, and Holger BocheNon-Linear Transceiver
Optimization with Individual MSE
RequirementsEmanuele ViterboBlock Coded
Modulation of the Golden codeMikael Skoglund
and Erik G LarssonOptimal Modulation for Known
Interference