May%202000

1
Summary of activities at ICSI Oct/99 to May/00
May 2000
José M. Páez
2
Introduction
Myself

• José M Páez-Borrallo, Professor of Signal
  Theory and Communications in
• the Technical University of Madrid
• On leave at ICSI in the period Oct/99-Jun/00

Research interests

• Signal and Array Processing for Communications
• Wireless Communications (WLL, WLAN, )
• Physical Layer Modulations, Equalizers, Smart
  antennas
• Access Techniques CDMA, OFDMA, Hybrid, .
• Mobile Digital Systems (DECT, GSM, UMTS,
  TD-UTRA ...)

My original project at ICSI

• Design and analysis of MUD receiver
  architectures for CDMA. Exploiting
• dynamic Space-Time diversity to improve the
  radio link

3
Activities carried out during my stay at ICSI
Research

• Design and analysis of MUD receiver
  architectures for CDMA with direct
• application to UTRA-TDD
• Design of new waveforms for CDMA. Something
  intermediate between
• DS-CDMA and MC-CDMA
• Prospective analysis and information gathering
  on Ultra Wide Band Radios

Definition of future infrastructure for Research

• Design and viability study of a Research
  Institute on Communications
• with partnership of Industry, Telecom
  Operators and University

4
Contacts in Berkeley during my stay at ICSI
Cooperation with BWRC

• Berkeley Wireless Research Center (BWRC).
  http//bwrc.eecs.berkeley.edu/
• Involved in technologies for new generations of
  radio systems
• Cooperation in new CDMA techniques
  (October-February)
• Continued assistance to internal BWRC seminars
  and tech discussions
• Invited tutorial on MC Modulation in BWRC
  Winters Retreat in Monterey, CA
• Poster on new waveforms for CDMA in BWRC
  Winters Retreat
• Contact at BWRC Scientific Director, Robert
  Brodersen

5
Multiple Access Sharing the medium
Fix services
Mobile services
Multiple Access Defined

• A scheme that enables many users to share a
  common access point
• Requires orthogonalization of users so that
  they can be distinguished
• Orthogonalization can be done in in time,
  frequency, space, or any
• combination (e.g. using codes)

6
Multiple Access CDMA vs. TDMA and CDMA
7
MUD Receiver Architectures for CDMA systems
8
MUD receiver architectures for CDMA (1)

• Definition of an extended multisensor signal
  model for Multiple Access
• Interference MAI including multipath
• Design and analysis of new front-ends with MAI
  cancellation
• Design and analysis of Decorrelation and
  Optimal (MMSE) vector receivers
• Performance evaluation of previous receivers
  (mono and multisensor)
• MATLAB implementation of a radio-link simulator
  with channel impairments
• Adaptive implementation of receivers.
  Reduced-rank versions of it
• Analysis of simulated results versus receivers
  complexity

9
MUD receiver architectures for CDMA (2)
MUD efficiency It quantifies the amount of
Intracell Interference removal
performed by the specific MUD receiver
Cancelable Total Interference ratio
10
MUD receiver architectures for CDMA (3)
A feasible vector MUD Decorrelation receiver
architecture
Front-End
Decorrelation
User 1, L paths
1
User 1
Signature space-time Decorrelation
Space-time filtering
. . .
P
User K
User K, L paths
KL
KL
KL
KLP
1
Multipath amplitude estimation
DOA estimation
To downlink in TDD systems
KL
P
Channel estimation branch
Front-end Capture the incoming active
users Channel branch Estimate space-time info
from the incoming signals Space-time filters
Isolate different users and paths Decorrelation
Remove the residual MAI from others users (eff.
Signature)
11
MUD receiver architectures for CDMA (4)
A feasible vector MUD MSE receiver architecture
Minimize MSE
Front-End
Filtering
GOAL
User 1, L paths
KLP
K
Rank reduction
. . .
User filters
User K
User K, L paths
K
K
KLP
Training sequence
Computing matrix transformation
Further refinement
Channel estimation branch
Front-end Capture the incoming active
users Channel branch Estimate rank-reduction
matrix from space-time info Space-time filters
Isolate different users and recover the desired
signals Training Provides the info sequence to
form the goal function
12
MUD receiver architectures for CDMA (5)
Some simulated results
Spreading gain 15 1000 QPSK symbols 5 active
users 2 paths per user (span 1/4 symbol) Linear
array of 8 sensors 4 different architectures
13
New TF waveforms for CDMA
14
New TF waveforms for CDMA CDMA concept (1)

• Any symbol is carried by a set of waveforms
  (basis)
• This set of carriers is needed to provide
  virtual separation among users
• Can choose any orthogonal basis -- each user is
  not limited to one time or frequency slot
• This generalization results in a code space
  where users may overlap in frequency/time
• Much research has been done and many different
  flavors exist DS-CDMA, MC-CDMA, etc.

where
continuous
chips
basis (waveforms)
symbol
discrete
where
15
New TF waveforms for CDMA Standard DS-CDMA (2)
16
New TF waveforms for CDMA Multicarrier CDMA (3)
17
New TF waveforms for CDMA Time-frequency chips
(4)
18
New TF waveforms for CDMA Example (5)
DS-CDMA is tiled in time and is susceptible to
distortion caused by time-dispersive
channel MC-CDMA is tiled in frequency is
susceptible to channel variations and frequency
noise/jitter
Discrete chirps
Waveform set discrete chirps
Long waveforms
-1
-1
1
1
2D Code chips
-1
-1
1
1
-1
1
1
1
T4Tc
Time
Symbol
Chips are time-frequency located
Provides intermediate level of resistance to time
and frequency impairments
19
Research Institute on Communications
20
Research Institute on Communications (1)
Foundation statements

• Provide an environment for research into the
  issues necessary to support Technology, Networks
  and Applications for future mobile communication
  systems.
• Center for education of new engineers in the
  field of Wireless Technologies and Services
• Mixed participation of Industry, Telecom
  operators and University
• Promote start-up companies from the Institute
  on own results
• The Institute will pursue an active policy of
  placing research results in the public domain as
  determined by the Scientific Board

Center lifetime of at least 7 years with yearly
reviews and 43 year informal commitments
21
Research Institute on Communications (2)
Specific Projects Consortia
Medium and long term RD
Ph D Theses Master Theses
Specific needs Multidisciplinary projects
Institute
Utilization of known RD Continuous Analysis of
extern RD Prediction of tech needs
Product Focalization Own development in the CI
Creation of start-ups
Prospective Innovation
22
Research Institute on Communications (3)
Research issues at the Institute
Prospective Innovation Office
3 research areas
This is a 1st proposal. It has to be
determined within the Advisory Board
23
Research Institute on Communications (4)

• 2 Workshops per year
• Research staff is committed to present their
  results publicly
• Monographic sessions, Poster presentations and
  Tech discussions
• Open assistance to extern public (other
  companies, institutes, etc)
• 2 montly Seminars
• Given by CI staff, company partners and
  invited speakers
• Semester calendar of seminars (as continuing
  education for some companies)
• Non-partners assistance by invitation or
  payment
• 2 montly technical meetings per research area
• Development of research agenda and summary of
  work done
• Paper Reading and discussion sessions
• Restricted assistance to associated partners

24
Research Institute on Communications (5)
Initial Participation (commitment for 4 years)
Associated
Affiliated
Funding
150 Keuro/year
50 Keuro/year
Fix lab seats up to 2 engineers
Only for Seminars
Human resources
Types of Company Partnership
25
Research Institute on Communications (6)
University
Head Telecom Faculty Head Dpt. Radiotech Head
Dept. Electronics Head Dept. Networks Head Dept.
Basic Sci
Industry
10-12 Associated Companies
Institute
Head of Institute Vicehead of Institute Technology
Responsible Network Responsible Application
Responsible
Individual
4 known professionals
Institute Advisory Board
26
Research Institute on Communications (7)
Benefits of being an Institute partner
Asso.
Affi.

• Member of the Advisory board
• Member of the Scientific board (Impact on the
  research agenda)
• Participation in large pre-competitive,
  interdisciplinary research
• effort with relatively modest investment
• Free Assistance to workshops and seminars
• Free assistance to periodic technical meetings
• In-situ involvement with a large number of
  outstanding students
• and staff (1-2 engineers per year)
• Co-utilization of lab installations and tech
  resources
• First hand knowledge of Center results and free
  development of them
• Unlimited explotation rights of developed
  technologies and prototypes
• Direct access to a number of graduates who have
  expertise in the area
• of mobile communications

27
Research Institute on Communications (8)
University hired personal and infrastructure
funding
3.5 Meuros
Others
Univ 0.89 (25) Taxes 0.27
(8) Misc. 0.19 (5)

1.35 (38)
28
Research Institute on Communications (9)
Institute permanent staff (estimation)
From University
Hired Personal
4 Faculty staff
1 MBA
Exec. team (5)
13 Faculty staff (4 per area 1) 12 PhD
students (4 per area) 24 Master students (8
por área)
14 Doctors (4 per area2) 12 Industry engineers
Resarch staff (82)
4 engineers 5 Administrative staff
Administr. support staff (13)
3 Grad. students
Total 56 (61)
Total 36 (39)
Employment creation 24/80 (30)
29
Research Institute on Communications (10)
Feasible funding sources
3.5 Meuros