INFOWARE-2011-MObile computation- panel- Intro-EB-v1.0€¦ · Challenges in Signal Processing,...
Transcript of INFOWARE-2011-MObile computation- panel- Intro-EB-v1.0€¦ · Challenges in Signal Processing,...
INFOWARE 2011 Luxembourg, June 2011
INFOWARE 2011PANEL 1: Mobile Computation
Challenges in Signal Processing, Global Informationand Mobile Communications
INTRODUCTION
Eugen Borcoci, University Politehnica Bucharest
INFOWARE 2011 Luxembourg, June 2011
PANEL
Mobile ComputationChallenges in Signal Processing, Global Informationand Mobile Communications
Moderator:Eugen Borcoci, University Politehnica of Bucarest, Romania
Guest Panelists:Dragana Krstic, University of Nis, SerbiaGyörgy Kálmán, ABB AS - Akershus, NorwayYasunori Iwanami, Nagoya Institute of Technology, JapanEugen Borcoci, University Politehnica Bucuresti, Romania
Slide 3
INFOWARE 2011 Luxembourg, June 2011
Panel topics
� Short presentations:
� Dragana Krstic: Methods for reducing the impact of fading in mobile telecommunications
� György Kálmán: Bearer or Service Provider: Future role of the Mobile Operator
� Yasunori Iwanami: Coded MIMO OFDM
� Eugen Borcoci: Service mobility- case study example
� Q/As
Methods for reducing theimpact of fading in mobiletelecommunications
Dragana KrstićFaculty of Electronic Engineering,
University of NišAleksandra Medvedeva 14, 18000 Niš, Serbia
ICWMC 2011
Wireless channels are simultaneouslyaffected by short-term fading and long-term fading (shadowing).
In wireless communication systems,various techniques for reducing fadingeffect and influence of shadow effect areused:
ICWMC 2011
They are:
- diversity reception,
- dynamic channel allocation and
- power control.
Upgrading transmission reliability andincreasing channel capacity withoutincreasing transmission power and bandwidthis the main goal of diversity techniques
ICWMC 2011
We are studing wireless communicationsystem following microdiversity to mitigate theeffects of fast fading and macrodiversityprocessing to reduce shadowing effects.
Fast signal variations are described byseveral distributions such as Rayleigh, Rice,Nakagami-m, α-μ, Weibull and Hoyt.
ICWMC 2011
Considering Nakagami-m and Hoytdistribution multipath scattering with relativelylarge delay-time spreads, with differentclusters of reflected waves are described.
Also the consideration of propagation paths,consisting of one strong direct line-of sight(LoS) signal, and many randomly reflected,usually weaker signals is included trought theRicean distribution.
ICWMC 2011
The non-linearity fading effects are includedtrought consideration of Weibull and α-μfading distributions.
If base stations at the macrodiversity level arewidely located, due to sufficient spacingbetween antennas, long-term fading can bemoddeled with joint distribution of twostatistically independent Gammadistributions.
ICWMC 2011
The contribution of this analysis fordualbranch macrocombiner, is that it hasbeen done for general case of α-μ(Generalized Gamma) distribution, whichincludes as special cases important otherdistributions such as Weibull and Nakagami-m (therefore, the One-Sided Gaussian andRayleigh are also special cases of it).
ICWMC 2011
Diversity reception, based on using multipleantennas at the receiver (space diversity, withtwo or more branches) is a very efficientmethod used for improving system’s quality ofservice
Multiple received copies of signal could becombined on various ways.
ICWMC 2011
System model 1
ICWMC 2011
System model 2
r,rp r,r
SCmicrodive
rsitySC
macro
diversity
SCmicrodiv
ersity
© ABB GroupJuly 14, 2011 | Slide 1
Bearer or Service ProviderFuture role of the Mobile Operator
György Kálmán, ABB Corporate Research Norway
Status
Market trends:
Smartphones in all segments
Content consumption and creation is increasing
Lower data traffic charges allow continous onlineservice access
Online payments are usual for consumers
Two IT companies have taken over the completecontrol of the phone market
© ABB GroupJuly 14, 2011 | Slide 2
Role of the Provider
Past: the degradation of the fixed telephony network to abearer service, mobile provider sells VoIP for circuitswitched price
Present: VoIP over 3G data is available
Regulatory and provider-initiated restrictions may bepresent
Continous internet access common
Future?
Bearer (e.g. iTunes) or content provider (e.g. TelenorMobileTV) or trusted party (e.g. BankID)
© ABB GroupJuly 14, 2011 | Slide 3
Trust, payments and user experience
Key features!
The trust relationship already exists between the providerand the subscriber
Payments: telecom standardization is slow, missing theopportunity window.
Smartphones can access ”desktop” services with gooduser experience
Do we need more complex hardware? – Paypal,webbank
© ABB GroupJuly 14, 2011 | Slide 4
Questions
Where is the place of the mobile provider?
In the payment area, is it worth to use trusted hardware?
Is NFC the future for payments or missed the window?
What can the telco industry learn from IT?
Future of standardization?
© ABB GroupJuly 14, 2011 | Slide 5
© ABB GroupJuly 14, 2011 | Slide 6
INFOWARE 2011Conference, June 19-24, 2011, Luxembourg
INFOWARE 2011
Panel: Mobile Computation
Challenges in Signal Processing, Global Informationand Mobile Communications
Mobility issuesService Mobility- case study
Eugen Borcoci, University Politehnica Bucharest
Slide 2
INFOWARE 2011Conference, June 19-24 Luxembourg
Challenges in Signal Processing, Global Informationand Mobile Communications: Mobility
� Mobility� Terminal mobility (IP based solutions)
� Technologies: IEEE 802.11x, 15x, 16x, LTE, 3G, …� Solutions
• Mobility solved at L2/L3 layer
• solved at higher layer ( e.g SIP supported mobility)
� Within the same L1-L2 technology• Horizontal mobility
Micro-mobility (L2)Macro-mobility (L2+L3)Handover styles :
Hard HO (Break before make)Soft/seamless HO (session preservation)
� Q: what are the major unsolved problems related to terminal
mobility in horizontal HO?� Between different L1-L2 technologies
• Vertical mobility- more challenging
• New solution (MIH) Media Independent HOQ: How much is currently deployed ?
Slide 3
INFOWARE 2011Conference, June 19-24 Luxembourg
Challenges in Signal Processing, Global Informationand Mobile Communications: Mobility
� Mobility
� Service/personal mobility� End user (EU) has registered for some services to an SP� EU can get registered services
• conforming his/her user profile and SLA concluded with SP• through a residential gateway
� EU is visiting another (IP) network� Service mobility:
• EU can get the same services with the same user profile• Using its portable terminal or a fixed terminal in the visited network
• Through the visited gateway
� Q: Can be the service mobility better served by integrated mobility solutions?
� New issue (also in FI-oriented studies) : � Identity/location decoupling attempts/solutions� What is the impact of this approach on mobility ?
Slide 4
INFOWARE 2011Conference, June 19-24 Luxembourg
Challenges in Signal Processing, Global Informationand Mobile Communications: Mobility
� Case study example: Service Mobility
� The system: MediA Ecosystem Deployment Through Ubiquitous Content-Aware Network Environments ALICANTE, 2010-2013, Integrated Project (IP)
� Complete system ( network + services) - media oriented, multiple domain, …� Virtual networking: parallel virtual planes customised for different
type of media content ( based on CA properties)
� Business Entities:� End Users (EU)� Service/Content Providers� Virtual network Providers� Infrastructure Network providers
� EU connection – to the core network� via Home Boxes ( enhanced Residential GWs)
Slide 5
INFOWARE 2011Conference, June 19-24 Luxembourg
Challenges in Signal Processing, Global Informationand Mobile Communications: Mobility
� Case study example: Service Mobility
� The system: MediA Ecosystem Deployment Through Ubiquitous Content-Aware Network Environments ALICANTE, 2010-2013, Integrated Project (IP)
Slide 6
INFOWARE 2011Conference, June 19-24 Luxembourg
Challenges in Signal Processing, Global Informationand Mobile Communications: Mobility
� Case study example:
� Service Mobility
1. Initial SLA EU1-SP
2. EU1 move to HB23. 3-..8 : service
handover
9. Media consumption
in the same way as at home
AccessNetwork 1
AS1
HB (home
AS2
CANP NP
CANP NP
CAN layer
move
Multi-domain VCAN
HB Mgmt
1
HB + SP Environment
EU1 host
PHY connection
Contract (SLA/SLS)
or interaction
MANE MANE
AR
SP/CP
9. Media flow consumption
MANE MANE
ANP
Service/Users Management
2
HB2
(visited)
EU1’ host
AN2
Content Server
3
4
5
6
7
8
Slide 7
INFOWARE 2011Conference, June 19-24 Luxembourg
Thank you
On coded MIMO OFDM
- How to obtain spatial and frequency diversity effects -
Yasunori Iwanami
Nagoya Institute of Technology
Nagoya, Japan
IFF
TIF
FT
IFF
TIF
FT +CP
Transmitter
P/S
FF
TF
FT
FF
TF
FT
ZF
Nu
lling
, MM
SE
Nu
lling
or M
LD
Receiver
+CP
+CP
+CP
- CP
- CP
- CP
- CP
S/P
OFDM demodulation for each Rx antenna
OFDM modulation for each Tx antenna
Transmitter and Receiver configuration of MIMO OFDM
Av
era
ge
po
wer
t
Quasi-static Rayleigh
fading with equal
power L delay paths GI
0
(0)ijh
/4T
(1)ijh (2)ijh ( 1)ijh L
Delay profile between each Tx and Rx antenna
Input data bits
QA
M M
od
.LD
PC
e
nco
din
g
Input data bits
Uncodedcase
LDPCcoded case
LD
PC
d
eco
din
g
QA
M D
em
od
.
Output data bits
Output data bits
Uncodedcase
LDPCcoded case
Soft value
1.0E-05
1.0E-04
1.0E-03
1.0E-02
1.0E-01
1.0E+00
0 5 10 15 20 25 30
Bit
Err
or
Rat
e
Average Eb/N0 per receive antenna [dB]
4×4-OFDM ZF Nulling
2×2-OFDM ZF Nulling
4×4-OFDM MMSE Nulling
2×2-OFDM MMSE Nulling
1×1-OFDM ZF Nulling
1×1-OFDM MMSE Nulling
1×1-OFDM MLD
Flat Rayleigh Theory
2×2-OFDM MLD
4×4-OFDM MLD
BER characteristics for uncoded MIMO OFDM
BPSK modulation
1
0 0 0
1 111 1/ 1 , 1( / )2 4
b bb
b
E EP
E N N N
210 / 5 (dB)
Diversity order=4
110 /10 (dB)
Diversity order=1
110 / 5 (dB)
Diversity order=2
1E-05
0.0001
0.001
0.01
0.1
1
0 5 10 15 20 25 30 35 40
BE
R
Average Eb/N0[dB] / receive antenna
1×1 OFDM
2×2 OFDM
4×4 OFDM
4×4 MC-CDM
2×2 MC-CDM
1×1 MC-CDM
MIMOMC-CDMA
MMSE Nulling
MIMOOFDMMMSENulling
BER characteristics for uncoded MIMO MC-CDMA
BPSK modulation
A
A
A
Uncoded:BPSKLDPC coded:QPSKL=16 equal power delay pathsNumber of subcarriers =64CP length=T/4
(n, k)=(1024,512) Regular LDPC code(j, k)=(4,8)Loop 4,6 excluded
Theory
About 7.5dB improvement
BER characteristics for LDPC coded MIMO OFDM with MLD
Flat Rayleigh
ZF, MMSE or MLD uncoded
• By using MLD, spatial diversity effect is obtained.
• By using MC-CDMA, frequency diversity effect is
obtained.
• By using LDPC code with MLD, spatial and frequency
diversity effects with coding gain is obtained.
Obtained results