Student Member, IEEE

Download Report

Transcript Student Member, IEEE

Network Service Provisioning in UWB
Open Mobile Access Networks
Dario Di Sorte, Student Member, IEEE,
Mauro Femminella, Student Member, IEEE,
Gianluca Reali, Associate Member, IEEE, and
Sven Zeisberg, Member, IEEE
IEEE JOURNAL ON SELECTED AREAS IN COMMUNICATIONS, VOL. 20, NO. 9, DECEMBER 2002
Pages 1745-1753
Outline
WHYLESS.COM
Open Mobile Access Network (OMAN)
Requirements for PHY
UWB as PHY
UWB Domain Parameters
Current work
WHYLESS.COM
Cofunded by European Union
Current paradigmatic chain


Infrastructure – Service – Terminal (user)
Rigid: prevent a quick response to changing user needs
and business models
Open Mobile Access Network (OMAN) paradigm



Efficient information transport
Modern, small scale, electronic business, evolutionary
growth
Decouples user, info transport resources, content
provision service
Goal: ‘standardize’ a new commodity dealing with
the transportation of electronic information
OMAN Concepts
Administratively independent IP domains

external characterization of traffic in terms of QoS
parameters
Virtual delay (d)


Network commodity
Standard measure of the level of the service
Differentiated Services (DiffServ) compliant
approach to guarantee QoS (RFC 2475)
Usage-based pricing


Actually used and/or reserved network resources
QoS charge beyond the flat access charge
OMAN Network Model
Entities involved
End Users

Consumers needing service
Network Service Providers (NSPs)

Provide the network infrastructure
Network Resource Manager (NRM)


Manage the domains owned by the NSP
Guarantee a set of edge-to-edge services over domain
Application Service Providers (ASPs)

Application service and IP connectivity to End User
Information Brokers (IBs)


End User and ASP mediator
Identifies ASPs and provides ASPs history
Resource Brokers (RBs)

ASP and NSP mediators
OMAN Reference Environment
End to End Service Provisioning
Procedure
End user (through IB)
Identifies an ASP for a specific service
Requests service (quality of desired service and willingness to pay)
End to End Service Provisioning
Procedure
ASP
Transforms customer’s qualitative requirements into quantitative technical parameters
Requests RB to find the “best” path to deliver the application service (eBDL)
End to End Service Provisioning
Procedure
RB
Identifies the potential domains and checks their offers with e-Table
Runs an interdomain routing algorithm
e-Table
f(d) : Technical cost of information transfer in commodity
unit

f(∞) = 0
α : Cost of each commodity unit

Network parameters, end points, domain policies
γ : Price variation factor

Market fluctuations
Bres,i = xiPs (0<xi<1)
End to End Service Provisioning
Procedure
RB -> ASP
Path found
Teriffs charged
End to End Service Provisioning
Procedure
ASP -> End user : Relevant price
End user -> ASP : Accept/Reject
ASP -> Each NAP involved : Service contract
PHY Requirements
Advanced and scalable radio transmission
technology
Heterogeneous traffic with specific QoS

PHY should not be a bottleneck
Short-term trading of wireless and wired
information transport resources




Wireless: Ownership not clear
On demand spectral usage shared with others
Synchronization probs (TDMA)
Combine resources with single radio front end
(FDMA)
PHY Layer for OMAN
PHY layer requirements:




Single PHY channel has to support a wide range of net data rates
Re-configurable on-the-fly, to maintain QoS
Support continuous and packet-oriented channels
Mobility of terminals has to be considered.
PHY layer functionality




Coding according to the appropriate FEC scheme
Pulse modulation by the appropriate mapping of the encoded bits
onto symbols
Insert/remove pilot symbols/preamble to/from the symbol train
Perform radio transmission and reception including
synchronization
UWB Characteristics
Large data rates over short distances without
allocating dedicated spectral resources
Software controllable parameters


Adapt bit rate according to terminal location,
propagation condition and service requirements
Enables highly resilient, scalable and flexible
networks
Other features




Low mean transmission power
Through the wall penetration
Precision location
Minimized hardware complexity
UWB for OMAN
Unsynchronized UWB
Fulfills wide range of requirements




End user service contracts
Dynamic domain requirements from NRM
Broad range of data rates
Classes of services
No sophisticated cell planning

Same frequency
On the fly modification of transmission
parameters

Processing gain, TH sequence, duty cycle, temporal
pulse shape, code rate
IR-UWB
Spread Spectrum technology with BW in GHz
Impulse Radio



Best studied
Relatively simple implementation
Large crest factor but with low average power
Whole band to every service provider




Improved spectral efficiency
Wider range of rates possible
Same equipments for all providers
Disadvantages:
Synchronization not possible
Power monitoring for billing
Less affected by fading
PHY design considerations
IR with FCC regulations
Centralized wireless network
Up- and down-link have the same air
interface definition
Only continuous pulse transmission with
hard handovers are currently considered
UWB Domain: Per Domain Behavior
NRM functions


Admission control able to support mobility
Provide bandwidth estimation
LMG: Local Mobility
Gateway
UWB Domain: Admission Control
Function
Guage&Gate Reservation with Independent Probing
(GRIP)


Verifies RB claims before NRM stipulates contract
Handoffs
UWB Domain: Bandwidth
Estimation
Discovery probe packets

Tradeoff: Overhead v/s Information accuracy
Per Domain Behavior (PDB) table
UWB Transmitter
Design parameters


Cannot be changed once the project is finalized
w(t), Tm, Tc, Na, cn,k, dn,k, Np, B
Control parameters


Can be varied by MAC
Em, Ns, Nh, Nd
Derived parameters


Values dependent on Control parameters
Tf, Ps, Nc, Ts, M, K, Rs, Rb, Es, Eb
Current Work
Performance degradation of UWB systems can
be analytically predicted
UWB DS-IR more robust against existing radio
technologies
Controlled licensed mode for higher
transmission power
Spectral shaping to avoid interference


TH codes investigated
Spectrogram based technique + threshold detector
Turbo encoding was found to improve
performance
References
www.whyless.org
Deliverable D5_2b, “Transceiver
architecture and algorithms”, Jan 2003
Deliverable D5_3b, “Physical layer
architecture and performance”, Jan 2003
Deliverable D5_4b, “Ultra-Broadband
coexistence (final)”, Dec 2003