Roma 17/10/08 WORLD Project KO Meeting Laura Galluccio WORLD Project – KO Meeting University of Catania.

Roma 17/10/08WORLD Project KO Meeting

Laura Galluccio

WORLD Project – KO MeetingUniversity of Catania


Laura Galluccio

Research Activities (AR)• UNICT proposal

• WP1: System Architecture– T1.2: WORLD system plan

• WP5: Cross-layer Wireless Multimedia and Active QoS-Mapping– T5.2: Opportunistic access and node auto set up– T5.3: Game-Theory-based distributed radio resource management

• WP6: Applications and secure End-to-End QoS-provisioning (leadership)– T6.2: Opportunistic scalable video and audio coding for adaptive streaming

applications– T6.4: Multimedia Peer-to-Peer (P2P) WLAN connections

T1.2: World System Plan

We will collaborate to the definition of• System and network architecture

(Context awareness and reconfigurability techniques, QoS support mechanisms)

• Protocol architecture

(Cross layer interactions, QoS transversal support)


Laura Galluccio

T5.2: Opportunistic Access and Node auto set up


Laura Galluccio


• Nodes opportunistically enter and exit the P2P network

• Mechanisms to let nodes self-reconfigure and discover other devices and services will be needed

• Discovery procedures should be characterized both in terms of performance and algorithms


• To automatically set up nodes, it will be required to:

– Localize

– Distribute a catalogue of the available resources


Laura Galluccio

Localization• Infostation nodes are also available into the network• Some nodes are equipped with GPS• Other nodes do not• Location information is needed to implement a

distributed algorithm for resources propagation into the network

• We will develop a middleware allowing nodes localization through some fixed position-aware nodes


Laura Galluccio

Localization• This middleware will allow to

– Perform an algorithm to let the nodes calculate their position exploiting position-aware nodes

• Send beacon packets• Use these packets to estimate the packet loss

probability (PLP)• Use the PLP to evaluate the bit error probability (BEP)• Estimate the SNR• Obtain the distribution of the SNR


Laura Galluccio

Resources catalogue: A mobile world

• Rapid increase in the amount of multimedia contents created “on the road”

• Pervasive use of personal portable devices

• Users typically share contents by non-mobile systems (P2P sharing program; social network)

• User friendly interface to easily share and exchange data through heterogeneous information systems and communication technologies

in the near future

in the near future

butbut

What is our target?

• To provide a general purpose platform whose aim is to create data sharing services in mobile environment in a modular and adaptive way

• Some scenarios:– Music festivals: where people could share contents (photos, songs, videos)

– Municipal services

– Cultural activities

– Mobile social network

Existing solutions

• Several works in literature deal with the problem of data sharing in mobile environments

• Solutions based on novel protocols or different transmission techniques are proposed to solve the architectural and technical problems:

– Web-service based platform (Internet access is required)– Instant messaging architecture (server based scheme)– Mobile P2P solutions (using the cellular network)

• The majority of the existing solutions are related to specific applications and use an infrastructured scheme

Idea

ShareAnywhere (SA) will be a platform which provides an

application framework where mobile devices make use of user-

friendly methods to share and exchange digital contents among

heterogeneous systems and through different communication

technologies.

Logical Architecture

Mobile Nodes: Cellular phones, PDAs and Notebooks running an appropriate software

InfoStations: Network nodes revealing the presence of mobile nodes and providing resources

CMS:Non-mobile system front-end to upload the resources

Database: element storing all the system resources

T5.3: Game-Theory based distributed radio resource management


Laura Galluccio


• Infostations will be used by mobile nodes to download the required information when located in the proximity

• In case of multimedia applications playout buffers length should be taken into account when considering the possibility to download frames during the limited proximity time

• Requests being sent to the infostation nodes should be managed according to a priority-like mechanism taking into account the limited proximity time, the remaining energy and the above mentioned playout buffer residual length

• A game modeling of this problem will be provided


• A traditional CSMA/CA model will be modified to allow support of priority differentiation in traffic management

• Network management will be performed by means of a jamming policy

• Game modeling will imply

– Use of a utility function (U)=node throughput (r )

– Use of a penalty function (P)=difference in throughput due to nodes deviation from the expected throughput

– Use of a payoff function (f) f=U-P to be maximized based on the different nodes priority (q)

– -node priority (q) depending on playout buffer residual length (Q), residual energy (E) and residual time in the proximity of the infostation(T)


Laura Galluccio


• Strategy: identified with the values assumed by the contention window size at the observation time

• Access probability

• Node throughput


Laura Galluccio

ii W

1

2

iii

iii cc

cr

32

1



Laura Galluccio

• Payoff

• Priority

• Expected throughput

iiii E

k

T

k

Q

kq 321

otherwisep

rqqifpr

rji

ij

qqdif

i

jiqq

difi

ji)(

)(

exp,

)1(

)1(

otherwise

rrif

r

rrPP

iji

i

iji

ijijN

ijjrf iiii

exp,

exp,

,,,11

1

1)1(


Definition of a distributed algorithm for reaching the Pareto Optimal Equilibria and letting the system work around it

T6.2: Opportunistic scalable video and audio coding for adaptive streaming

applications


Laura Galluccio


applications• The main target of this activity will be evaluation of

the impact of using opportunistic techniques for support of multimedia traffic in P2P wireless networks

• Multimedia Coding impacts on– File size and consequent storage requirements– The way the resource is distributed into the network (many

copies at different coding quality can be distributed)

• Resource distribution should consider inter-contact time distribution!


applications

• Two strategies will be prosecuted:• Software implementation

• Simulation

T6.4: Multimedia Peer-to-Peer (P2P) WLAN connections


Laura Galluccio


• Due to the limited bandwidth resources in wireless channels redundancy should be reduced

• To this purpose network coding can be used in P2P multihop communication environments

• Network coding allows to combine different files’ chunks so as to reduce energy consumption, increase capacity and thus throughput, and should be variable according to network conditions

• Drawbacks are longer processing and thus increase in delivery delay

• We will analitically characterize the problem• Then, we will define and test an algorithm for estimating system

performance


• Energy efficient data routing will be addressed as well

• Knowledge of position is required and can be obtained by mean of localization techniques as discussed above

• Once known the position, geographical forwarding can be exploited to forward information


• Same Payload size SD

• Header size equal to SH and SHNC~2 SH

• Node density around node A equal to A















Summarizing…

• WP5: Cross-layer Wireless Multimedia and Active QoS-Mapping– T5.2: Opportunistic access and node auto set up

• Localization (just started!)• Resource catalogue (ongoing!!!)

– T5.3: Game-Theory-based distributed radio resource management (ongoing!!)

• WP6: Applications and secure End-to-End QoS-provisioning (leadership)– T6.2: Opportunistic scalable video and audio coding for

adaptive streaming applications (just started!)– T6.4: Multimedia Peer-to-Peer (P2P) WLAN connections

(ongoing!)

Roma 17/10/08 WORLD Project KO Meeting Laura Galluccio WORLD Project – KO Meeting University of Catania.

Documents

network nodes

presence of mobile nodes

nodes localization

nodes self

mobile devices

localization infostation

mobile world rapid increase

system resources