Research on Wireless Networks at NKUA Prof. Ioannis Stavrakakis Communication Networks Laboratory Dep. of Informatics and Telecommunications National & Kapodistrian University of Athens (NKUA) Leonidas Tzevelekas Dagstuhl Seminar, March 2004
Research on Wireless Networks at NKUA
Prof. Ioannis Stavrakakis Communication Networks Laboratory
Dep. of Informatics and TelecommunicationsNational & Kapodistrian University of Athens (NKUA)
Leonidas TzevelekasDagstuhl Seminar, March 2004
Leonidas TzevelekasDagstuhl seminar, March 2004
2
Outline
Scalability of Ad hoc Routing Protocols, C. Santivanez, I. Stavrakakis
On multi-hop extensions of WLANs,A. Panagakis, E. Balafoutis, I. Stavrakakis
IST Broadway Project,A. Vaios, I. Stavrakakis
Wireless Sensor Networks, L. Tzevelekas, I. Stavrakakis
Leonidas TzevelekasDagstuhl seminar, March 2004
3
Scalability of ad hoc routing protocols
Work’s objectives:1. To define and measure scalability in ad hoc
networks
2. To examine existing routing protocols for ad hoc networks with respect to scalability
3. To propose new, ‘better’ (more scalable) routing protocols for ad hoc networks
Leonidas TzevelekasDagstuhl seminar, March 2004
4
Definitions of Scalability Network’s scalability (quantitatively, qualitatively)
Routing protocol’s scalability (quantitatively, qualitatively)
Examples Network Scalability Factor:
Routing protocol Scalability Factor:
Scalability of ad hoc routing protocols
def1 2log ( )
limlogi
i i
Tr …
def1 2log ( )
limlogi
i
X ov
i
X …
Leonidas TzevelekasDagstuhl seminar, March 2004
5
Scalability of ad hoc routing protocols Family of Fuzzy Sighted Link State algorithms
(FSLS Algorithms)Protocol Characteristics the frequency of Link State Updates (LSUs) propagated to
distant nodes is reduced a node transmits a Link State Update (LSU)
only at particular time instants the Time To Live (TTL) field of the LSU is set to a value
that is a function of the current time Hazy Sighted Link State (HSLS)
latency versus distance curve: optimal performance when linear
optimal balance between proactive and sub optimal routing overhead
Leonidas TzevelekasDagstuhl seminar, March 2004
6
Scalability of ad hoc routing protocolsRelated Publications
C. Santivanez, B McDonald, I. Stavrakakis, Ram Ramanathan, “On the Scalability of Ad Hoc Routing Protocols. Part II: Asymptotic Analysis with Respect to Traffic”, IEEE/ACM Transactions on Networking [About to be submitted or submitted]
C. Santivanez, Ram Ramanathan, I. Stavrakakis, B McDonald, “On the Scalability of Ad Hoc Routing Protocols. Part I: Making Link State Routing Scale”, IEEE/ACM Transactions on Networking [ submitted]
C. Santivanez, B McDonald, I. Stavrakakis, Ram Ramanathan, “On the Scalability of Ad Hoc Routing Protocols”, IEEE INFOCOM 02, N. York, USA
C. Santivanez, Ram Ramanathan, I. Stavrakakis, “Making Link-State Routing Scale for Ad Hoc Networks”, MobiHoc 2001, Oct. 4-5, 2001, Long Beach, CA, USA
Leonidas TzevelekasDagstuhl seminar, March 2004
7
The main problem: Need for higher capacity/user
Motivation : to support high-rate apps to allow for direct communication between MTs to support Hotspots requiring higher capacity
“In-fashion” solution: Ad hoc in WLANs? Via a Multi-hop Cellular Network (MCN)
On multi-hop extensions of WLANs
Leonidas TzevelekasDagstuhl seminar, March 2004
8
On multi-hop extensions of WLANsEvaluation of MCN (simulation environment) ns simulator (version 2.1b7) MAC : IEEE 802.11 DCF mode Propagation model: two ray path loss model Routing: DSR Transport protocol: UDP packet size: 512 bytes default parameters of the simulator (WaveLan Card)
(except transmission range, range of interference)
Leonidas TzevelekasDagstuhl seminar, March 2004
9
AP
On multi-hop extensions of WLANs The performance of an MCN is rather poor
(in terms of throughput, for inter-cell traffic)
Doughnuts “taste” better than cookies (as expected: spatial reuse up to the edge, lower max hop count)
“Taste” of Doughnut depends on: Interference Dimensions of Doughnut
Additional “taste” comes at the cost of increased power consumption (in case where last hop longer)
Leonidas TzevelekasDagstuhl seminar, March 2004
10
Related Publications Antonis Panagakis, Elias Balafoutis and Ioannis Stavrakakis,
“Study of the capacity of multi-hop cellular networks”, QofIS 2003, 1-3 October 2003, Stockholm, Sweden
Athanasios Vaios, Konstantinos Oikonomou and Ioannis Stavrakakis, "A Centralized Routing Scheme Supporting Ad Hoc Networking in Dual Mode HIPERLAN/2 System", IST Mobile Summit 2003, 15-18 June, Aveiro, Portugal
On multi-hop extensions of WLANs
Leonidas TzevelekasDagstuhl seminar, March 2004
11
IST Broadway Project
IST-Broadway:IST-Broadway: “the way to broadband access at 60GHz”
Duration: 01/01/2002 – 31/12/2004
PARTNERS:1. CRM, Motorola Labs Paris2. Intracom3. National and Kapodistrian University of Athens4. TNO Physics and Electronics Laboratory5. Farran6. Technische Universität Dresden7. Institut für Mobil und Satellitenfunktechnik - IMST
Leonidas TzevelekasDagstuhl seminar, March 2004
12
IST Broadway Project
Motivation Standard ETSI BRAN HiperLAN/2 features 5.2 GHz AP is the bottleneck when the number of users demanding very
high bit-rates is increased All the connections have to pass through the AP even when
MTs are very close peer-to-peer communications are penalized
Objective: Offload the 5 GHz mode of operation Based on HiperLAN/2: Enhanced HiperLAN/2 System with two
modes of operation Mode 1: Standard HiperLAN/2 Mode 2: HiperLAN/2-like operation at 60 GHz
Leonidas TzevelekasDagstuhl seminar, March 2004
13
IST Broadway Project Basic Principle: the 5 GHz mode is considered as
the default and signalling mode
When connectivity exists the AP manages the switching between different modes for all MTs
Two types of equipment For the AP: operates at both frequency-bands
simultaneously For the MTs: operate either at 5 GHz or 60 GHz (requiring a
switching between the two bands)
Leonidas TzevelekasDagstuhl seminar, March 2004
14
IST Broadway Project
Centralized Ad hoc Network Architecture Centralized Ad hoc Network Architecture (CANA)(CANA)
AP Access Point
CH Cluster Head
FN Forwarding Node
Mobile Terminal
5 GHz Range
60 GHz Range
5 GHz Data Flow
60 GHz Data Flow
AP
CHCH
FNCH
CHCH
CH
FNFN
CH
Leonidas TzevelekasDagstuhl seminar, March 2004
15
IST Broadway
Related Publications K. Oikonomou and I. Stavrakakis, “Analysis of a Probabilistic
Topology-Unaware TDMA MAC Policy for Ad Hoc Networks”, IEEE JSAC Special Issue on Quality of Service Delivery in Variable Topology Networks
K. Oikonomou, A. Vaios, S. Simoens, P. Pellati and I. Stavrakakis, “Centralized Ad-hoc Network Architecture (CANA) Based on an Enhanced HiperLAN/2 System”, PIMRC 2003, 7-10 September, Beijing, China
K Oikonomou and I. Stavrakakis, “Throughput Analysis of a Probabilistic Topology-Unaware TDMA MAC Policy for Ad Hoc Networks”, QofIS, 1-3 October 2003, Stockholm, Sweden
Leonidas TzevelekasDagstuhl seminar, March 2004
16
Wireless Sensor Networks
Related Project: E-NEXTE-NEXT Network of Excellence in emerging network
technologies IST FP6 programme Duration: 2004 – 2005 Task Forces:
1. TF1 – scalable routing 2. TF2 – distributed router architecture to gain
scalability3. TF3 – traffic driven scalabilit4. TF4 – new task force on sensor networks
proposed by us
Leonidas TzevelekasDagstuhl seminar, March 2004
17
Wireless Sensor Networks
My personal interests
1. Power aware / energy efficient routing for WSNs
2. Design of cross-layer algorithms for improved power efficiency
3. Real-life deployment of test networks
Questions
1. Direction of WSN related research
2. Impact of this technology on future generations
3. Trends & Hype: Where is the border?
Leonidas TzevelekasDagstuhl seminar, March 2004
18
More Information
Our web addresses
http://cnl.di.uoa.gr – Communication Networks Lab
http://cgi.di.uoa.gr/~istavrak – Prof. I. Stavrakakis web page
http://users.cnl.di.uoa.gr/~ltzev – Personal Web Page