ࡱ> VXW` (bjbj b>+ $PD,p$J,2( +++++++$|-h/x++,P++*|+ `Zl8d*+,0J,*\0\0$+\0+@C|)++*XJ,$   GENI Development Project Proof-of-Concept Prototyping of Methods for Wireless Virtualization and Wired-Wireless Testbed Integration 1. Project Summary: The GENI project aims to provide a flexible and programmable shared experimental infrastructure for investigation of future Internet protocols and software. GENI will consist of a global-scale wired network with programmable and virtualizable network elements (routers, switches, servers) along with several wireless access network deployments intended to support experimentation with mobile computing devices, embedded sensors, radio routers, etc. This project is aimed at finding solutions for two important technical issues related to the integration of wireless networks into GENI: Virtualization of wireless network resources to provide capabilities for simultaneous support of multiple concurrent experiments (slices) on the same set of radio devices; and Integration of control and management across wired and wireless networks, providing research users with a single programming interface and experimental methodology. Both these technical problems are on the critical path for development of wireless capabilities in GENI, and we believe that the proposed proof-of-concept prototyping effort on the ORBIT and PlanetLab testbeds will help mitigate technical risk on the GENI project as a whole. Virtualization of wireless networks is recognized to be a difficult problem due to the fact that radios interact with their neighbors at various layers of the protocol stack, making strict isolation of virtual networks (or slices) quite challenging. Several candidate techniques which exploit all three available radio resource dimensions (time, space and frequency) have been identified and will be prototyped for comparative evaluation at scale using the 400-node ORBIT radio testbed. Key software components in ORBIT (radio node drivers, NodeAgent, NodeHandler) will be modified in this project to accommodate different forms of virtualization including VMAC, SDMA, FDMA and TDMA (the last in collaboration with University of Wisconsin). Prototypes developed will be used to evaluate trade-offs between the number of slices supported and the degree of isolation given practical equipment constraints. The second project on integrating ORBIT and PlanetLab as an example of wired-wireless experimental network integration represents another important development need for GENI. In particular, PlanetLabs service model based on programmability and virtualization has been adopted as a baseline for GENI, but has so far been used mostly in homogeneous wired network scenarios. Incorporating device heterogeneity and alternative experimental usage models into the PlanetLab framework would significantly improve understanding of GENIs overall architecture and control/management protocols. In this project, ORBIT and PlanetLab will be integrated using two alternative approaches intended to provide a single experimental framework to existing users of both systems. The prototype implementation will involve enhancements to major ORBIT software modules including the NodeAgent and NodeHandler, along with modest extensions to the PlanetLab control framework where needed. Prototypes developed will be used to evaluate alternative concepts for integrating wired and wireless components in GENI, and should help drive the design of the control and management framework. Proof-of-concept demonstrations from the above prototyping projects will be prepared in support of the GENI MREFC readiness process. Design information resulting from these studies will also feed into design and system engineering work being carried out by the GENI Facility Architecture and Wireless Working Groups. Dipankar Raychaudhuri, Sanjoy Paul, Marco Gruteser, Ivan Seskar WINLAB, Rutgers University in collaboration with Larry Peterson Princeton University Suman Banerjee University of Wisconsin 2. Project Objectives & Deliverables/Demos: 2.1 Wireless Network Virtualization Project: The first project on wireless virtualization has the following objectives: Prototype feasible virtualization methods from proposed alternatives (VMAC, SDMA, FDMA, TDMA) taking into account practical radio hardware and software constraints. Design upgrades to ORBIT software components to implement selected methods on the testbed. Quantitatively evaluate performance of virtualization methods, considering factors such as number of slices supported and degree of isolation between slices. Make recommendations for specific virtualization methods to be used in GENI wireless subnets, along with guidance for GENI system engineering work. Proof-of-concept demos of wireless virtualization technology feasibility in support of the MREFC readiness process Major demos and deliverables in the virtualization project are: Development of upgraded 802.11x radio driver for VMAC, porting of VM software to ORBIT nodes and ORBIT infrastructure software upgrades (e.g. radio resource manager and other related modules for virtualization) - 12/06 Proof of concept demo of VMAC virtualization on ORBIT for star networks of APs - 1/07 [DEMO] Initial informal demo of TDMA virtualization with U Wisconsin - 2/07 [DEMO] Design of centralized resource management software for SDMA/FDMA and client driver upgrade; integration of RF signal monitoring - 4/07 Proof of concept demonstration for SDMA/FDMA/TDMA on ORBIT grid with arbitrary wireless network topology - 5/07 [DEMO] Wireless virtualization design document as input to GENI system engineering - 1/07, 7/07 2.2 Wired-Wireless Testbed (PlanetLab-ORBIT) Integration: The second project on wired-wireless testbed integration has the following objectives: Prototype example end-to-end wired + wireless network testbed by integrating PlanetLab and ORBIT. Investigate alternative design choices for integration such as ORBIT slice vs. PlanetLab-ORBIT proxy. Identify extensions to PlanetLab and ORBIT experimental control software, and extract recommendations for GMC (GENI control & management) specs needed to support heterogeneous wireless devices. Conduct several proof-of-concept wired + wireless experiment demos in support of GENI MREFC readiness process. Major demos and deliverables in the wired-wireless testbed integration project are: Software extensions to ORBIT NodeHandler and NodeAgent to implement ORBIT slice integration model - 12/06 Proof-of-concept demonstration of ORBIT slice approach using location-aware mobile networking application as an example - 1/07 [DEMO] Software extensions to ORBIT NodeHandler and NodeAgent to implement PlanetLab-ORBIT proxy integration model - 2/07 Proof-of-concept demonstration of PlanetLab-ORBIT proxy approach using selected PlanetLab applications such as CDN with mobile users, or grid/sensor experiment - 3/07 [DEMO] Proof-of-concept demos with multiple research experiments, and possible service release to ORBIT and PlanetLab users. Integrate with virtualization capability on ORBIT grid. - 6/07 [DEMO] Reports with recommendations for experimental wired + wireless service models and control and management protocols in GENI - 1/07, 7/07 3. GENI Risk Reduction Impact: 3.1 Wireless Virtualization Project: This project is expected to lead to validation of wireless virtualization approaches which are considered to be a critical path technology in the development of GENI. The effort should also result in prototype Linux code for virtualization that can later be ported to various GENI wireless implementations. Quantitative studies will also help guide the selection of virtualization methods for different types of experimental scenarios. 3.2 PlanetLab-ORBIT Integration Project: This project is aimed at reducing design risk associated with extending GENIs slice virtualization model to incorporate wireless subnetworks in a seamless way. In particular, the results of this project are expected to guide the design of an experimental service model/API for heterogeneous wired + wireless network experiments in GENI. The effort should also provide design guidance for wireless-specific aspects of the GENI management and control (GMC) protocol.     GENI Development Project Summary Page:  PAGE 1 WINLAB, Rutgers University  DATE \@ "M/d/yyyy" 10/25/2006 p | u~FPT_%;Of}~ui`ih5CJaJh7:&hd35CJaJh,Th,T5CJaJh,T5CJaJhGh,T5CJaJh-Ah,TCJaJhGh,T6CJaJh>wuh,T6CJaJh,TCJaJh7:&CJaJh7:&h7:&5CJaJ h,T5h]`h,T5CJaJh]`h,T5CJaJh]`h,T56CJaJ  > T t%Q~# $xxa$gdgdd3gd,T$0^`0a$gd,T$a$gd,T $a$gd,T $ & Fa$gd,T$a$gd,Tgd,T$a$gd,T+((!"#33TUwd59;AaCV`fo$Ю裗裗ź裗hhd35CJaJhh$5CJaJh7:&h$CJaJh7:&h$5CJaJh7:&hpm{CJaJh7:&h,bCJaJh7:&hd35CJaJh7:&h7:&5CJaJh7:&hd3CJaJh7:&h,T5CJaJ6##UB${H y $ & Fa$gdd3 $ & Fa$gdpm{$a$gdd3gdd3$ & F 8x^a$gdd3 $ & Fa$gdd3$ & Fxxa$gd(./8?EFGi    * 7 L M x  !!!!3!8![!g!p!!!!!!!!!""""0"1"4"W"a"ӿӴӴhh$5CJaJh7:&hpm{6CJaJh7:&hCJaJhCJaJh7:&h$6CJaJh7:&hpm{CJaJh7:&hd36CJaJh7:&h$CJaJh7:&hd3CJaJ9y 9!!5""#+$,$-$L$M$(&)&)(*(+(-(.(0(1(3(4(6(gdd3gd7:&$a$gdd3$ & F x^a$gdd3 $xa$gdd3a"m"v""""""""""""#Q#W########### $$$ $*$,$-$.$K$L$M$N$p$O%b%%%)&*&Q&c&t&&&һݳҳǫһݠh7:&hd35CJaJh7:&h7:&5CJaJhH5CJaJh7:&h7:&CJaJhCJaJhHCJaJhh$5CJaJh7:&hCJaJh7:&h$CJaJh7:&hd3CJaJh7:&hpm{CJaJh7:&h6CJaJ2&&&c'l'z'''''((*(+(,(.(/(1(2(4(5(7(_(`(f(g(h(i(j(k(((((((((((òÒÃÃsÒh$hCJaJmHnHujh$hCJUaJhh0JCJaJmHnHuh$h0JCJaJ!jh$h0JCJUaJh$hCJaJhW jhW Uh7:&hd3CJaJhCJaJhHCJaJh7:&h7:&CJaJ'6(7(j(k(((((21h:p,T/ =!"#$% ,1h/ =!"#$% @@@ ,TNormalCJ_HaJmH sH tH DA@D Default Paragraph FontRi@R  Table Normal4 l4a (k@(No List:O: ,T Body CharxCJaJJoJ d3 Heading first & Fx<54@4 $Header  !4 @"4 $Footer  !.)@1. $ Page Number~ > 4> >Tt %Q~##UB${H y95+,-LM()) * + - . 0 1 3 4 6 7 j k 000 0 0000000000000 0 0 0 0 00 0 0 0 0 0 0000 0 0 0 00 0 0 0 0 0 0000000000@0y00@0y00@0y00@0y00@0@0@0@0y00 >Tt %Q~##U{H y95+, @0@0@0@ 0@ 0@0@0@0@0@0@0@0@0@0@ 0@ 0@ 0@ 0@ 0@0@ 0@ 0@ 0@ 0@ 0@0@ 0@ 0@ 0@ 0 @0A 0oA 0oA 0oA 0oA 0oA 0o@ 0 0 @@a"&(#y 6(((4;=\q|!8@0(  B S  ? OLE_LINK3  kk\"k_) kIJk<k*k,k<]klkԝkTkJk kZ) k,K##1;;Effts s {      -::DOOp}}z    = *urn:schemas-microsoft-com:office:smarttags PlaceName=*urn:schemas-microsoft-com:office:smarttags PlaceType9*urn:schemas-microsoft-com:office:smarttagsplace 0     ry5<u~  K T F O T _ QVW_!IP/8_hW`HQ-6+ + - - . . 0 1 3 4 6 7 9E+ + - - . . 0 1 3 4 6 7 3~.HZ+ + - - . . 0 1 3 4 6 7 _ k + + - - . . 0 1 3 4 6 7 , OX9PdvF^X{H&Pkv[2tS88^8`o(. ^`hH.  L ^ `LhH.   ^ `hH. xx^x`hH. HLH^H`LhH. ^`hH. ^`hH. L^`LhH.^`o() ^`hH. pLp^p`LhH. @ @ ^@ `hH. ^`hH. L^`LhH. ^`hH. ^`hH. PLP^P`LhH.h^`OJPJQJ^J.h^`OJQJ^Jo(hHohpp^p`OJQJo(hHh@ @ ^@ `OJQJo(hHh^`OJQJ^Jo(hHoh^`OJQJo(hHh^`OJQJo(hHh^`OJQJ^Jo(hHohPP^P`OJQJo(hHh 88^8`o(hH.h ^`o(hH.h  ^ `OJQJo(hHh  ^ `OJQJo(hHhxx^x`OJQJ^Jo(hHohHH^H`OJQJo(hHh^`OJQJo(hHh^`OJQJ^Jo(hHoh^`OJQJo(hH^`o(() ^`hH. pLp^p`LhH. @ @ ^@ `hH. ^`hH. L^`LhH. ^`hH. ^`hH. PLP^P`LhH.hh^h`o(. 88^8`hH. L^`LhH.   ^ `hH.   ^ `hH. xLx^x`LhH. HH^H`hH. ^`hH. L^`LhH.PkOX9vF[2t, {H         d                 Bx        0          W 7:&d3,T,bpm{$H@++ Ҡ++ p@UnknownGz Times New Roman5Symbol3& z Arial?5 z Courier New;Wingdings"1h9ͪfBͪf{^:^:!4  2qHX ?,T2Project SummaryrayDipankar Raychaudhuri$      Oh+'0 $ D P \ ht|Project SummaryrayNormalDipankar Raychaudhuri3Microsoft Office Word@ԭ@^F@V}@MW^՜.+,0 hp  WINLAB, Rutgers University:  Project Summary Title !"#$%&')*+,-./0123456789:;<=>?@BCDEFGHJKLMNOPSRoot Entry F@klUData  1Table(0WordDocumentb>SummaryInformation(ADocumentSummaryInformation8ICompObjq  FMicrosoft Office Word Document MSWordDocWord.Document.89qRoot Entry Feb=[@Data  1Table(0WordDocumentb> !"#$%&')*+,-./0123456789:;<=>?@BCDEFGHZY SummaryInformation(ADocumentSummaryInformation8CompObjq  FMicrosoft Office Word Document MSWordDocWord.Document.89q՜.+,D՜.+,T hp  WINLAB, Rutgers University:  Project Summary Title4 $,