A Policy-based Unified Multi-radio Architecture for Agile Mesh Networking
PUMA is a declarative constraint solving platform for policy-based routing and channel selection in multi-radio wireless mesh networks. We have developed a prototype of the PUMA system using the RapidNet declarative networking system deployed on the ORBIT testbed. RapidNet can be downloaded here. PUMA consists of the following two building blocks:
Policy-based adaptive routing. Using the declarative networking framework, network protocols and adaptation policies are concisely specified in the declarative language to build hybrid protocols, where the policies dictate the dynamic selection of different protocol components based on network conditions.
Extensible dynamic channel selection. Users formulate channel selection policies as optimization goals and constraints that are concisely declared using the PawLog declarative language. To efficiently execute PawLog programs in a distributed setting, PUMA integrates the high performance Gecode constraint solver with the RapidNet declarative networking engine. PUMA further enables distributed protocols that cross optimize across channel selection and routing.
The above figure shows an overview of PUMA from the perspective of a single PUMA node. The role of the channel manager is to assign available channels to wireless links to optimize a performance goal (e.g. minimize network interference, minimize the number of unique channels) while subjected to constraints (e.g. regional policies on spectrum usage, yield to primary users who own exclusive rights to certain spectrums in white space networks). In PUMA, we use the PawLog language for declaratively expressing goals and constraints as a constraint optimization problem (COP).
At the network layer, the RapidNet declarative networking engine is deployed within the control plane to implement a variety of neighbor discovery and routing protocols also expressed in PawLog. The channel abstraction layer interacts with multiple radios and presents upper layers with a multi-channel communication interface. In order for packets to be routed to neighbors using appropriate interface/channel. The output of the channel manager is then used to initialize the channel assignment table at the channel abstraction layer. Finally, the output of declarative routing is a forwarding table (next-hop for each destination) used by the forwarding agent.
Students
Changbin Liu
Lu Ren
Faculty
External collaborators:
Prithwish Basu, BBN Technologies
Yun Mao, AT&T Labs Research
Alumni
Tanveer Gill (MSc. 2013, Netsil)
Xiaozhou Li (MSc. 2010, pursuing Ph.D. in Computer Science at Princeton University)
Richardo Correa (MSc. 2008, Lockheed Martin)
Mihai Oprea (MSc. 2009)
Shiv Muthukumar (MSc. 2010, Amazon)
Taher Saeed (MSc. 2011, DramaFever)
Yash Saini (BSE 2011, Amazon)
Aditi Jain (BSE 2011)
Publications
Cologne: A Declarative Distributed Constraint Optimization Platform. (PDF)
Changbin Liu, Lu Ren, Boon Thau Loo, Yun Mao, and Prithwish Basu.
38th International Conference on Very Large Databases (VLDB’12).
PUMA: Policy-based Unified Multi-radio Architecture for Agile Mesh Networking. (PDF)
Tanveer Gill, Changbin Liu, Lu Ren, Harjot Gill, Boon Thau Loo, Prithwish Basu
4th International Conference on Communication Systems and Networks (COMSNETS) (demo), Jan, 2012.
PUMA: Policy-based Unified Multi-radio Architecture for Agile Mesh Networking. (PDF)
Changbin Liu, Ricardo Correa, Harjot Gill, Tanveer Gill, Xiaozhou Li, Shivkumar Muthukumar, Taher Saeed, Boon Thau Loo, Prithwish Basu
4th International Conference on Communication Systems and Networks (COMSNETS), Jan, 2012.
Declarative Policy-based Adaptive Mobile Ad Hoc Networking. (PDF)
Changbin Liu, Richardo Correa, Xiaozhou Li, Prithwish Basu, Boon Thau Loo, and Yun Mao.
ACM/IEEE Transactions on Networking (TON), 2011.
A Policy-based Constraint-solving Platform Towards Extensible Wireless Channel Selection and Routing.
(PDF)
(Talk)
Changbin Liu, Xiaozhou Li, Shivkumar C. Muthukumar, Harjot Gill, Taher Saeed, Boon Thau Loo, and Prithwish Basu.
ACM Workshop on Programmable Routers for Extensible Services of TOmorrow (PRESTO), Philadelphia, PA, 2010.
Declarative Policy-based Adaptive MANET Routing.
(PDF)
(Talk)
Changbin Liu, Rick Correa, Xiaozhou Li, Prithwish Basu, Boon Thau Loo, Yun Mao.
The 17th IEEE International Conference on Network Protocols (ICNP), Princeton, New Jersey, Oct, 2009.
A Declarative Perspective on Adaptive MANET Routing. (PDF) (Talk)
Changbin Liu, Yun Mao, Mihai Oprea, Prithwish Basu, and Boon Thau Loo.
ACM SIGCOMM Workshop on Programmable Routers for Extensible Services of TOmorrow (PRESTO), Seattle, WA, Aug 2008.
Funding
This work is sponsored by:
NSF CNS-0721845
NSF CCF-0820208
NSF CAREER CNS-0845552
FIA grant NSF CNS-1040672