No abstract

No abstract

This paper presents PUMA, a novel declarative constraintsolving platform that achieves efficient policy-based channel selection and routing for multi-radio wireless mesh networks. PUMA is based on declarative networking, a databaseinspired extensible infrastructure using query languages to specify behavior. In PUMA, users specify high-level declarative policies that dictate their channel selection constraints and routing protocol behavior. We demonstrate that channel selection can be expressed in a compact fashion and implemented efficiently. We have developed a PUMA prototype based on the RapidNet declarative networking engine with enhancements to handle multi-channel communication and integration with an open-source constraint solver. We perform preliminary evaluation of PUMA using the emerging ns-3 network simulator, and describe our ongoing research in ORBIT testbed deployment, distributed channel selection protocols, and distributed optimizations that combine routing and channel selection.

No abstract

No abstract