SHF:Small:Software-Defined Radio: From High-level Language to Hardware

Drexel University, Philadelphia PA

Investigators

Abstract

Software-defined radio (SDR) promises to bring the flexibility of software development to radio protocol design. Unfortunately, SDR platforms that support high-data rate protocols are difficult to program and require low-level, expert knowledge that has little to do with radio protocol design. This project makes high-data rate radio protocol design accessible to a wider community of developers by developing a new high-level language for expressing SDR applications, freeing protocol designers from worrying about the low-level details of a particular SDR platform. The intellectual merits are a new understanding of the language abstractions and runtime facilities needed to translate high-level code to efficient low-level implementations. The project's broader significance and importance are to drastically lower the barrier to entry for high-speed software-defined radio platforms and to produce new tools for the SDR community that facilitate both research and education. The proposed research builds on Ziria, a language for writing radio physical layer (PHY) implementations. Ziria will be extended in to support new radio protocol applications and to support new hardware platforms, including FPGAs. A key project goal is portability: protocols written in Ziria should be portable across SDR platforms without significant loss of performance. Furthermore, users must be able to make use of all parts of heterogeneous SDR platforms---from FPGAs to CPUs to DSPs---seamlessly and without rewriting code. Finally, this project applies knowledge gained in the SDR domain to new domains, such a video encoding and decoding, where low-level hardware is also commonly targeted.