A Scalable Baseband Platform for Energy-Efficient Reactive Software-Defined-Radio

Abstract

The spectrum data-mining and agile air interface requirements of cognitive radios claim for software-defined-radio (SDR) implementations. The need to detect and/or generate virtually any kind of waveform in any band pushes the specification of such SDR to the limit. This makes cost-effective and energy-efficient implementation very challenging. As a first step toward effective embedded cognitive radio, we present the design, based on a top-down methodology, of a scalable SDR platform that enables spectrum environment awareness and reactive transmission. Hardware cost and energy-efficiency are contended by opportunistic partitioning and aggressive power management leading to 10 mW-range standby power and 100 mW-range active power