Automating the Implementation of Software Defined Radios at Northrop Grumman

The design of a flexible or multi-mission SDR requires a great deal of development time, resources, and engineering expertise. Broadly speaking, SDR creation taps multiple domains involving signal processing, RF design, mixed-signal integration, FPGA development, and software engineering. Moreover, communications engineers with the necessary skills are in short supply. At Northrop Grumman, we set out to address these challenges by building a system called Hotrod that automates the implementation of SDR devices (Figure 1). The core of this integrated framework builds on the modeling and simulation capabilities of MATLAB, Simulink, and Xilinx® System Generator. Our main objective for this effort was to enable inexpert users without an in-depth knowledge of the nuances of communications theory, such as junior engineers and technicians, to specify and construct an operational radio design at a high level of abstraction using parameters such as data rate, modulation scheme, and available bandwidth. Beyond this, Hotrod has shown it can save considerable development time by generating an initial design architecture that experienced engineers can later modify. Utilizing MATLAB and Simulink in the Hotrod framework, we automated the complex SDR design process and demonstrate virtually unattended radio synthesis, reducing design time for our reconfigurable platforms by an order of magnitude.

By Gus DiPierro, Doug Jaeger, Patrick Ring, and Matt Vondal, Northrop Grumman
 
This article was published in The MathWorks News & Notes, 2008