UCLA Extension

Software-Defined Radio

A software-defined radio (SDR) is a communication system that performs many of its required signal processing tasks in a programmable digital signal processing (DSP) engine. The engine is coupled to the air interface consisting of analog circuits and antennae by analog-to-digital and digital-to-analog converters. Software changes reprogram the DSP segment of the radio’s physical layer to reconfiguring the radio and can thus synthesize multiple radios. Software control also is employed to select and alter the air interface segment as well as the higher-level data processing layers of the radio system. Why field a reconfigurable radio? The answer, of course, is to support communications between a wide range of communication systems. Further, as new waveforms are developed and deployed, or as standards incorporate new features rather than field another waveform specific radio, an SDR can be programmed to be the new radio. By permitting existing hardware and software to incorporate new wireless features and capabilities through software upgrades, the SDR offers the promise of reduced cost, wider utilization, and delayed obsolescence.

Complete Details

The timely confluence of multiple major components and technology enables software radio! One major enabling component is multirate signal processing: the flexibility afforded by multirate systems with varying filter capabilities provides SDR much of its power and flexibility. A second major component is a large degree of interdisciplinary knowledge. The general-purpose computer (GPP) is the core of many SDR systems, such as GNU Radio. To make the most out of the computational capabilities of a GPP, good programming skills are needed as well as knowledge of its computational architecture, current libraries, and techniques developed to incorporate them into SDR. The programmer’s familiarity with these libraries and their efficient usage is hugely important in making good-quality, robust, and efficient software.

A third component is the need to be knowledgeable about communication, information, and coding theory that go into building good-quality SDR systems. A fourth major—and often neglected—component is a working knowledge of RF and an appreciation of RF capabilities, limitations, and nonlinear effects associated with analog hardware. The designer must be aware that DSP and clever programming cannot help the radio violate the laws of physics. An analog radio loses the signal when the link experiences a deep fade. An SDR does the same. What the SDR and its embedded software-controlled DSP engines can do is expand utility, extend operating regime, reduce cost, and impress its users by enhancing capabilities as well as reward its proponents and practitioners.

This course presents an overview of the software radio systems and the new DSP-based architectures leading radio innovation. The instructors develop and illustrate the flexibility afforded the SDR by multirate signal processing techniques. Real-time MATLAB simulations illustrate essential concepts for a number of SDR systems. The GNU radio, a low-cost but powerful way to get involved in SDR, is used to demonstrate many of the important SDR concepts. The Ettus Research Company’s Universal Software Radio Peripheral (USRP) product family serves as the platform to illustrate the flexibility of the software-hardware interface.

Coordinator and Lecturer

fredric j. harris, PhD, Cubic Signal Processing Chair Professor of Electrical and Computer Engineering, San Diego State University, California. Dr. harris is a recognized expert in the field of Digital Signal Processing (DSP), especially as applied to underwater acoustics, radio surveillance, satellite communications, radar, real-time acoustics, vibration monitoring equipment, and laboratory instrumentation. He is the author of the text, Multirate Signal Processing for Communication Systems. Since 1970, he has been a consultant to such organizations as the U.S. Navy Ocean Systems Center, Lockheed, ESL, Cubic, Hughes, BAE, Scientific Atlanta, Rockwell, Northrop Grumman, Boeing, and Inritsu. He also has presented courses on fast algorithms, adaptive algorithms, error-correcting codes, and control theory. Dr. harris has conducted seminars in DSP for Motorola, Northrop Grumman, BAE, Lockheed, Hewlett Packard, General Electric, Rockwell, Spectral Dynamics, and the U.S. Navy Research Laboratory.



Daily Schedule

  • Radio System Architectures
  • Software-Defined Radio Architectures
  • Physical Layer Tasks and Processing
  • Higher-Layer Tasks and Processing
  • Digital Filter Design for Interpolators
  • Signal Conditioning for SDR Radios
  • Multirate Filters for Arbitrary Interpolators
    —Polyphase filter design and implementation
    —Farrow filter design and implementation
  • Signal Enhancement for SDR
  • GPP-Based Software Radio
  • Hardware vs. Software in SDR and Where They Meet
  • The GNU Radio Approach
    —Software suite of signal processing tools
    —Flexible, inexpensive hardware for the air interface (the USRP)
  • Limitation of this Approach
  • Benefits of the Approach
  • Overview of the Data Stream Approach: Using a Flowgraph
  • The Multidisciplinary Requirements of SDR
  • Issues in Computational Complexity and Programming Issues
    —Using available libraries
    —Using advanced processor concepts, such as SIMD
  • Examples of GNU Radio
    —Focus on filtering: how this fundamental concept shows off the levels of complexity and programming issues
  • Advanced Programming Concepts
    —Message-passing architecture: how this enhances the SDR
    —Vectorized routines: the new libvector tools

For more information contact the Short Course Program Office:
shortcourses@uclaextension.edu | 310-825-3858