Introduction

Coined by Joe Mitola in 1991 to refer to the class of reprogrammable or reconfigurable radios [1], the term “software radio” has evolved into the following encompassing definition given in [2]:

“a radio that is substantially defined in software and whose physical layer behavior can be significantly altered through changes to its software.”

The evolution of software radio has been motivated by the following goals: supporting multiband multimode radios (MBMMR), facilitating global roaming, runtime reconfigurability, over-the-air-programming, providing a common radio architecture, and improving spectrum utilization.

MBMMR started to become increasingly important to commercial interests in the late 1980’s with the emergence of digital cellular standards. To leverage the better coverage provided by the older analog systems, service providers needed a single phone that would operate at both the analog and digital frequencies (hence multiband) and with both waveforms (hence multimode). The need for commercial MBMMR has continued with the creation of tri-mode and quad-mode phones and has been expanded further to support global roaming phones.

Frequency agility in terms of the center frequencies and bandwidths supported by the radio is critical to the MBMMR and global roaming concepts. Since traditional RF components are not frequency agile, software radio has had to introduce new approaches to RF design, specifically MEMS (micro electro-mechanical switches) based RF and digital based RF. In MEMS based RF, hundreds of miniature switches are employed to vary the capacitance, inductance, and resistance of RF components to achieve the desired frequency response. In digital based RF, digital processing is used to compensate for any deficiencies in the RF.

Of course digital based RF places enormous constraints on the processing elements. Thus while the term “software radio” implies significant microprocessor content, significant operations may still be performed in ASICs or in FPGAs when particularly high performance is required. However, ASICs and FPGAs are not particularly well-suited for either run-time reconfiguration or over-the-air-programming. To address the relative drop in processor flexibility, software radio has developed alternative processing solutions, such as custom computing machines (CCMs) that provide the massive parallelization of FPGAs but with a coarser granularity that permits rapid reconfiguration – on the order of microseconds [3].

The proliferation of different technologies has heightened the need for a common radio architecture.  Championed by the US military through the JTRS program, the Software Communications Architecture (SCA) [4] has been adopted as a standard software radio architecture and now forms the basis of the SDR Forum’s Software Radio Architecture (SRA) [5]. Recognizing that different radios have different capabilities, the SCA and SRA define a number of common functionalities while providing for differentiation across categories (or clusters) of software radios. Significant development efforts continue on the SCA and SRA with the handheld cluster presenting the greatest challenges.

By adapting their operation to their environment, software radios also have the promise of improving spectrum utilization – a fact that has not gone unnoticed by the FCC. The FCC is in the process of liberalizing their regulations for radio operation, and has issued initial guidelines for facilitating the reprogramming of software radios [6]. Cognitive radios – software radios that are aware of their environment and abilities and are capable of independently changing their operation – are also receiving attention from the FCC [7], though adoption of this technology is understandably proceeding at a slower pace.

While this discussion appears speculative, significant numbers of software radios are now being built daily. Having successfully demonstrated prior prototypes, Boeing is currently constructing cluster 1 (large form factor) software radios [8]. Cluster 2 (smaller form factor) software radios are being constructed by Thales [9]. On the commercial side, Vanu Inc., has fielded a PDA based software radio [10], and Spectrum Signal Processing sells a testbed for testing SCA/SRA compliant waveform [11]. With all these software radio successes, your next cell phone may just be a software radio too!

Tutorial Slides on Software Radios

The links below are power-point presentations on different aspects of software radios from Dr. Reed's lectures.

• Overview and challenges

• Introduction

• RF issues

• Data Conversion

• Processing hardware

• Software issues

• Network integration issues

• Examples of software radio implementation and future trends

• Smart antennas

• Multirate digital signal processing

• Standardization of software radios

• Software communications architecture

• Network enhancement with SDR

References

[1] Joseph Mitola, III, Software Radio Architecture: Object Oriented Approaches to Wireless Systems Engineering, John Wiley and Sons, 2000.

[2] Reed, Jeffrey H.  Software Radio: A Modern Approach to Radio Engineering, Prentice Hall, 2002, p 2.

[3] Neel, James O., Simulation of an Implementation and Evaluation of the Layered Radio Architecture,” Master’s Thesis Virginia Tech, December 2002.

[4] JTRS Program

[5] SDR Forum

[6] FCC Docket No. 00-47, “Authorization and Use of Software Defined Radios,” Adopted Dec. 7., 2000, Released Dec. 8, 2000.

[7] Cognitive Radio Technologies Proceedings, May 19, 2003.

[8] Boeing JTRS Backgrounder Press Release

[9] JTRS Cluster 2 Overview Available Online

[10] Chiu, Andrew, and Jessica Forbess, “A Handheld Software Radio Based on the IPAQ PDA: Software,” SDR Forum 03, November 2003.

[11] Spectrum Signal Processing Press Release, “Spectrum Signal Processing Selected to Provide JTRS Representative Hardware for U.S. Department of Defense JTRS Test and Certification Activities,” Available Online