Over the last few months I've noticed a significant uptick in stories related to software-defined radio (SDR). CEVA has a new SDR platform for handsets. Octasic has one for base stations. Imagination is bringing SDR to TV and radio receivers. And the list goes on.
Each of these announcements is quite impressive, but put them all together and I conclude that SDR has become a mainstream technology. I ran this idea past industry experts like Will Strauss of Forward Concepts and Jeff Bier of BDTI, and they disagreed with me. To them, SDR still has a way to go before it can be considered mainstream.
So what gives? Am I drawing the wrong conclusions, or has the age of SDR has arrived without anybody noticing?
To answer that question, we first need to define SDR. The SDR Forum defines it as "Radio in which some or all of the physical layer functions are software defined." That's a very broad definition. Under this definition, you could argue that every handset modem on the market is a software-defined radio—and in fact the SDR Forum sees the world this way. This seems like a silly argument to me. If every radio is a software-defined radio, then what is so special about SDR?
On the other extreme, some people think of SDR as a system where the entire signal chain is done in software. (You still need an analog front-end, of course). This sort of system is not practical for commercial applications, and is unlikely to become practical any time soon. That's one reason Jeff Bier and Will Strauss disagreed with me—if this is what you mean by SDR, then SDR is obviously not mainstream.
My own idea of SDR lies somewhere between these extremes. I like the Wikipedia definition: "…a radio communication system where components that have typically been implemented in hardware (e.g. mixers, filters, amplifiers, modulators/demodulators, detectors. etc.) are instead implemented using software." By this definition, the CEVA, Octasic and Imagination platforms all qualify as SDR. All three of these platforms do things in software that were previously done in hardware.
If you can accept this middle-of-the-road, somewhat fuzzy definition of SDR, you can make the case that SDR has become a mainstream technology. So, I'll stick to this view, even after talking to Jeff Bier and Will Strauss. What do you think?
Next time: I explain how changes in technology and market forces have aligned to bring SDR to a tipping point of mass adoption.
PS—I am excited to announce a HUGE discount on the Embedded Systems Conference. We've dropped the prices 25% for DSP DesignLine readers. That's a savings of up to $400! Use code CTMSP15 when you