Three large organizations in the financial community have recently asked Semiconductor Insights about the status and the implications of Software Defined Radio (SDR) on the wireless industry. The reason for the request is that the ability to define radios by software lets wireless providers reduce operating costs, efficiently support additional services such as video and camera phones, and boost revenues by carrying traffic previously supported on multiple separate technologies.
Well-designed SDRs will potentially reduce time-to-market for deploying new services or features. The potential impact of SDR on the wireless industry is significant, and while SDR was just a buzz word five years ago amongst a small group of RF designers, it's now become a key technology.
SDR impacts not only consumer wireless devices and their support networks, but also governmental, military, public safety, and transportation wireless operators. The SDR forum (www.sdrforum.org) now has more than 100 members. The business case for SDR has already been accepted and technology gurus are working to deploy SDR across wireless platforms, probably starting with cellular handsets.
According to the FCC, SDR is alarmingly simple. It states, "In an SDR, functions that were formerly carried out solely in hardware, such as the generation of the transmitted signal and tuning and detection of received radio signals, are performed by software that controls high-speed signal processors."
Similarly, the SDR Forum defines an SDR device as one "that functions independently of carrier frequencies and can operate within a range of transmission-protocol environments."
Architecturally, these definitions suggest transceivers that perform up/down conversion between baseband and RF exclusively in the digital domain, reducing the RF interface to a transmit-channel power amplifier, low noise amplifier (LNA) for the receive path, and minimal analog filtering.
The forum defines different classes of SDRs based on the level (or tier) of capability and flexibility. Tier 0 devices, known as hardware radios (HRs), implement all the radio functionality in hardware and any changes in functionality require physical intervention to implement. All radios can be tuned to pick up a specific (limited) frequency range and it takes some amount of time to change the tuning. Traditionally, these tuning characteristics are defined in the radio hardware. The goal is to make these characteristics programmable.
Tier 1 devices are known as software-controlled radios (SCRs). At this level of the SDR hierarchy, only the control functions are implemented in software. For a given modulation standard, the baseband processing and radio front-end are fixed. Using multiple transceivers in the same device provides mutli-standard support. Software controls which transceiver to activate. An example of this type of device is the dual-mode cell phone supporting both CDMA and GSM.
Reconfigurable SDRs are the natural progression from SCRs and are now common in basestation applications. Known as Tier 2 SDRs, these devices employ software to control various modulation techniques, wideband and narrowband operation, security, and the waveform requirements of current and evolving standards over a broad frequency range. When people talk about SDR, they're generally referring to Tier 2 devices.
The ideal software radio, known as Tier 3, includes all the features of the of the Tier 2 reconfigurable SDRs, but eliminates the analog amplification or heterodyne mixing prior to digital-to-analog conversion. Programmability extends over the entire system, with all analog conversions taking place at the antenna, speaker, and microphone in the case of a cellular handset
Looking at far the industry is from delivering a Tier 3 radio, we must quantify the gap between today's SDRs and the Tier 3 radio. This requires the definition of some measurements metrics. One criterion is the RF circuitry's complexity. A larger quantity of RF/analog functionality provides less SDR flexibility, while reduced RF/analog circuitry implies more SDR flexibility. A potentially more effective criterion is to characterize the radio's reconfigurability, which comes from digitizing the signal-processing functions typically in the RF/analog domain.
SDR technology can be considered as the driving force behind the next generation of RF systems. For today's SDR applications, much of the computational power is coming from improvements in the baseband-processing designs, faster interface buses, and process-node advances. Software radios are still currently limited by specialized RFICs that target particular frequency bands and waveforms. To realize low-cost solutions, digital functionality must use the latest process node in deep-submicron CMOS technology such as 130, 90, and 65 nm in the future.
To provide the required dynamic range and power, RF and analog circuits require higher supply voltages than deep-submicron CMOS technologies can support. With supply voltages in digital processes falling to 1.5 V and below and threshold voltages around 600 mV, there's little headroom available to conduct significant signal processing in the analog domain. As digital processing moves to the latest process node to reduce cost and increase functional density, analog and RF functions lag behind at earlier process nodes, resulting in lower system integration levels, increased system cost, and reduced SDR capabilities.
To quote Bogdan Staszewski from TI's Wireless Analog Technology Center at RFIC2005, "When designing highly integrated RF circuits in deep submicron CMOS processes, we are faced with the paradigm shift that the time domain resolution of a digital-signal edge transition is superior to the voltage resolution of analog signals."
TI has exploited this paradigm shift extensively in their new Digital Radio Processor architecture. The company has created a design that compensates for the weaknesses of modern digital process nodes with respect to analog and RF functions by exploiting the strengths of the process in terms of transition speed and logic density. Traditional analog and RF functions are being migrated into the digital domain to increase integration levels, reduce system cost, and ultimately push toward the goal of Tier 2 and beyond SDR capabilities. These and other advances in software-tunable/configurable RF technology will lead to truly cost-effective front-end designs. At that point, the commercialization of SDRs will really begin.
Currently, most commercial SDR technology is implemented in base-station systems but it's only recently that SDR technology is being developed for the cellular handset market. The performance and cost challenges related to RFIC design for cellular handsets as well as the requirement to pack more functionality on a die may be addressed with SDR by using digital logic to implement traditional analog functions. For such designs, the key tradeoffs shift from performance and design cost to flexibility, small form factor, and product variances. While work is progressing, practical, commercially viable Tier 2 and 3 SDR systems are still a few years off.
If widespread use of SDRs is still off in the future, why are we hearing about cognitive radios? Cognitive Radio is generally recognized as a radio system that adapts to the spectral environment and the users activity to select the correct radio interface, channel, data rate, etc., to provide maximum data throughput to the user's application. SDR, on the other hand, is a radio system that adapts only to the network environment.
Today, using the same radio, you can't utilize both 2.4GHz and 5GHz wireless LAN signals without implementing two separate radio paths to deal with their respective waveforms. Assuming that SDR technology will address this problem within the next few years, the next hurdle will be to find a solution for SDR to dynamically receive or transmit different kind of waveforms automatically, adapting itself to the local environment and seeking out open frequencies to communicate. Cognitive Radio is the subject of extensive academic and R&D efforts.
Cedric Paillard is the joined Director TECHinsights at Semiconductor Insights. He holds a Bachelor and Master of Science in VLSI System Engineering from the University of Manchester Institute of Science and Technology. He also has a Bachelor of Engineering from the Ecole Superieure de Technolgie Electrique. Paillard can be reached at cedricp@semiconductor.com.
