Inching toward SDR

If you've ever restored to sitting on your over-stuffed suitcase in order to force it closed, just so you could fit that last pair of Bermuda shorts, then you'll have a sense for what's currently happening in the world of handset manufacturing. What started out as a simple device with a single radio operating at one frequency has, in less than a decade, evolved to pack more than five radios into a package slighter bigger than a pack of cigarettes.

 

Jump forward two or three years and its foreseeable that a mid-range handset sold into developed markets will need to support: GSM, GPRS, EDGE, W-CDMA, HSDPA, LTE, GPS, mobile TV, Wi-Fi, Bluetooth, FM radio and UWB. Add Wimax to the mix and the number of waveforms that a phone will need to support is mind-blowing - something has to give.

 

With the sophistication of today's mobile phones it is easy to forget that the handset, at heart, is a radio: the phone's basic function is to send and receive radio signals carrying voice or data information. Traditionally, radios have been implemented entirely in hardware, consisting of an antenna, RF front-end (amplifiers, receiver, transmitter) and baseband. New waveforms have been added by integrating new hardware.

 

At some point, it will become impractical to continue squeezing new, disparate radio hardware onto a handset circuit board because of the increase in handset size, complexity and cost. The attraction of software defined radio (SDR) is its ability to support multiple waveforms by re-using the same hardware while changing its parameters in software. This has enormous benefits for handset size, cost, development cycle, upgrade and interoperability.

 

True software defined radio solutions are not yet a commercial reality. While re-configurability of the baseband has been demonstrated by several vendors, power consumption of reconfigurable RF front-ends is still high. Antennas are the farthest away from re-configurability, although design developments such as phased array antennas, steerable antennas, polarization diversity and multi-band antennas are bringing this achievement closer.

 

Vested interest

 

One reason SDR in handsets has been slower to take off than many had expected is that the industry has invested heavily in (and consequently become very good at) tightly integrated solutions. Single ASICs that combine 2G and 3G functions are widely available and semiconductor suppliers like TI, Marvel and CSR have products that combine technologies like Wi-Fi, Bluetooth and UWB.

 

It cannot be denied that the cost, power and real-estate implications of such an exhaustively modular architecture have the potential to render the product unattractive.

 

Because these players have invested so heavily in the integrated ASIC approach, they have been more inclined to milk the R&D assets they have built around this, instead of moving rapidly toward SDR. In a recently published report, ARCchart examined the unique set of challenges posed by some of these new waveforms:

 

  • Mobile TV waveforms require dedicated antennas, RF front-end, baseband and codecs

  • LTE places enormous performance pressure on the antenna and RF front-end, requiring them to dynamically switch center frequencies, bandwidths, power scheduling as well as data rates

  • UWB requires extraordinary performance from the RF front-end to be able to process a bandwidth of 500 MHz

  • GPS antennas have specific structural and orientation requirements

The technical challenges of implementing these technologies will catalyze the shift to SDR. ARCchart believes that SDR will first be adopted in smartphones as these devices will be better able to absorb the higher cost of early SDR implementation. Across all devices, ARCchart estimates that SDR phone shipments will grow to more than 11% of the market in 2011, with Western Europe and North America being early adopter markets.

 

Matt Lewis is research director of ARCchart

 

Suggested Articles

Wireless operators can provide 5G services with spectrum bands both above and below 6 GHz—but that doesn't mean that all countries will let them.

Here are the stories we’re tracking today.

The 5G Mobile Network Architecture research project will implement two 5G use cases in real-world test beds.