Software Defined Radio
How does it help?
Many of the visions of future wireless communications involve devices connecting to a wide range of different networks such as 2G, 3G, WiFi, BlueTooth and perhaps others. They may even involve devices modifying their behaviour, perhaps as they discover new types of network or as home networks add additional functionality.
SDR Technology Today
A simplified block diagram of a SDR, implemented using technology available today, is shown in Figure 1. An antenna system is required to handle the air/electrical interface. This might range from a single, passive antenna to an array of ‘active’ antennas which are reconfigurable, controlled by software. The interface between the analogue and digital domains is be handled by analogue-to-digital converters (ADCs) on the receive side and digital-to-analogue converters (DACs) on the transmit side.
Figure 1 A block diagram of an SDR implemented with today’s technology
Due to limitations in these devices and for the foreseeable future at least, the radio frequency (RF) front-end will still have to be implemented in the analogue domain. Once in the digital domain, a SDR might use a device such as a field-programmable gate array (FPGA) to implement certain ‘high-speed’ functions, such as mixing, filtering and sample rate conversion and/or ‘general purpose’ processors or digital signal processors (DSPs) to implement ‘low speed’ functions. The demodulated data are passed to/received from higher level ‘application’.
The ability of devices to work in multiple different environments is often termed “multi-modal” capability. At present this is achieved by incorporating into the handset the chipsets from each of the different standards, e.g. 3G and BlueTooth. While such an approach works, it is relatively expensive and inflexible and will likely become more so as the range of different wireless technologies grows. An alternative might be for communication devices to be designed more like computers with general purpose processing capabilities and different software for different applications. Such a device could then call up, or download, the appropriate software for the particular communications requirement currently in use. It could also seek new versions of software, or “patches”, in the case where changes had been made to the communications technology. Another benefit might be increased economies of scale, with manufacturers producing a single core chipset for the entire world market and then installing software appropriate for each country, or operator.
The underlying architecture needed to achieve this is often termed software-defined radio (SDR). SDR is heralded by some as a significant step in the evolution of radio. In the future this flexibility might enable the more efficient use of the available spectrum through rapid deployment of the latest radio technologies.
Ofcom is researching SDR to:
- Understand the likely progress of the technology and the limits of what can be achieved as an input to predicting the likely deployment patterns and timescales for new services.
- Examine whether there might be any regulatory roadblocks for such radical new technology.
As it is mentioned earlier, SDRs might simplify the implementation of multi-modal devices and bring additional flexibility by allowing patches and enhanced functionality to be downloaded to the terminal. We also noted that SDR might be an enabling technology for cognitive radios and perhaps other advances in spectrum usage.
However, there are many issues with the implementation of SDR including the difficulties in implementing antennas, in the speed of chipsets, in the battery power required and in the cost, particularly of handsets. This leads us to conclude that SDR will be gradually implemented, starting with devices able to download minor protocol changes, likely around 2008, through the devices able to make major changes to the radio solutions they deploy around 2015.