Software defined radio: defining the challenges

Software defined radio (SDR) has become a hot topic in the industry and is seen as a solution for many of the development problems for new radio platforms.  It is generally perceived that SDR is the answer to the need for flexibility, upgradability, and the problems of implementing multiple radio standards alternatively and even running several services in parallel. An SDR-based platform can also reduce the high development costs of a full new ASIC-based solution.

But what is really meant by software defined radio? It is not only a system on a chip (SoC) radio receiver with "software inside." This approach is already state of the art. Modern radio receiver chipsets already count on embedded cores, where software implements many parts of the functionality. Even radio receivers for the old fashioned FM radio are implemented digitally as software solutions on embedded cores or standard DSPs. The software may be embedded in a chip or chipset without being visible to the outside world.

The advantages of such solutions are clear. A software-based solution can be modified in the very late stages of the development process, and to some extent be reused across SDR platform generations. Certain standard software functions and libraries on the market are licensable as intellectual property (IP) or core developer kits (CDK). Audio and video decoder software libraries are a typical example of such IP. Another advantage of these software solutions is the possibility for updates or bug fixes after delivery of a development, when a problem encountered in the field requires countermeasures.

Defining SDR
Software defined radio is more than just an implementation in a software solution. It means the hardware platform obtains its functionality only when the software and the hardware development is independent, but it is influenced by the software development. The concept relies on the split between a suitable hardware platform and a suitable set of functionalities to be implemented on the platform.

There are two principal scenarios:

First, the hardware platform to be used is a standard platform, which can be ordered as a standard component. Examples for such hardware platforms are the C64+ DSP-family from Texas Instruments, the Blackfin DSP-family from Analog Devices or Infineon’s SDR20, among others.

This is different to the second case of proprietary programmable platforms with DSP IPs, like ARM, Tensilica, MIPS or others inside. All silicon vendors have such devices in their portfolio for existing or future radio applications. The main difference is that the knowledge of how to use the device most effectively requires a very close cooperation with the silicon vendor. This is a different business relationship to the silicon vendor compared to ordering standard components. The devices are often very heterogeneously built based on one or more embedded cores. The advantage is the possibility to optimize the platform for an application or family of applications. Normally the intended feature-set or applications are known upfront and the hardware has to be targeted for the purpose. Therefore, such SDR platforms are in principle a multi-standard terminal, which can be used for different radio applications alternatively or even in parallel.