Software-defined Testing to Drive T&M Equipment R&D

By: STEPHEN LAS MARIAS, EDITOR
( 1 Jan 2007 )

Developments in the wireless world— such as emergence of newer standards as well as convergence between existing ones—continue to offer a great challenge for test and measurement (T&M) equipment manufacturers. Time-tomarket pressures are also fueling up the need for stronger R&D on T&M solutions, for as product development begins even before the wireless standards are finalized, the solutions should also have the flexibility to accommodate further changes as the standards evolve.

Separate stand-alone instruments are traditionally used for every communications standards to be tested. However, the current pace of technology development in the wireless space is making this approach unproductive and not costeffective.

In an interview with Wireless Design and Development Asia, Chandran Nair, Managing Director, Southeast Asia, National Instruments, explains the advantages of software-defined route to testing, and how this can increase flexibility, and cost-effectiveness, and offer faster time-to-market to electronics manufacturing companies. Excerpts:


In the wireless world, what is the current scenario regarding software-defined testing and dedicated testing? What is the market size of each of these, and what is the growth rate?
The general assumption is that the RF test market is about $5 billion and $10 billion, with the softwaredefined test segment as the fastest growing, as shown by the market forces such as the US and European military forces looking into synthetic instruments, and the largest telecommunication companies moving into software-defined test. Another factor to the growth of software-defined testing is the continuing adoption of PXI.

What do you think are the most important technical and economic requirements in wireless-related test instruments and how far can SDR meet these requirements?
The cost of systems is one of the major challenges for T&M equipment manufacturers. Whether it is a phone, DVD player or plasma TV, prices will definitely come down. But the biggest challenge T&M equipment companies are facing is that test system pricing has not reduced at exactly the same rate. For instance, from a semiconductor perspective, the cost of transistors is declining, but the cost of test per transistor is almost the same. In the past, the artificiality of traditional instrumentation gives the customers no leverage to bargain with test equipment suppliers. The way electronics manufacturers can get over it is by aligning the software defined test metrology that is based on modern instrumentation, which leverages commercial off-the-shelf technology.

With such a confusing multiplicity of wireless standards, how does the test engineer get the best out of software-defined testing?
The best way for a software engineer is to invest in the broad base of hardware instruments that can plug into a software-defined model. It will also depend on the direction of the company; also being trained in software helps you to write these new standards.

If communications standards change during the development phase, how fast is SDR in adapting itself to the new requirements?
The complexity of standards and the proficiency of the person developing the product will be among the factors that will come into play if communications standards change during the development phase. But the good thing is that there are a lot of companies whose business is to develop to these emerging standards. When there are new protocols, there are companies whose complete business model is to make softwares for these protocols. You can outsource, or buy from them; or if you want to do this in-house, it depends on your proficiency in softwares or the proficiency of your business unit. You can either buy or you can develop yourself.

Are there any specific applications that you can say that are better for SDR than hardware-ware defined test?
For any area of test that there are a large number of standards or there are many emerging standards, I think software-defined test is better, because it reduces time and costs, and it also allows you to test to these standards quickly. Hardware-defined test could be proven to be advantageous in some instances, for example, for an established technology that has been very widely adopted many years ago, which is not going to change. However, in the wireless domain, things change so fast; making something proprietary may not be advantageous.

Please say something about NI’s perceptions of the next-generation wireless software-defined testing? What kind of work is going on at NI on this front?
Looking into the future, the development of software-defined test will be very closely based on the evolution of commercial technology, such as in the field of commercial computers and analog-todigital and digital-to-analog chipsets. The biggest concentrated effort from National Instrument is to be able to not only operate the software-defined test, but also to use the software platform from the design, prototype, testing, and deployment phase, which will enhance the productivity of engineers, and most importantly reduce time-to-market.