Multi-DUT PXI approach reduces small cell manufacturing cost: Page 3 of 7

November 24, 2014 //By Thomas Deckert, National Instruments
Small cell base stations are a key technology that will increase capacity and coverage of today’s cellular mobile networks. In fact, the growth in the small cell industry is creating a new cost structure of base station manufacturing test.
The majority of test plans for base stations will execute each measurement sequentially – performing the entire suite of test steps on a single device before moving on to test the next unit. The problem with this approach is that RF test equipment is used only a fraction of the time to do calibration as well as transmitter and receiver tests. Depending on the complexity of the product and the test plan, RF measurements typically take between 30 and 50 percent of the time; the other 50 to 70 percent the RF equipment remains idle (Figure 3).

For large base station equipment that is capable of spanning macro cells, poor test equipment utilization might be acceptable: Such products are made in moderate volumes and typically sell for a lot more than a small cell base station. Small cell base stations more resemble a consumer-grade product in that they are made in larger quantities and come at a lower price than macro equipment. Multi-DUT testing is a cost-efficient approach used by handset manufacturers for some time now. The same technique holds substantial benefits for small cell testing as well.

Multi-DUT testing

Multi-DUT testing (also known as “multi-site”) is an advanced test procedure that has the explicit goal of maximizing the utilization of the more valuable test assets such as RF vector signal analyzers. Implementing a multi-DUT test plan requires test engineers to reorganize their test sets and procedures such that they can pipeline the DUTs through the individual test phases (Figure 4). From these modifications, they can reap tremendous benefits in utilization of the instrument and, ultimately, in the cost of test.


Figure 4: Parallel testing and auto-scheduling reduce test time and, thus, increase test throughput. Click image to enlarge.

There are several questions one may ask in this process: How many devices should be tested in parallel? How must a test be structured for efficient pipelining of the individual phases? What about software and processing requirements? What is a good switching concept to route all the relevant signals from the multiple DUTs to the test equipment? Let us address each of these questions in a bit more detail.


Design category: