Design Article
Tell us What You Think
We want to know what you thought about this Design. Let us know by adding a comment.
Case study: Threat simulation for multi-port radar and electronic warfare systems
John Eckert, Agilent Technologies
11/22/2011 4:53 PM EST
For many types of multiple-input devices, precise testing requires accurate control over the amplitude and phase of all stimulus signals applied to the inputs of the unit under test (UUT). Examples include phased-array radar systems and multiple-input/multiple-output (MIMO) devices used in commercial communications and wireless networking. In these applications, precise control makes it possible to accurately simulate parameters such as the angles of arrival of incident signals.
This type of testing is useful in research, product development, design verification, manufacturing, and calibration. In all cases, a test solution that provides flexible synchronization of test signals can help reduce test time—an advantage that is especially beneficial in the manufacturing process. As an example, let's consider the process of developing a test platform to evaluate the performance of a four-port UUT. The results provide significant benefits from technical and operational perspectives.
Defining the problem
The testing of multiple-input receivers often presents a key challenge: delivering the desired signal at the end of long test cables. Creating an accurate signal simulation has two major requirements. One is the need to measure changes in phase and amplitude at the UUT, not in the test equipment itself. The other is the need to make real-time corrections to the amplitude and phase of each waveform. This level of control ensures proper alignment of the simulated incident signal in the presence of reflected signals from the inputs of the UUT.
For the four-port radar receiver, the challenge was to create a system that could serve as both a threat simulator and a receiver calibrator. To thoroughly test the UUT, the system had to provide phase accuracy of less than 1 degree from 100 MHz to 20 GHz—and do so at the end of 6-ft test cables. To support the intended usage model, the system had to meet two additional criteria: maintain its calibrated accuracy for at least 12 hours, and eliminate the need to manually attach and detach calibration standards during a test (i.e., support hands-free calibration).
Sketching the solution
The system has two major sections: coherent wideband stimulus and RF correction (see figure 1). The stimulus section generates two sets of complex waveforms that simulate real-world signals. The other section performs 12-term error-corrected stimulus measurements at the UUT and provides feedback for correction of the RF and microwave signals.
Although it isn’t shown, the system also includes a host PC running a test executive. The PC is linked to the LXI-compliant instruments through LAN connections and a router.
The left-hand side of figure 1 includes two types of instruments: vector signal generators and arbitrary waveform generators (AWGs). For the vector signal generator, the key specifications are phase accuracy of less than 1 degree and an amplitude imbalance of less than 0.1 dB. Key attributes of the AWG are fast waveform switching and waveform resolution of less than 0.25 ns.
Each generator/AWG pair produces a complex waveform that simulates a combination of reference and offset signals. This is accomplished by defining two sets of complex waveforms in each two-channel AWG and using those signals to drive the I and Q wideband modulation inputs of the vector signal generator.
To simulate angle of arrival, the AWGs are loaded with multiple copies of the reference signal. Reference-channel memory is loaded with a single reference waveform while offset-channel memory contains multiple copies of the reference signal, each of which has a different phase offset (see figure 2). The test executive simply calls up and plays the desired phase sequence.
We created the individual waveforms using MATLAB from The MathWorks, based on Ashland-generated examples (see figure 3).
This type of testing is useful in research, product development, design verification, manufacturing, and calibration. In all cases, a test solution that provides flexible synchronization of test signals can help reduce test time—an advantage that is especially beneficial in the manufacturing process. As an example, let's consider the process of developing a test platform to evaluate the performance of a four-port UUT. The results provide significant benefits from technical and operational perspectives.
Defining the problem
The testing of multiple-input receivers often presents a key challenge: delivering the desired signal at the end of long test cables. Creating an accurate signal simulation has two major requirements. One is the need to measure changes in phase and amplitude at the UUT, not in the test equipment itself. The other is the need to make real-time corrections to the amplitude and phase of each waveform. This level of control ensures proper alignment of the simulated incident signal in the presence of reflected signals from the inputs of the UUT.
For the four-port radar receiver, the challenge was to create a system that could serve as both a threat simulator and a receiver calibrator. To thoroughly test the UUT, the system had to provide phase accuracy of less than 1 degree from 100 MHz to 20 GHz—and do so at the end of 6-ft test cables. To support the intended usage model, the system had to meet two additional criteria: maintain its calibrated accuracy for at least 12 hours, and eliminate the need to manually attach and detach calibration standards during a test (i.e., support hands-free calibration).
Sketching the solution
The system has two major sections: coherent wideband stimulus and RF correction (see figure 1). The stimulus section generates two sets of complex waveforms that simulate real-world signals. The other section performs 12-term error-corrected stimulus measurements at the UUT and provides feedback for correction of the RF and microwave signals.
Figure 1: The simulator/calibrator system has two major sections: the coherent, wideband stimulus side and the RF correction side.
Although it isn’t shown, the system also includes a host PC running a test executive. The PC is linked to the LXI-compliant instruments through LAN connections and a router.
The left-hand side of figure 1 includes two types of instruments: vector signal generators and arbitrary waveform generators (AWGs). For the vector signal generator, the key specifications are phase accuracy of less than 1 degree and an amplitude imbalance of less than 0.1 dB. Key attributes of the AWG are fast waveform switching and waveform resolution of less than 0.25 ns.
Each generator/AWG pair produces a complex waveform that simulates a combination of reference and offset signals. This is accomplished by defining two sets of complex waveforms in each two-channel AWG and using those signals to drive the I and Q wideband modulation inputs of the vector signal generator.
To simulate angle of arrival, the AWGs are loaded with multiple copies of the reference signal. Reference-channel memory is loaded with a single reference waveform while offset-channel memory contains multiple copies of the reference signal, each of which has a different phase offset (see figure 2). The test executive simply calls up and plays the desired phase sequence.
Figure 2: Referencing the centerline, these 8-ns raised-cosine pulses have precise offsets of 90 and 180 degrees.
We created the individual waveforms using MATLAB from The MathWorks, based on Ashland-generated examples (see figure 3).
Figure 3: MATLAB produced these 300 MHz I and Q chirps.
|
|
|
|
|
Navigate to related information