Compressive sense based reconstruction algorithm for non-uniform sampling based data converter
Compressive sensing is an emerging field that attempts to prevent the losses associated with data compression and improve efficiency overall, and compressive sensing looks to perform the compression before or during capture, before energy is wasted. Here, a reconstruction algorithm is proposed for a compressive sensing successive approximation register (SAR) analog-to-digital converter (ADC). Accordingly, an analog signal is converted to a first digital signal at a sampling frequency that is less than a Nyquist frequency for the analog signal, and a second digital signal is constructed from the first digital signal with a box constrained linear optimization process such that the second digital signal is approximately equal to an analog-to-digital conversion of the analog signal at the Nyquist frequency for the analog signal.
1. An apparatus comprising:
an analog-to-digital converter (ADC) that is configured to generate a first digital signal from an analog signal; and
a controller that is coupled to the ADC so as to provide a sample signal to the ADC and to receive the first digital signal from the ADC, wherein the frequency of the sample signal is less than a Nyquist frequency for the analog signal, and wherein the controller generates a second digital signal from the first digital signal by employing a box constrained linear optimization process, and wherein the second digital signal is approximately equal to an analog-to-digital conversion of the analog signal at the Nyquist frequency for the analog signal,
wherein the ADC further comprises a successive approximation register (SAR) ADC.
2. The apparatus of claim 1 , wherein the SAR ADC further comprises:
a sample-and-hold (S/H) circuit that is configure to receive analog signal and the sample signal;
a comparator that is coupled to the S/H circuit;
SAR logic that is coupled to the comparator and the controller; and
a capacitive digital-to-analog converter (CDAC) that is coupled to the SAR logic and the comparator.
3. The apparatus of claim 2 , wherein the controller provides a clocks signal to the SAR logic and the comparator.
4. The apparatus of claim 3 , wherein the box constrained linear optimization processes minimizes a frequency sparse signal ({right arrow over (α)}) subject to the following constraint:
| y i −( A {right arrow over (α)} ) i |≦e i, i=1, . . . , M
where y is the second digital signal, A is the measurement matrix, e is an error, and M is the number of measurments.
5. The apparatus of claim 4 , wherein the controller further comprises a processor having a memory with a computer program embodied thereon.
6. A method comprising:
converting an analog signal to a first digital signal at a sampling frequency that is less than a Nyquist frequency for the analog signal; and
constructing a second digital signal from the first digital signal with a box constrained linear optimization process such that the second digital signal is approximately equal to an analog-to-digital conversion of the analog signal at the Nyquist frequency for the analog signal,
wherein the step of converting further comprises:
sampling the analog signal at a plurality of sampling instants; and
determining a digital value for the analog signal at each sampling instant by successive approximation.
7. The method of claim 6 , wherein the sampling instants occur as non-uniform intervals.
8. The method of claim 7 , wherein the box constrained linear optimization processes minimizes a frequency sparse signal ({right arrow over (α)}) subject to the following constraint:
| y i −( A {right arrow over (α)} ) i |≦e i ,
where y is the second digital signal, A is the measurement matrix, and e is an error.