Linear Programming Algorithms for Sparse Filter Design

被引:111
作者
Baran, Thomas [1 ]
Wei, Dennis [1 ]
Oppenheim, Alan V. [1 ]
机构
[1] MIT, Dept Elect Engn & Comp Sci, Cambridge, MA 02139 USA
关键词
Sparse filters; linear programming; FIR digital filters; linear arrays; FIR DIGITAL-FILTERS; RESPONSE MASKING APPROACH; COEFFICIENTS; APPROXIMATION; INTERPOLATION; DECIMATION; MAGNITUDE; ARRAYS; BAND;
D O I
10.1109/TSP.2009.2036471
中图分类号
TM [电工技术]; TN [电子技术、通信技术];
学科分类号
0808 ; 0809 ;
摘要
In designing discrete-time filters, the length of the impulse response is often used as an indication of computational cost. In systems where the complexity is dominated by arithmetic operations, the number of nonzero coefficients in the impulse response may be a more appropriate metric to consider instead, and computational savings are realized by omitting arithmetic operations associated with zero-valued coefficients. This metric is particularly relevant to the design of sensor arrays, where a set of array weights with many zero-valued entries allows for the elimination of physical array elements, resulting in a reduction of data acquisition and communication costs. However, designing a filter with the fewest number of nonzero coefficients subject to a set of frequency-domain constraints is a computationally difficult optimization problem. This paper describes several approximate polynomial-time algorithms that use linear programming to design filters having a small number of nonzero coefficients, i.e., filters that are sparse. Specifically, we present two approaches that have different computational complexities in terms of the number of required linear programs. The first technique iteratively thins the impulse response of a non-sparse filter until frequency-domain constraints are violated. The second minimizes the 1-norm of the impulse response of the filter, using the resulting design to determine the coefficients that are constrained to zero in a subsequent re-optimization stage. The algorithms are evaluated within the contexts of array design and acoustic equalization.
引用
收藏
页码:1605 / 1617
页数:13
相关论文
共 36 条
[1]   SOME EFFICIENT DIGITAL PREFILTER STRUCTURES [J].
ADAMS, JW ;
WILLSON, AN .
IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS, 1984, 31 (03) :260-266
[2]   A NEW APPROACH TO FIR DIGITAL-FILTERS WITH FEWER MULTIPLIERS AND REDUCED SENSITIVITY [J].
ADAMS, JW ;
WILLSON, AN .
IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS, 1983, 30 (05) :277-283
[3]  
BARAN TA, 2007, P 41 AS C SIGN SYST, P1088
[4]   AN APPROACH TO PROGRAMMABLE CTD FILTERS USING COEFFICIENTS 0, +1, AND -1 [J].
BATEMAN, MR ;
LIU, B .
IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS, 1980, 27 (06) :451-456
[5]  
Bertsimas Dimitris, 1997, Introduction to linear optimization, V6
[6]   THINNING DIGITAL-FILTERS - A PIECEWISE-EXPONENTIAL APPROXIMATION APPROACH [J].
BOUDREAUX, GF ;
PARKS, TW .
IEEE TRANSACTIONS ON ACOUSTICS SPEECH AND SIGNAL PROCESSING, 1983, 31 (01) :105-113
[7]   OPTIMAL-DESIGN OF FIR FILTERS WITH THE COMPLEX CHEBYSHEV ERROR CRITERIA [J].
BURNSIDE, D ;
PARKS, TW .
IEEE TRANSACTIONS ON SIGNAL PROCESSING, 1995, 43 (03) :605-616
[8]   Robust uncertainty principles:: Exact signal reconstruction from highly incomplete frequency information [J].
Candès, EJ ;
Romberg, J ;
Tao, T .
IEEE TRANSACTIONS ON INFORMATION THEORY, 2006, 52 (02) :489-509
[9]   A trellis search algorithm for the design of FIR filters with signed-powers-of-two coefficients [J].
Chen, CL ;
Willson, AN .
IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS II-ANALOG AND DIGITAL SIGNAL PROCESSING, 1999, 46 (01) :29-39
[10]  
Crochiere R., 1983, MULTIRATE DIGITAL SI