In compressed sensing, the nullspace property gives necessary and sufficient conditions on the reconstruction of sparse signals using the techniques of -relaxation. The term "nullspace property" originates from Cohen, Dahmen, and DeVore.[1] The nullspace property is often difficult to check in practice, and the restricted isometry property is a more modern condition in the field of compressed sensing.

The technique of -relaxation edit

The non-convex  -minimization problem,

  subject to  ,

is a standard problem in compressed sensing. However,  -minimization is known to be NP-hard in general.[2] As such, the technique of  -relaxation is sometimes employed to circumvent the difficulties of signal reconstruction using the  -norm. In  -relaxation, the   problem,

  subject to  ,

is solved in place of the   problem. Note that this relaxation is convex and hence amenable to the standard techniques of linear programming - a computationally desirable feature. Naturally we wish to know when  -relaxation will give the same answer as the   problem. The nullspace property is one way to guarantee agreement.

Definition edit

An   complex matrix   has the nullspace property of order  , if for all index sets   with   we have that:   for all  .

Recovery Condition edit

The following theorem gives necessary and sufficient condition on the recoverability of a given  -sparse vector in  . The proof of the theorem is a standard one, and the proof supplied here is summarized from Holger Rauhut.[3]

  Let   be a   complex matrix. Then every  -sparse signal   is the unique solution to the  -relaxation problem with   if and only if   satisfies the nullspace property with order  .

  For the forwards direction notice that   and   are distinct vectors with   by the linearity of  , and hence by uniqueness we must have   as desired. For the backwards direction, let   be  -sparse and   another (not necessary  -sparse) vector such that   and  . Define the (non-zero) vector   and notice that it lies in the nullspace of  . Call   the support of  , and then the result follows from an elementary application of the triangle inequality:  , establishing the minimality of  .  

References edit

  1. ^ Cohen, Albert; Dahmen, Wolfgang; DeVore, Ronald (2009-01-01). "Compressed sensing and best 𝑘-term approximation". Journal of the American Mathematical Society. 22 (1): 211–231. doi:10.1090/S0894-0347-08-00610-3. ISSN 0894-0347.
  2. ^ Natarajan, B. K. (1995-04-01). "Sparse Approximate Solutions to Linear Systems". SIAM J. Comput. 24 (2): 227–234. doi:10.1137/S0097539792240406. ISSN 0097-5397. S2CID 2072045.
  3. ^ Rauhut, Holger (2011). Compressive Sensing and Structured Random Matrices. CiteSeerX 10.1.1.185.3754.