1: \begin{abstract}

2: To improve the convergence property of approximate message-passing (AMP), 

3: convolutional AMP (CAMP) has been proposed. CAMP replaces the Onsager 

4: correction in AMP with a convolution of messages in all preceding iterations 

5: while it uses the same low-complexity matched filter (MF) as AMP. This paper 

6: derives state evolution (SE) equations to design the Bayes-optimal denoiser 

7: in CAMP. Numerical results imply that CAMP 

8: with the Bayes-optimal denoiser---called Bayes-optimal CAMP---can achieve 

9: the Bayes-optimal performance for right-orthogonally invariant sensing 

10: matrices with low-to-moderate condition numbers.  

11: \end{abstract}

12: