A T-matrix based method of solution of the multiple scattering problem
presented in Gumerov and Duraiswami (J. Acoust Soc. Am., 112, 2002, 2688-
2701) can in practice be applied to computation of relatively small
problems (up to hundreds of scatterers), since the number of operations it
requires grows with the the number of scatterers N as O(N^3), and with the
sixth power of the wavenumber. In this study we present a method, which 
combines iterative techniques with the multilevel fast multipole method
that employs fast translation algorithms. We show that in this case the
number of operations for a matrix-vector multiplication grows with N as O(N
log N) and with the third power of the wavenumber. We present details of
the method. We also discuss convergence of the iterative techniques,  
selection of the truncation number, errors in the solution, and other
issues related to the use of the method in practice. Results of the
solution of test problems obtained with the method for N can be
substantially large (N ?? 102 to 104; for different ranges of
wavenumbers).
UMIACS-TR-2004-42