The Poisson equation dx2d2u=f in the unit interval [0,1], with u(0)=u(1)=0, is discretised with the formula
ui−1−2ui+ui+1=h2fi,1⩽i⩽n,
where u0=un+1=0,h=(n+1)−1, the grid points are at x=ih and ui≈u(ih).
(a) Write the above system of equations in the vector form Au=b and describe the relaxed Jacobi method with relaxation parameter ω for solving this linear system.
(b) For x∗ and x(ν) being the exact and the iterated solution, respectively, let e(ν):=x(ν)−x∗ be the error and Hω be the iteration matrix, so that
e(ν+1)=Hωe(ν)
Express Hω in terms of the matrix A and the relaxation parameter ω. Using the fact that for any n×n Toeplitz symmetric tridiagonal matrix, the eigenvectors vk(k=1,…,n) have the form:
vk=(sinikx)i=1n,x=n+1π
find the eigenvalues λk(A) of A. Hence deduce the eigenvalues λk(ω) of Hω.
(c) For A as above, let
e(ν)=k=1∑nak(ν)vk
be the expansion of the error with respect to the eigenvectors (vk) of Hω.
Find the range of the parameter ω which provides convergence of the method for any n, and prove that, for any such ω, the rate of convergence e(ν)→0 is not faster than (1−c/n2)ν when n is large.
(d) Show that, for an appropriate range of ω, the high frequency components ak(ν) (2n+1⩽k⩽n) of the error e(ν) tend to zero much faster than the rate obtained in part (c). Determine the optimal parameter ω∗ which provides the largest supression of the high frequency components per iteration, and find the corresponding attenuation factor μ∗ assuming n is large. That is, find the least μω such that ∣∣∣∣ak(ν+1)∣∣∣∣⩽μω∣∣∣∣ak(ν)∣∣∣∣ for 2n+1⩽k⩽n.