minimize a 2d function including phase

I am trying to minimize a function in which I use the angle() function. The function mathematically is smooth and monotonous, and therefore if someone search for the min there should be only one answer. However for different starting values I get different results.

In a nutshel, this is the function I try to minimize,

function y = delta(n1)

      n = n1(1)-1i*n1(2);
      T = ( 4*n/(n + 1)^2 ) * exp(-1i*(n - 1) * 3) ;

      deltaRho = log( sqrt( T * conj(T)) ) - log(0.05);
      deltaPhi = angle(T) - 10;

      y = deltaRho^2 + deltaPhi^2;

end

and I call,

fminsearch(@delta,n0)

Is there any way to fix the angle() jumps and find the unique solution to that monotonous function.

George.

[Matt J]

Your function is from R^2-->R. Please give your definition of monotonicity of such a function.

Also, since your function uses complex arithmetic, it would probably be wise to insert some calls to REAL to get rid of complex-valued floating point errors.

[Matt J]

Isn't the obvious minimum at n=[0;0] where y=-inf