1function [W,rhohat]=qsys_gig1_approx_klb(lambda,mu,ca,cs)
2% [W,rhohat]=QSYS_GIG1_APPROX_KLB(LAMBDA,MU,CA,CS)
4% kramer-langenbach-belz formula
7 g=exp(-2*(1-rho)*(1-ca^2)^2/(3*rho*(ca^2+cs^2)));
9 g=exp(-(1-rho)*(ca^2-1)/(ca^2+4*cs^2));
11W=1/mu*((rho/(1-rho))*((cs^2+ca^2)/2)*g+1);
12rhohat = W*lambda/(1+W*lambda); % so that M/M/1 formulas still hold