generate temporally correlated channel？

usually we can produce the independent channel as following
h =(randn(multipath,num_block_channel)+j*randn(multipath,num_block_channel))/sqrt(2)/sqrt(multipath);
how to including the temporal correlation?
any useful type will be appreciated!

h(t+delta_t) = h(t)*sqrt(c)+h_random*sqrt(1-c)
c 是相关系数，和delta_t有关。

% Auto-correlation function is specified
clear all;n_data=25;
%time varying channel model; a is Gaussian process
a_leng=5; a(:,1)=[5 4 3 2 1]';%initial values of channel coefficient
std=[5 4 3 2 1]; %standard deviation of channel coefficient
for rho_case=1:5
rho=1-(rho_case-1)*0.001; % Five Cases of rho
for n =2:n_data+a_leng-1
a(:,n)=rho*a(:,n-1)+sqrt(1-rho^2)*randn(a_leng,1)std(:);
end
figure(rho_case)
for i=1:a_leng
plot(a(i,:));hold on
end
hold off
string_=strcat('Correlation Coefficient rho=',num2str(rho)); title(string_);
ylabel('Channel Coefficients');xlabel('Symbol Index'); axis([0, 25, 0, 8.5]);
end