Multi step ahead forecasting with Radial basis function neural network

I want to forecast multi step (k=6) ahead with Radial basis function neural network (RBFNN) with time series data. I have one output and 2 input variables. The code is given below:

net = newrb(p, t, 0.5)
outputs = net(p)
perf = perform(net, outputs, t)

I am stuck with what to do after this to make multi step ahead prediction?

Please help and thanks in advance.

[KSSV]

What is your inpu and output?

[KSSV]

anurag kulshrestha commented:

The input and output, both are time-series. Input is 90*2 and output is 90*1 time series data.

[KSSV]

Save the net..and give the input, you shall get the predicted value.

[anurag kulshrestha]

I want to do multi step forecatsing recursively.