I am not sure if I understand your question(), so i looked at some of your publications and they seem to focus on cross section or panel data. I hope this is not too low of an answer to get you going.
Consider adding a variable, let's call it time, where time = 0 for your first observation and increases by one with each observation. If your data starts in 2004 - quarter 1 (20051), then use 0 for 20051, use 1 for 20052, 2 for 20053 and so forth. Then as you forecast beyond the latest data, your use the estimated coefficients including the time variable to estimate the value of the dependent variable. Obviously, in addition to the variable time, you will need variables that tend to reveal the value of the dependent variable well in advance or information that is given (for example, a dummy variable that takes on a value of 1 if it is a 4th quarter variable and 0 otherwise. The quarters are known in advance and seasonal patterns might be able to be modeled by using up to three of these quarterly dummy variables. By the way one cannot use 4 quarterly dummy variables, if the intercept is also estimated because each of these quarterly dummy coefficients becomes additive with the intercept to effectively modify the intercept. See the dummy variable trap if you are interested.
The simple mathematics of inserting known values for your variables into the equation with your coefficients allows forecasting in this really simple approach. There are MUCH better way to forecast; however, it seemed like this might help to get you thinking of simple time series forecasting. I hope it helps and if I missed the point of your question, please let me know.
Time Series is a sequence of well-defined data points measured at consistent time intervals over a period of time. Data collected on an ad-hoc basis or irregularly does not form a time series. Time series analysis is the use of statistical methods to analyze time series data and extract meaningful statistics and characteristics about the data