Demand Forecasting

Demand Forecasting
Forecasting future demand for bandwidth is an estimate of something that will happen in the future is based on the existing data in the present and the past (historical data).
Many forms of trend data. For example, in forecasting methods, there are four common forms of trend data.
1. The Linear Model
Linear trend is the trend of the data in which changes based on the time is fixed. Yt = β0 + β1T

Figure 2.8 Example Graph Linear Models
2. Quadratic Model
Quadratic trend is the trend of the data curve patterned arch (curvature). Has a quadratic trend following model: Yt = β0 + + β2T2 β1T

Figure 2.9 Example Graph Quadratic Model
3. Models Exponential Growth
Exponential growth trend is the trend of the data in which changes progressively increases exponentially. Y = β0eβ1T or ln (Y) = lnβ0 + β1T
(2.9 )

Figure 2.10 Example Graph Exponential Growth Model
4. Curve Model S
S-curve characteristic is slow growth initially, then increased rapidly up to a certain point and then slows down again and tend to stay. Trend curve S has the following model: Yt = E (β0 + (β1 / T)) or ln (Y) = β0 + (β1 / T)

Figure 2.11 Example of Graph Model S curve

Forecasting future demand for bandwidth is an estimate of something that will happen in the future is based on the existing data in the present and the past (historical data), to forecast customer demand data is needed bandwidth bandwidth users during the last 5 years .