A DYNAMIC BANDWIDTH ALLOCATION SYSTEM IN XG-PONS BASED ON STATISTICAL ESTIMATION OF BUFFER INPUTS AND OUTPUTS

The exchanges in the New Generation Optical Access Network (NG-PON), are mainly based on bandwidth request by Units (ONU) and its granting Line Termination (OLT). This process (request-grant) not being synchronized, will create ONU buffer a long delay loss of sensitive packets. Many authors, from signaling to prediction, have proposed models control these parameters, however only those prediction reduce queues but also waste over- or under-allocating. Our goal is propose model that accurately allocates while minimizing their queue delay. As such, we Statistical Queue Estimation Buffer Inputs Outputs (ESTEST). In execution ESTEST memorizes requirements used transmit data ith upstream frames. With two parameters (requirements used), our algorithm computes traffic intensity (𝜌) predicts size ONU. On 100 observations compute mean standard deviation 𝜌th values with reduced centered normal distribution, find confidence interval estimation.