The TileCal Online Energy Estimation for the Next LHC Operation Period

This work presents a study on the TileCal online energy estimation for the next LHC operation period (Run2). Due to hardware limitation, the algorithm that performs the online energy estimation must remain based on the linear and fast Optimal Filter (OF) technique, which envisages the use of the background covariance matrix in its design. During Run1, TileCal made use of an identity matrix to describe and the background covariance matrix, which showed to be a reasonable approximation for the low luminosity scenario. Under such conditions, the background for most of its cells comprised mainly electronic noise which can be approximated as a uncorrelated Gaussian process. However, as the pile-up introduces a non-Gaussian component to the background, the OF method presents larger variance and it becomes biased. The use of the background covariance matrix in the design of the OF weights is expected to improve the energy estimation performance. The results for the scenario considered show that for high occupancy cells, an improvement of 60% in terms of estimation error is achieved by replacing the identity matrix to the background covariance matrix.