Enhancemant of the Statistical Learning Automated Healing (SLAH) technique using packet scheduling

Automated healing aims to reduce cost of network operations by automated fault diagnosis and rectification. This paper investigates the use of Packet Scheduling (PS) in automated healing. PS has been integrated into a previously proposed scheme of Statistical Learning Automated Healing (SLAH) for LTE. SLAH locally optimizes the Radio Resource Management (RRM) parameters of the faulty eNodeBs (eNBs). SLAH uses Logistic Regression (LoR) to extract the closed form relationship between the RRM parameters and the network measurements which are in the form of Key Performance Indicators (KPIs). This paper uses PS in SLAH to achieve the required minimum coverage constraint on an eNB by coverage/capacity compromise. Similarly, if this minimum coverage requirement of an eNB is already satisfied, additional capacity gain for an eNB can be achieved. This enhanced SLAH methdology has been used to rectify faults due to excessive interference suffered by an eNB from its first tier neighbours. Simulation results of the case study done prove that this technique converges in few iterations.