Hidden Markov model based resource estimation and call admissions in mobile cellular networks

A new call admission algorithm with better performance guarantees is proposed with emphasis on decoupling of the given cell from communication with its neighbours. The algorithm is formulated with the view that the bandwidth requirements of anticipated incoming and outgoing handoff calls can be trained. Accordingly, a hidden Markov model (HMM) based resource estimation and call admission scheme is suggested. A novel way of maintaining the imposed performance metrics is outlined. Simulation results validate the performance of the algorithm against the existing ones