In this paper, an Artificial Neural Network (ANN) structure using Multi-Layer Perceptron with Back Propagation (MLP-BP) and Legendre Neural Network (LeNN) applied to location management problem in the cellular system is proposed. The nonlinear Neural Network (NN) that is computationally-efficient solves the issues associated with location management. A large amount of computational time for learning in the case of feed-forward NN such as Back Propagation (BP) is a major issue. The proposed LeNN, basically a single layer structure in which nonlinearity is introduced where the input pattern is enhanced with nonlinear functional expansion, is simpler than MLP-BP. The novelty of the proposed work is: less computational time is required in LeNN than in MLP-BP. The simulation results show that LeNN outperforms the other proposed two techniques on performance error. Using simulation, the effectiveness of subscriber prediction is evaluated. It is also shown that the proposed network is computationally cheap and gives better classification accuracy than MLP classifier.

Identification of the location of the Mobile Terminal (MT) in a cellular network is the main objective of the location management algorithm. In cellular network, when a call arrives to a user in order to route the call appropriately, it is necessary to trace the mobile user correctly and efficiently (Bar-Noy et al., 1994). There are two fundamental operations in locating a mobile terminal in cellular mobile networks, i.e., location tracking (paging) and location update (registration). In ‘Paging’ technique, when a call arrives, the network sends a signal to investigate the subscriber, but in ‘Location Update’ technique, the subscriber sends the signal to inform the network (Rose and Yates, 1995). When a subscriber moves to a new location area, basically location update occurs. When there is an incoming call for a mobile terminal, paging process occurs which is initiated by a mobile terminal. When location updates are frequent, a tradeoff exists between the frequency of location update and paging costs; fewer number of calls need to be paged as there is less uncertainty about the user’s position. On the other hand, the cost of paging increases when location updates are infrequent (Akyildiz and Ho, 1995).

Telecommunications Journal, Location prediction, Neural Network (NN), Multi-Layer Perceptron (MLP), Legendre Neural Network (LeNN).