Hopfield Neural Network Algorithm for Dynamic Resource Allocation in WCDMA Systems

This paper proposes a dynamic resource allocation (DRA) algorithm for WCDMA systems for non-real time data services using a Hopfield neural network (HNN). Previous works have demonstrated the utility of HNN to implement DRA algorithms in the uplink due to the high computational speed of its hardware implementation and the maximization of the total allocated bandwidth. This paper presents an enhanced version of these algorithms by introducing a packet delay control technique, guaranteeing a better QoS and taking into account both uplink and downlink.