This paper proposes a novel algorithm for dynamic physical resource allocation based on the use of positioning information during carrier aggregation (CA) in a semi-and uncoordinated deployment of heterogeneous networks (HetNet). The algorithm uses the known Q-learning method enhanced with information but the locations of the mobile users and neighboring cells to solve the problem of dynamic physical resource assignment in uncoordinated scenario while accounting for improved allocation and scheduling. The results are compared to the performance when known scheduling algorithms are employed and show increased cell throughput, while maintaining an adequate user throughput when employing Q-learning with positioning information.
Wireless Communications, Vehicular Technology, Information Theory and Aerospace & Electronic Systems (vitae), 2013 3rd International Conference on, 2013, p. 1-6
2013 3rd International Conference on Wireless Communications, Vehicular Technology, Information Theory and Aerospace & Electronic Systems (VITAE), 2013