Energy Efficient Heterogeneous Network Selection Method Based on Fuzzy Neural Network

Abstract

Heterogeneous network selection algorithm can be expected to play an increasingly important role in the future wireless communication network. Due to the difference in energy consumption of different networks, selecting appropriate access network according to energy consumption characteristics and applications for Mobile Terminal (MT) can effectively extend battery life. In this paper, a fuzzy neural network based energy efficient network selection method is proposed. To select the optimal network and achieve the best balance between performance and energy consumption, the proposed method takes network bandwidth, signal strength, Quality of Service (QoS) and energy consumption into account. A novel energy efficiency utility function is proposed to calculate the energy utility value of different applications based on real-time and non-real-time service. And the calculated energy utility value together with other QoS parameters is input into Fuzzy Neural Network (FNN) for subsequent processing. Finally, the most appropriate network can be selected by comparing the inference results of fuzzy neural network. Computer simulations show that the proposed algorithms can reduce the energy consumption for MT effectively, and has superior performance in terms of access blocking probability and packet dropping probability.