Abstract
Long-Term Evolution (LTE) network is an improved standard for mobile telecommunication system developed by the 3rd Generation Partnership Project (3GPP) requires an efficient handover framework which would reduce hysteresis and improve quality of service (QoS) of subscribers by maximizing scarce radio resources. This paper compares the performance of two ANN prediction algorithms (LevenbergMarquadt and Bayesian regularization) based on received signal strength (RSS) and the hysteresis margin parameters for neuro-adaptive hysteresis margin reduction algorithm. The Bayesian regularization algorithm had a lower mean error when compared with the Levenberg-Marquadt (LM) prediction algorithm and as such a better option for neuro-adaptive hysteresis margin reduction algorithm.
Highlights
IntroductionIn Mobile communications systems, the user equipment (UE) regularly moves through several base stations (BS)
In Mobile communications systems, the user equipment (UE) regularly moves through several base stations (BS).As a UE enters the region served by a new BS, the call connection is reassigned from the BS previously serving that UE to the new BS [1]
This paper compares the performance of two Artificial Neural Network (ANN) prediction algorithms (LevenbergMarquadt and Bayesian regularization) based on received signal strength (RSS) and the hysteresis margin parameters for neuro-adaptive hysteresis margin reduction algorithm
Summary
In Mobile communications systems, the UE regularly moves through several base stations (BS). The research proposes that the processes involved in the handover (preparation, execution and completion) tend to take time as it can lead to a breakdown and loss of data in the search for and selection of a target eNodeB This analysis led to development of a prediction model that predicts the direction of signals and angles of the eNodeB using the MLP neural network. The utilization of ANN in vertical handoff decisions have proved very useful and efficient and the adoption of Levenberg Marquardt algorithm have made the entire training period of the neural network model faster. The regression plot in figure 7 shows a good correlation between the inputs (RSSS and RSST) and the target (hysteresis) using the Bayesian regularization training algorithm. These R values signify a close relationship between the input and target values
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
