Fuzzy-Neural based Cost Effective Handover Prediction Technique for 5G-IoT networks
Rashad. T.S.1, A. Ch. Sudhir2

1Rashad T.S., Scientist ‘F’ RCI, DRDO, Hyderabad (Telangana), India.

2A. CH. Sudhir, Assistant Professor, Department of Electronics and Communication Engineering, GITAM Deemed to be University, (Telangana), India.

Manuscript received on 23 November 2019 | Revised Manuscript received on 11 December 2019 | Manuscript Published on 30 December 2019 | PP: 191-197 | Volume-9 Issue-2S3 December 2019 | Retrieval Number: B10481292S319/2019©BEIESP | DOI: 10.35940/ijitee.B1048.1292S319

Open Access | Editorial and Publishing Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open-access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)

Abstract: Most of the existing works related to handover prediction in 5G networks, depends on huge mobility patterns collected over several periods of time, which will be tedious and complex to classify and analyze these patterns to predict the future locations of mobile users. Hence the main objective is to design a HO prediction technique which accurately predicts the next cell location with least amount of mobility history or patterns. In this paper, we design handoff prediction and target network selection scheme for 5G-IoT networks. For VHO triggering condition, Multi-layer Feed Forward Network (MFNN) is applied which will predict the user mobility based on distance, RSS, mobile speed and direction parameters. For target cell selection, Fuzzy decision model is applied based on the network level metrics such as traffic load, handover latency, battery power and user level metrics such as security and cost. The proposed approach will be implemented in NS3 and the performance is measured in terms of network throughput, handoff delay, handoff cost and prediction accuracy.

Keywords: 5G; IoT; Fuzzy; Cost; Prediction.
Scope of the Article: Regression and Prediction