Heterogeneous Lifi–Wifi with multipath transmission protocol for effective access point selection and load balancing.

In recent days, the emerging ideas of hybridized network are used, which isbased on light fidelity (LiFi) and wireless fidelity (WiFi). The merging concept of LiFi and WiFirequires huge coverage area and reduce traffic overload. The classical Load Balancing (LB) technique aids to optimize the network throughput. Considering the handoveroverhead issue, this paper devises an effective access point selection (APS) and LB technique. The hybrid Lifi–Wifi is employed and location prediction is executed with Deep Long Short Term Memory (Deep LSTM). After accomplishing the location prediction, APS is carried out using the proposed Sewing Training Inspired Optimization (STIO), where the objective function considersSignal Noise Ratio and Shannon capacity. Further, the joint optimization of LB and handover is performed using STIO, where the user is handover to the selected AP. Finally, the multipath transmission control protocol is employed to transit the data. The proposed STIO-Deep LSTM outperformed with a smaller delay of 0.113 ms, ahigh energy efficiency of 0.097 Mbits/joules, asmall handover occurrence of 6.100, and ahigh network throughput of 0.905. [ABSTRACT FROM AUTHOR]