Energy efficient routing and secured data transmission in the IoV: Improved deep learning model for energy prediction

In the Internet of Vehicles (IoV), vehicle data exchange for cooperative assessment can enhance the experience of driving along with service quality. Nevertheless, concerns with security, bandwidth, and privacy prevent service providers from actively participating in the knowledge exchange procedure. This paper proposes energy-efficient routing and secured data transmission through a deep learning model. In the initial stage, this work adopts the Bidirectional Long Short Term Memory- Modified Attention Layer (Bi-LSTM-MAL) approach for energy prediction of each node. The optimal cluster head selection is the crucial step, for which this work proposes the Combined Beluga with Black Widow Optimization (CBBWO) algorithm by considering distinct constraints like energy, risk, distance and delay. This ensures precise routing, which is optimally done by taking account of constraints including trust, mobility and link quality. Also, this work concerns secure data transmission by protecting the data with a security standard, Fusion Key induced in Elliptic Curve Cryptography (FK-ECC) that aids the model to transmit the data securely between source and destination. Finally, the work evaluates its performance in optimal routing and secured data transmission over conventional models. The CBBWO recorded the minimized cost value of 1.4850 at the final (25th) iteration while existing methods such as BWO, BWOA, BMO, MFO, SBO, ROAONC and SDO-BM obtained maximum cost rates respectively.