Path loss estimation in wireless networks using partial derivative based convex optimisation method

In this paper, we develop a multi-dimensional mathematical optimisation problem for predicting the optimised path loss in general wireless systems under non-linear network operating constraints. The proposed nonconvex optimisation problem is reformulated through convex relaxation programming in order to guarantee global optimal solution. The underlying convexity of the transformed reformulated problem is derived through the application of second-order partial derivatives. Numerical simulation experiments are conducted to compute the propagation path loss and optimal values of system parameters using free space and two-ray ground models. The graphical comparison between the two propagation models and the previously existing wireless channel models is illustrated in terms of various channel performance metrics like optimal path loss, signal-to-noise ratio (SNR), bit error rate (BER), and transmit power consumption. Finally, simulation results are provided to demonstrate the effectiveness of our optimisation model in achieving significantly lower path loss than the previous models based on conventional machine learning systems.