Run time acceleration for adaptive algorithm for OFDM bit-loading for OWC channels

Effective power and bit loading strategies are indispensable for mitigating throughput losses attributed to the low-pass frequency response of the light emitting diodes (LED)s and photodiodes (PD)s in optical wireless communication (OWC) systems. To enable the widespread adoption of a high-speed wireless technology using LEDs, the development of precise, computation-friendly algorithms is required. Although expressions for the power allocation in a continuous channel exist, these show to not be the most appropriate in real scenarios as the OWC chipsets are digital devices with finite precision. Moreover, response of the transimpedance amplifiers (TIA)s shows to distort the power spectrum density of the noise, a factor often assumed to be flat in existing expressions. In this research, we delve into the adaptation and acceleration of an iterative algorithm in the past used in radio frequency (RF) systems to optimize performance of LED-based OWC systems. The strategy shows to be efficient, less computationally demanding then commonly used optimization solvers and it intrinsically consider the finite precision of the chipsets.