Executive Summary : | The digital divide in India is evident in rural areas, with only 29.3% internet penetration. The COVID-19 pandemic has exacerbated this gap, particularly in the northeast region where people walk several kilometers to access internet. To improve network coverage in these areas, a significant boost in wireless signal strength is needed. Traditional methods include multiple antenna base stations or repeaters, which are power-consuming and expensive. To address this issue, a proposal combines the benefits of intelligent reflecting surfaces (IRS) and antenna selection. A base station equipped with fewer radio frequency chains than the number of antennas performs antenna selection, while two IRSs, made up of passive reflective elements, help improve coverage. IRS replaces costly radio-frequency chains with low-cost passive reflectors, improving coverage in a power-efficient manner. The design aims at joint operation of the two IRSs, yielding a 16x increase in signal strength by doubling the number of passive elements compared to a 4x increase with a single IRS. This allows a base station located on one side of a hill to reach users on the other side. However, the number of pilot transmissions required for a communication system aided by two IRSs is proportional to the product of the number of antennas at the base station and the number of elements at each IRS. Conventional and machine-learning algorithms are proposed to address this challenge and bring the benefits of antenna selection and two IRSs into practice. This proposal combines the cost and power consumption benefits of the IRS and antenna selection to improve internet speeds in rural and hilly areas and reduce the digital divide. |