White space detection and spectrum characterization in urban and rural India

Broadband penetration in developing countries such as India is still very low, especially in rural areas, because of the prohibitive cost of laying cable and fiber. Recently, wireless solutions based on Wi-Fi have been proposed as a cost-effective solution to bridge the last-mile problem. However, Wi-Fi normally operates in the 2.4 GHz band and above, where signal propagation is not as good as in sub-GHz bands such as terrestrial TV bands. Using TV white spaces for broadband connectivity can be game changing, bringing down the cost of such networks to an affordable level. Few studies exist, however, which characterize sub-GHz spectrum usage in India. In this paper, we make two main contributions. First, we present results of several day-long spectrum measurements in the sub-GHz band made at different urban and rural locations in the National Capital Region (i.e. near New Delhi) of India. Second, we have developed an inexpensive spectrum measurement setup based on a commercially available hand-held spectrum analyzer called RF Explorer. Our main findings are that about 85% of the TV band in the 470–698 MHz range are not used in the heart of urban Delhi, and as much as 95% are unused in rural areas. At any point in time, the largest contiguous TV white space varies between 66 MHz to 136 MHz in an urban location, and between 51 MHz and 242 MHz in a rural location. This suggests the need for development of very wideband, and low-cost wireless solutions which can exploit these valuable white spaces to provide affordable rural connectivity.