There continues to be much debate over the allocation of the upper 6GHz band. Some countries have allocated it to cellular (e.g. China), others to unlicensed devices which are typically dominated by Wi-Fi (e.g. US), and some have adopted interim measures awaiting further insights (e.g. UK). The debate continues because both technologies have a case. In this paper we consider the benefits of allowing both technologies to share spectrum and set out approaches that would allow this to happen.
Our premise is that cellular will only deploy 6GHz in areas of network congestion, which tend to be in denser urban areas. Outside these areas, Wi-Fi would be free to use the whole band. Within these areas, there are differences of use – cellular tends to be deployed on outdoor base stations whereas Wi-Fi is deployed on indoor access points (APs). There are also differences of need. Residential areas likely have greater need for Wi-Fi whereas shopping and tourist areas have greater need for cellular. There are likely differences of timing – in business districts there may be greater need for Wi-Fi during the morning and afternoon when workers are in the office, but greater need for cellular during commute and lunch times when they are outdoors. An ideal approach would allow the 6GHz spectrum to flow between cellular and Wi-Fi dynamically according to time and location, on a minute-by-minute and block-by-block basis.
This paper proposes an approach whereby spectrum is shared by cellular and Wi-Fi through use of a centralized database. This database determines the fraction of shared spectrum that is available for cellular and the fraction available for Wi-Fi, and that fraction can vary from location to location, year to year, and hour to hour.
This paper uses a simulation model to examine how this form of dynamic sharing can affect performance for both cellular and Wi-Fi, infrastructure required, and ultimately economic value. We compare spectrum management strategies in which each spectrum band is statically allocated for either cellular or unlicensed, strategies where allocations vary over time, and strategies where allocations vary over space. The results show that dynamic sharing can vastly outperform static spectrum allocations.