Grant: Discovery Grants Program - Individual

The adoption of spectrum sharing in massive MIMO based 5G networks has the potential to drastically enhance the efficiency of 5G networks. Indeed, this technology allows the entire available spectrum to be available at every location, thereby increasing network capacity. Moreover, spectrum sharing allows new network operators to use massive MIMO based 5G spectrum in regions that are not served by existing network operators. Last but not least, incumbent (primary) spectrum users could continue using their spectrum in a specific location, while making it available for others at other locations. This research program will focus on designing spectrum sharing and leasing for broadband multiuser massive MIMO networks. To this end, two main themes will be pursued. The first theme deals with learning the environment where massive MIMO networks are to share wideband spectrum. Such a learning process should be devised under the restriction that only minimal (preferably no) coordination between the networks exists. Such a restriction will ensure that the networks can cooperate with minimal interference to each other. The second theme in this research program deals with the design and development of efficient resource allocation schemes. Multiuser Massive MIMO networks that are to co-exist in the same RF spectrum, shall not require too much coordination among them. One aspect of this research program will focus on developing learning procedures through which each network will learn about the other networks blindly or semi-blindly as much as possible. In doing so, the challenges involved include using minimal information exchange between the networks to estimate the inter-network channel state information. Another aspect of this research program will rely on the channel state information obtained as mentioned above to devise power-optimal spectrally-efficient communication schemes that are robust against uncertainty in the estimated channel state information. To do so, we rely on worst-case performance optimization approaches with a variety of performance measures such as minimum mean squared error, sum-rate, signal-to-noise ratio, signal-to-interference-plus noise ratio, etc. The above two aspects will be investigated under different setups. Such different setups include single-antenna users versus multi-antenna users, single-carrier schemes versus multi-carrier schemes. Hybrid schemes, where a mix of single- and multi-antenna users are to be served, will also of interest. Majority of the studies conducted on spectrum sharing for massive MIMO networks, assume single carrier communication scheme. In practice, however, 5G networks and beyond will have to operate in a time-dispersive (frequency-selective) environment. As such, developing spectrum sharing schemes for multicarrier massive MIMO networks is a must and this is exactly what is missing from the literature on spectrum sharing schemes for massive MIMO networks.