Projects – 1-2-1 Gaussian mmWave networks

What are fundamental limits for arbitrary Gaussian MillimeterWave networks, and can we achieve or approximate them using efficient algorithms?

Millimeter-Wave (mmWave) communications are expected to play a vital role in 5G mobile communications expanding the available spectrum and enabling multi-gigabit services. To explore the information-theoretic capacity for mmWave networks,  we proposed a new deterministic model, which we call “1-2-1” network model, that we proved is within a constant gap from the Gaussian network capacity. 

We used this model to compute the approximate capacity of arbitrary half-duplex and full-duplex 1-2-1 networks; we showed that we can achieve the approximate capacity by simply routing information across a polynomial (in the number of network nodes) number of paths; and we derived polynomial-time algorithms that enable to find the optimal network schedule, i.e., which nodes should communicate at what fraction of time.