Interference-limited versus noise-limited communication over dense wireless networks

A network of n wireless communication links is considered. Rayleigh fading is assumed to be the dominant factor affecting the strength of the channels between nodes. In previous works it is shown that the maximum throughput of this network over all link activation strategies scales as log n. However, it is achieved by assigning a vanishingly small rate to each active link. The objective of this paper is to analyze the achievable throughput of the network when the data rate of each active link is constrained to be a constant λ > 0. A link activation strategy is proposed and analyzed using random graph theory. In the interference-limited regime, a throughput scaling as τ log n is achievable, where the scaling factor r approaches 1 as λ → 0 or λ - ∞. This implies the asymptotic optimality of the proposed scheme. In the noise-limited regime, it is shown that rate-perlinks scaling as log(Δ_0ρ) are achievable, where Δ₀ is a constant and ρ is the transmit signal to noise ratio. However, in this case the throughput decreases by a factor of log log n as compared to the interference-limited regime.