Chapter Summary

Chapter 17 Summary

Key Points

• 1.

  The interference channel models independent transmitter-receiver pairs sharing a common medium, where each transmission interferes with the others. Unlike the MAC and BC, the capacity region is unknown in general — even for the two-user Gaussian case.

• 2.

  The capacity is known in two extreme regimes: under very strong interference, both users achieve interference-free rates (decode and subtract); under strong interference ($a^2 \geq 1$), the capacity region equals the intersection of two MAC regions (joint decoding at each receiver).

• 3.

  The Han-Kobayashi (HK) scheme uses rate splitting — each user transmits a common part (decoded by both receivers) and a private part (treated as noise). The ETW power split achieves rates within 1 bit of capacity for all parameter regimes, making HK a universal near-optimal strategy.

• 4.

  At high SNR, the $K$-user IC has $K/2$ total degrees of freedom, achieved by interference alignment (IA) — a precoding technique that confines all interference to half the signal space at each receiver. IA requires global perfect CSI and large symbol extensions, severely limiting its practical applicability.

• 5.

  In the noisy interference regime ($\text{INR} \leq 1 + \text{SNR}$), the simplest strategy — treating interference as noise (TIN) — is sum-rate optimal to within 0.5 bits. This is the regime in which most practical cellular systems operate, validating the engineering approach of managing interference through power control, frequency planning, and MIMO spatial processing.

• 6.

  The GDoF characterization reveals a W-shaped curve: DoF = 1 at no interference, drops through the TIN-optimal regime, reaches a minimum at $\alpha = 1/2$, and recovers to DoF = 1 in the strong interference regime.