Chapter Summary

Chapter Summary

Key Points

  • 1.
    The broadcast channel models one-to-many communication

    A single transmitter sends independent messages to multiple receivers, each observing a different noisy version of the transmitted signal. The capacity region — the set of simultaneously achievable rate tuples — is known in closed form only for degraded broadcast channels, where X→Y1→Y2X \to Y_1 \to Y_2 forms a Markov chain.

  • 2.
    Superposition coding is the key achievability technique

    The weak user's message is encoded as a coarse "cloud center" and the strong user's message as a fine "satellite." The weak user decodes only the cloud; the strong user decodes both layers via successive decoding. The capacity region is R2≀I(U;Y2)R_{2} \leq I(U; Y_2) and R1≀I(X;Y1∣U)R_{1} \leq I(X; Y_1 | U) over all p(u)p(x∣u)p(u)p(x|u).

  • 3.
    The Gaussian BC has a clean closed-form capacity region

    For the scalar Gaussian BC with noise variances N1<N2N_1 < N_2, the capacity region is parameterized by a power-splitting ratio α∈[0,1]\alpha \in [0,1]: R2≀12log⁑(1+Ξ±P/((1βˆ’Ξ±)P+N2))R_{2} \leq \frac{1}{2}\log(1 + \alpha P / ((1-\alpha)P + N_2)) and R1≀12log⁑(1+(1βˆ’Ξ±)P/N1)R_{1} \leq \frac{1}{2}\log(1 + (1-\alpha)P/N_1).

  • 4.
    Superposition coding strictly outperforms TDMA

    The superposition coding boundary is a convex curve that lies strictly above the TDMA straight line connecting the corner points. The gain is largest when users have very different channel qualities. This is the information-theoretic foundation for NOMA.

  • 5.
    The converse uses the entropy power inequality

    Bergmans' converse shows that Gaussian inputs are optimal for the Gaussian BC. The entropy power inequality β€” N(X+Y)β‰₯N(X)+N(Y)N(X+Y) \geq N(X) + N(Y) with equality iff both are Gaussian β€” is the essential tool that forces Gaussian optimality in the converse argument.

  • 6.
    MAC-BC duality connects uplink and downlink

    The capacity region of the Gaussian BC with total power PP equals the capacity region of the dual MAC with individual powers summing to PP. This duality extends to MIMO and is the computational basis for practical beamforming design. The MIMO BC capacity is achieved by dirty paper coding.

Looking Ahead

Chapter 16 treats the general (non-degraded) broadcast channel, where superposition coding alone is insufficient. Marton's coding scheme combines superposition with binning, and the MAC-BC duality provides the computational framework for the MIMO broadcast channel. Chapter 16 also introduces dirty paper coding β€” the capacity-achieving strategy for the MIMO BC β€” which connects directly to the precoding algorithms in Book telecom, Chapter 17.