Paper 1, Section I, G

Coding and Cryptography
Part II, 2012

Let A\mathcal{A} and B\mathcal{B} be alphabets of sizes mm and a respectively. What does it mean to say that c:ABc: \mathcal{A} \rightarrow \mathcal{B}^{*} is a decodable code? State Kraft's inequality.

Suppose that a source emits letters from the alphabet A={1,2,,m}\mathcal{A}=\{1,2, \ldots, m\}, each letter jj occurring with (known) probability pj>0p_{j}>0. Let SS be the codeword-length random variable for a decodable code c:ABc: \mathcal{A} \rightarrow \mathcal{B}^{*}, where B=a|\mathcal{B}|=a. It is desired to find a decodable code that minimizes the expected value of aSa^{S}. Establish the lower bound E(aS)(j=1mpj)2\mathbb{E}\left(a^{S}\right) \geqslant\left(\sum_{j=1}^{m} \sqrt{p_{j}}\right)^{2}, and characterise when equality occurs. [Hint. You may use without proof the Cauchy-Schwarz inequality, that (for positive xi,yix_{i}, y_{i} )

i=1mxiyi(i=1mxi2)1/2(i=1myi2)1/2\sum_{i=1}^{m} x_{i} y_{i} \leqslant\left(\sum_{i=1}^{m} x_{i}^{2}\right)^{1 / 2}\left(\sum_{i=1}^{m} y_{i}^{2}\right)^{1 / 2}

with equality if and only if xi=λyix_{i}=\lambda y_{i} for all i.]