Past exam of the mathematics course of the University of Cambridge 2012 iii Paper 82 2 Solution Created 2026-10-03 Updated 2026-10-14
The printed interval has a typo: its right endpoint must be , rather than . Keep as an unnormalized indicator function; only is normalized to have integral one. Thus useWith the Boolean literal printed interval, the convolution vanishes at , so a sequence consisting entirely of is already a counterexample to the proposed majorization. For the corrected interval and every , all with belong to the support of . Hence there. Everywhere , giving , including interval endpoints.
Put . For , direct integration and the convolution theorem give the exact Fourier coefficientsIn particular,This also supplies the required justification of the Fourier series. The convolution is continuous: it is the overlap length of two moving intervals, divided by . Its Fourier coefficients are absolutely summable, so the corresponding Fourier series converges uniformly to a continuous function . Termwise integration gives ; the permitted uniqueness assertion applied to yields .
Let . Split the absolutely convergent Fourier series at the arbitrary real cutoff :There are fewer than low-frequency terms and each coefficient has magnitude at most . For the tail, and . ThereforeThis proves the requested interval discrepancy estimate without an unjustified sharp truncation of the Fourier series.
For a nonzero integer , let be the nearest integer to . The integer is nonzero, and . The suggested factorization yieldsThis is stronger than the stated lower bound for the fractional part, since the distance to the nearest integer is at most that fractional part. For , the geometric series and giveThus the low-frequency contribution is . Choose and ; these choices satisfy for every . Both and are , and so is . Hence