MathLabs
TheoremProved

Hardy–Littlewood maximal theorem

Statement

MM is of weak type (1,1)(1,1): ∣{x:Mf(x)>λ}∣≤Cλ∥f∥1|\{x : Mf(x) > \lambda\}| \le \frac{C}{\lambda}\|f\|_1 for a universal constant CC (one can always take C=5C=5). Consequently, by interpolation, MM is also bounded on Lp(R)L^p(\mathbb{R}) for every 1<p≤∞1<p\le\infty — but MM itself is never bounded on L1(R)L^1(\mathbb{R}).

Why is it true?

Mf(x)Mf(x) bounds, in one stroke, every average of ff that could ever be taken over an interval around xx — the running mean at every possible scale. Controlling that single quantity turns out to be exactly what is needed to prove the Lebesgue differentiation theorem (the average of ff over shrinking intervals around xx converges to f(x)f(x) for almost every xx): once MfMf is known to be finite almost everywhere, a short soft argument upgrades that to the full differentiation statement. This is the real-variable engine behind the Calderón–Zygmund theory below, playing the role that Plancherel's identity played for the easy L2L^2 estimate.

Proof sketch

Step 1 (a cover by good balls). Fix λ>0\lambda>0. For every xx with Mf(x)>λMf(x)>\lambda, by definition of the supremum there is some interval BxB_x centered at xx with 1∣Bx∣∫Bx∣f∣>λ\frac{1}{|B_x|}\int_{B_x}|f| > \lambda. These balls {Bx}\{B_x\} cover the set {Mf>λ}\{Mf>\lambda\}.

Step 2 (Vitali 5r5r-covering lemma). From any collection of balls of bounded radius, one can always extract a countable disjoint subcollection {Bi}\{B_i\} such that the 55-times dilated balls {5Bi}\{5B_i\} still cover the union of the whole original collection. Apply this to {Bx}\{B_x\} to get a disjoint subfamily {Bi}\{B_i\} with {Mf>λ}⊆⋃i5Bi\{Mf>\lambda\}\subseteq\bigcup_i 5B_i.

Step 3 (sum the disjoint pieces). Each BiB_i satisfies λ∣Bi∣<∫Bi∣f∣\lambda|B_i| < \int_{B_i}|f| by construction. Since the BiB_i are pairwise disjoint, summing over ii gives λ∑i∣Bi∣<∑i∫Bi∣f∣≤∫R∣f∣=∥f∥1\lambda\sum_i|B_i| < \sum_i\int_{B_i}|f| \le \int_{\mathbb{R}}|f| = \|f\|_1, so ∑i∣Bi∣<∥f∥1/λ\sum_i|B_i| < \|f\|_1/\lambda. Finally ∣{Mf>λ}∣≤∑i∣5Bi∣=5∑i∣Bi∣<5λ∥f∥1|\{Mf>\lambda\}| \le \sum_i|5B_i| = 5\sum_i|B_i| < \dfrac{5}{\lambda}\|f\|_1, which is exactly ∣{x:Mf(x)>λ}∣≤5λ∥f∥1|\{x : Mf(x) > \lambda\}| \le \frac{5}{\lambda}\|f\|_1.

Topics that use this theorem

Step-by-step proofs

No step-by-step proof yet for this theorem.

References

  1. Terence Tao (2003). Recent progress on the restriction conjecture · arXiv:math/0311181
  2. Elias M. Stein, Guido Weiss (1971). Introduction to Fourier Analysis on Euclidean Spaces
  3. Loukas Grafakos (2014). Classical Fourier Analysis · DOI:10.1007/978-1-4939-1194-3