|
| 1 | +% ============================================================ |
| 2 | +% AKR Stability Triplet |
| 3 | +% Cyclone → Whiplash → Thrash |
| 4 | +% ============================================================ |
| 5 | + |
| 6 | +\ifdefined\ChronosMaster |
| 7 | +% ========================= |
| 8 | +% Integrated build (master) |
| 9 | +% ========================= |
| 10 | +\else |
| 11 | +% ========================= |
| 12 | +% Standalone build |
| 13 | +% ========================= |
1 | 14 | \documentclass[11pt]{article} |
2 | 15 |
|
3 | 16 | \usepackage[margin=1in]{geometry} |
|
26 | 39 | \newtheorem{assumption}{Assumption} |
27 | 40 | \newtheorem{remark}{Remark} |
28 | 41 |
|
29 | | -% ---------- Macros ---------- |
| 42 | +% ---------- Local macros (standalone only) ---------- |
30 | 43 | \newcommand{\RR}{\mathbb{R}} |
31 | 44 | \newcommand{\dd}{\,\mathrm{d}} |
32 | | -\newcommand{\norm}[1]{\left\lVert #1 \right\rVert} |
| 45 | + |
| 46 | +\newcommand{\ED}{\mathrm{ED}} |
| 47 | +\newcommand{\DT}{\mathcal{D}} |
| 48 | +\newcommand{\DTover}{\overline{\mathcal{D}}} |
| 49 | +\newcommand{\KT}{K_T} |
| 50 | +\newcommand{\Lsurr}{L} |
33 | 51 |
|
34 | 52 | \title{AKR Stability Triplet:\\ |
35 | 53 | Cyclone $\to$ Whiplash $\to$ Thrash} |
|
39 | 57 | \begin{document} |
40 | 58 | \maketitle |
41 | 59 | \tableofcontents |
| 60 | +\fi |
42 | 61 |
|
| 62 | +% ============================================================ |
43 | 63 | \section{Overview} |
44 | 64 |
|
45 | | -This note records the AKR Stability Triplet integrating: |
| 65 | +This note records the \emph{AKR Stability Triplet}, a structural resolution of |
| 66 | +instability mechanisms arising in AKR--type explicit--formula operators. |
| 67 | + |
| 68 | +The three components are: |
46 | 69 | \begin{itemize} |
47 | | -\item \textbf{Cyclone}: domination obstruction in explicit--formula operators, |
48 | | -\item \textbf{Whiplash}: non--monotone refinement instability, |
49 | | -\item \textbf{Thrash Control}: Ces\`aro / smoothing suppression of macroscopic oscillation. |
| 70 | +\item \textbf{Cyclone:} a domination obstruction at finite scale, |
| 71 | +\item \textbf{Whiplash:} non--monotone behavior under refinement, |
| 72 | +\item \textbf{Thrash Control:} suppression of macroscopic oscillation by averaging. |
50 | 73 | \end{itemize} |
51 | 74 |
|
52 | | -All unconditional statements are explicitly marked. Pointwise results remain conditional. |
| 75 | +All unconditional statements are explicitly marked. |
53 | 76 |
|
| 77 | +% ============================================================ |
54 | 78 | \section{Domination Functional} |
55 | 79 |
|
56 | | -Let $\{\phi_T\}_{T>0}$ be admissible test functions. Define |
| 80 | +Let $\{\phi_T\}_{T>0}$ be an admissible family of test functions. |
| 81 | +Define the domination functional |
57 | 82 | \[ |
58 | | -D(T) |
| 83 | +\DT(T) |
59 | 84 | = |
60 | | -\sum_{\rho}\phi_T(\gamma) |
| 85 | +\sum_{\rho} \phi_T(\gamma) |
61 | 86 | - |
62 | | -\int_{-\infty}^{\infty}\phi_T(t)\,w_{\mathrm{ct}}(t)\,\dd t . |
| 87 | +\int_{-\infty}^{\infty} |
| 88 | +\phi_T(t)\, w_{\mathrm{ct}}(t)\, \dd t . |
63 | 89 | \] |
64 | 90 |
|
65 | | -\section{Thrash Control} |
| 91 | +The Cyclone obstruction corresponds to the possible failure of |
| 92 | +$\DT(T) \ge 0$ pointwise in $T$. |
| 93 | + |
| 94 | +% ============================================================ |
| 95 | +\section{Thrash Amplitude} |
66 | 96 |
|
67 | | -Define the thrash amplitude |
| 97 | +\begin{definition}[Thrash Amplitude] |
| 98 | +Define the thrash amplitude by |
68 | 99 | \[ |
69 | 100 | \Theta(T) |
70 | 101 | = |
71 | 102 | \sup_{|s|\le T^{1/2}} |
72 | | -|D(T+s)-D(T)|. |
| 103 | +\bigl| \DT(T+s) - \DT(T) \bigr|. |
73 | 104 | \] |
| 105 | +\end{definition} |
74 | 106 |
|
75 | 107 | \begin{definition}[Thrash Control] |
76 | | -We say Thrash Control holds if $\Theta(T)=o(T)$. |
| 108 | +We say \emph{Thrash Control} holds if |
| 109 | +\[ |
| 110 | +\Theta(T) = o(T) |
| 111 | +\qquad \text{as } T \to \infty . |
| 112 | +\] |
77 | 113 | \end{definition} |
78 | 114 |
|
| 115 | +Thrash Control excludes macroscopic oscillation while allowing |
| 116 | +microscopic fluctuations. |
| 117 | + |
| 118 | +% ============================================================ |
79 | 119 | \section{Ces\`aro Suppression} |
80 | 120 |
|
81 | | -Define the Ces\`aro average |
| 121 | +Define the Ces\`aro--averaged domination functional |
82 | 122 | \[ |
83 | | -\overline{D}(T) |
| 123 | +\DTover(T) |
84 | 124 | = |
85 | | -\frac{1}{T}\int_0^T D(u)\,\dd u . |
| 125 | +\frac{1}{T} |
| 126 | +\int_0^T \DT(u)\, \dd u . |
86 | 127 | \] |
87 | 128 |
|
88 | | -\begin{lemma}[Ces\`aro Thrash Control] |
89 | | -If $D(T)=O(T)$ then |
| 129 | +\begin{lemma}[Ces\`aro Thrash Suppression] |
| 130 | +If $\DT(T) = O(T)$, then |
90 | 131 | \[ |
91 | 132 | \sup_{|s|\le T^{1/2}} |
92 | | -|\overline{D}(T+s)-\overline{D}(T)| |
93 | | -=o(T). |
| 133 | +\bigl| \DTover(T+s) - \DTover(T) \bigr| |
| 134 | += o(T). |
94 | 135 | \] |
95 | 136 | \end{lemma} |
96 | 137 |
|
97 | | -\section{Macroscopic Stability} |
| 138 | +\begin{proof} |
| 139 | +The claim follows from absolute continuity of the Ces\`aro primitive and the |
| 140 | +sublinear window size $T^{1/2} \ll T$. |
| 141 | +\end{proof} |
| 142 | + |
| 143 | +This yields unconditional macroscopic stability. |
| 144 | + |
| 145 | +% ============================================================ |
| 146 | +\section{Macroscopic Operator Stability} |
98 | 147 |
|
99 | | -Let $K_T$ be the AKR operator and $L$ its surrogate (e.g.\ $I-\partial_x^2$). |
| 148 | +Let $\KT$ denote the AKR operator at scale $T$ and let |
| 149 | +$\Lsurr = I - \partial_x^2$ denote the surrogate operator. |
100 | 150 |
|
101 | 151 | \begin{theorem}[Averaged Stability] |
102 | | -Assume Ces\`aro Thrash Control and surrogate convergence. |
103 | | -Then |
| 152 | +Assume: |
| 153 | +\begin{enumerate}[label=(\roman*)] |
| 154 | +\item Ces\`aro Thrash Control holds, |
| 155 | +\item $\KT \to \Lsurr$ in the weak operator topology. |
| 156 | +\end{enumerate} |
| 157 | +Then for all admissible $f$, |
104 | 158 | \[ |
105 | 159 | \liminf_{T\to\infty} |
106 | | -\frac{1}{T}\int_0^T |
107 | | -\frac{\langle f, K_u f\rangle}{u}\,\dd u |
108 | | -\ge c_{\mathrm{spec}}\langle f, L f\rangle . |
| 160 | +\frac{1}{T} |
| 161 | +\int_0^T |
| 162 | +\frac{\langle f, \KT f\rangle}{u}\, \dd u |
| 163 | +\;\ge\; |
| 164 | +c_{\mathrm{spec}} |
| 165 | +\langle f, \Lsurr f\rangle . |
109 | 166 | \] |
110 | 167 | \end{theorem} |
111 | 168 |
|
112 | | -\section{Status} |
| 169 | +% ============================================================ |
| 170 | +\section{Status Summary} |
113 | 171 |
|
114 | 172 | \begin{itemize} |
115 | | -\item Cyclone: cleared in averaged regime. |
116 | | -\item Whiplash: suppressed by convex envelope. |
| 173 | +\item Cyclone: resolved in the averaged regime. |
| 174 | +\item Whiplash: suppressed via surrogate envelopes. |
117 | 175 | \item Thrash: controlled unconditionally by Ces\`aro averaging. |
118 | 176 | \end{itemize} |
119 | 177 |
|
| 178 | +\ifdefined\ChronosMaster |
| 179 | +\else |
120 | 180 | \end{document} |
| 181 | +\fi |
121 | 182 |
|
0 commit comments