Here you go. (With an explicit picture one can give an explicit answer.)
\documentclass[tikz,border=3mm]{standalone}
\usetikzlibrary{knots}
\begin{document}
\begin{tikzpicture}[fat line/.style={black, double=#1,double
distance=6pt,looseness=1.2,line cap=round}]
\begin{knot}[%draft mode = crossings, % uncomment to see where the crossings are
clip width = 0,
flip crossing/.list={1,3,5,7,9}]
\path foreach \X in {0,4.5} {foreach \Y in {0.2,-0.2}
{(1.6*\X,\Y) node[draw,fill=orange,inner ysep=1.5pt,inner xsep=8pt,rounded
corners=1.5pt]{}}};
\strand[fat line=white]
plot[domain=0:4.5,samples=251] (1.6*\x,{0.2*cos(\x*360)});
\strand[fat line=green!60!black]
plot[domain=0:4.5,samples=251] (1.6*\x,{-0.2*cos(\x*360)});
\end{knot}
\end{tikzpicture}
\end{document}

Here is a more complicated way with nicer line joins. (I am not implying that it is impossible to have these nice line joins with knots, I am just not versed enough to fix this within the time I have now.)
\documentclass[tikz,border=3mm]{standalone}
\begin{document}
\begin{tikzpicture}[fat line/.style={black, double=#1,double
distance=6pt,looseness=1.2,line cap=round}]
\path foreach \X in {0,4.5} {foreach \Y in {0.2,-0.2}
{(1.6*\X,\Y) node[draw,fill=orange,inner ysep=1.3pt,inner xsep=8pt,rounded
corners=1.5pt]{}}};
\draw[fat line=white]
plot[domain=0:4.5,samples=101,smooth] (1.6*\x,{0.2*cos(\x*360)});
\draw[fat line=green!60!black]
plot[domain=0:4.5,samples=101,smooth] (1.6*\x,{-0.2*cos(\x*360)});
\draw[fat line=white,line cap=butt]
foreach \X in {1,...,4}
{plot[domain=\X-0.4:\X-0.1,samples=7,smooth] (1.6*\x,{0.2*cos(\x*360)})};
\draw[white,line width=6pt]
foreach \X in {1,...,4}
{plot[domain=\X-0.5:\X,samples=11,smooth] (1.6*\x,{0.2*cos(\x*360)})};
\end{tikzpicture}
\end{document}

EDIT: Corrected the line width of the last plot, big thanks to @Symbol 1.
line width=6without-.4because howdouble distanceworks. So even the small tinny tick can be eliminated. – Symbol 1 Mar 17 '20 at 05:57