1: \begin{abstract}

2: Consider~\(n\) nodes~\(\{X_i\}_{1 \leq i \leq n}\)

3: independently distributed in the unit square~\(S,\)

4: each according to a distribution~\(f.\) Nodes~\(X_i\) and~\(X_j\)

5: are joined by an edge if the Euclidean distance~\(d(X_i,X_j)\) is less than~\(r_n,\)

6: the adjacency distance and the resulting random graph~\(G_n\)

7: is called a random geometric graph~(RGG). We now

8: assign a location dependent weight to each edge of~\(G_n\)

9: and define~\(MST_n\) to be the sum of the weights of the minimum spanning trees of

10: all components of~\(G_n.\) For values of~\(r_n\) above the connectivity regime,

11: we obtain upper and lower bound deviation estimates for~\(MST_n\)

12: and~\(L^2-\)convergence of~\(MST_n\) appropriately

13: scaled and centred. %converges to zero in mean as~\(n \rightarrow \infty.\)

14: 

15: 

16: %\(\frac{1}{\sqrt{n}}(MST_n - \mathbb{E}MST_n)\)

17: 

18: 

19: 

20: 

21: 

22:  %and analogous results hold for minimum spanning trees and paths.

23: 

24: 

25: %In this paper, we study the structure of left-right crossings

26: %of the random geometric graph \(G = G(n,r_n)\) of \(n\) nodes

27: %uniformly distributed in \(S = [0,1]^2\) with \(r_n = \epsilon\sqrt{\frac{\log{n}}{n}}\)

28: %for some \(\epsilon > 0.\) Tiling \(S\) horizontally and

29: %vertically into rectangles of length \(1\) and width \(Mr_n,\) we

30: %show that each rectangle has a left-right crossing of edges with

31: %high probability if \(M\) is sufficiently large.

32: %We call the resulting subgraph to be a ``backbone" of \(G.\)

33: 

34: %The techniques we use to construct the backbone has quite a few applications.

35: %As a first, we show that the diameter of second largest component in \(G\)

36: %is \(O(1)\) with high probability. Secondly,

37: %\vspace{0.1in} \noindent \textbf{Key words:} .

38: 

39: \vspace{0.1in} \noindent \textbf{AMS 2000 Subject Classification:} Primary:

40: 60J10, 60K35; Secondary: 60C05, 62E10, 90B15, 91D30.

41: \end{abstract}