多网络中，如果节点v1连接于节点v2，节点v2连接于节点v3，那么节点v3很可能与v1相连接。这种现象体现了部分节点间存在的密集连接性质。可以用聚类系数（CC）来表示，在无向网络中，聚类系数定义为： C = 2*n/(k*(k-1))其中，n表示在节点v的所有k个邻居间的边数。
用Java编写一个求聚类系数的程序，代码注释
Problem Description
A local area network (LAN) supplies networking capability to a group of computers in close proximity to each other such as in an office building, a school, or a home. A LAN is useful for sharing resources like files, printers, games or other applications. A LAN in turn often connects to other LANs, and to the Internet or other WAN.
In this contest, everybody is connecting with each others like the following figure.
The contest’s network was built as N rows and M columns, your computer locate at (0, 0) and the judger’s computer locate at (m-1, n-1) The code you submit would only transfer to up or right smoothly. It doesn’t matter if some accidents happened. Could you tell me how many ways from your computer to judger when a data wire was broken?
Input
There are multiple cases. Every case contains two integers in the first line, N, M (3<=N+M<=40). Second line contains four integers X1, Y1, X2, Y2 (0<=X1, X2<M, 0<=Y1, Y2<N, |X1+Y1-X2-Y2|=1), meaning the broken wire position.
Output
Print the answer in one line.
Sample Input
3 3
0 0 1 0
Sample Output
3
关于matlab实现无线传感器网络DV-HOP算法中如何计算能量损耗
用MATLAB实现无线传感器网络DV-HOP算法，然后根据下列文字编写代码计算能量损耗：
目前，在低能量无线电通信领域有大量的研究。无线电通信特性的不同假设，包括传输和接收模型中的能量消耗，将决定不同协议的优势所在。在我们的工作中，我们假设了一个简单的模型，无线电通信消耗的能量来运行传输或接收电路Eelec=50nJ/ bit，以及作为传输信号增益，Eamp=100pJ/bit/m2以达到一个可接受的Eb/Nb。
这些参数略高于无线电通信设计的目前的技术水平。我们同时假设了因信道传输造成的的能量损耗。因此，使用我们的无线电通信模型，为了传输一个k位的信息到距离为d的地方，无线电通信将消耗：
ETx（k，d）=ETx-elec（k）+ETx-amp（k，d） (1)
ETx（k，d）=Eelec*k+Eamp*k*d2
为了接收该信息，无线电通信将消耗：
ERx（k）=ERx-elec（k） (2)
ERx（k）=Eelec*k
根据这些参数值可以看出，接收一条信息代价很高，因此协议应该力求实现不但使传输距离最小化，而且让每条信息的传输数量和接收操作最小化。
%~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ DV-Hop算法 ~~~~~~~~~~~~~~~~~~~~~~~~
%~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
% BorderLength-----正方形区域的边长，单位：m
% NodeAmount-------网络节点的个数
% BeaconAmount---信标节点数
% Sxy--------------用于存储节点的序号，横坐标，纵坐标的矩阵
%Beacon----------信标节点坐标矩阵;BeaconAmount*BeaconAmount
%UN-------------未知节点坐标矩阵;2*UNAmount
% Distance------未知节点到信标节点距离矩阵;2*BeaconAmount
%h---------------节点间初始跳数矩阵
%X---------------节点估计坐标初始矩阵,X=[x,y]'
% R------------------节点的通信距离，一般为10-100m
clear,close all;
k=2000;
E_elec=50*0.000000001;
E_amp=100*0.000000000001;
E_da=5*0.000000001;
E_0=0.5;%初始能量
E_b=0.1;%波动能量
maxx=40;%节点分布范围
maxy=40;%节点分布范围
BorderLength=100;
NodeAmount=100;
BeaconAmount=9;
UNAmount=NodeAmount-BeaconAmount;
R=50;
% D=zeros(NodeAmount,NodeAmount);%未知节电到信标节点距离初始矩阵；BeaconAmount行NodeAmount列
h=zeros(NodeAmount,NodeAmount);%初始跳数为0；BeaconAmount行NodeAmount列
X=zeros(2,UNAmount);%节点估计坐标初始矩阵
%~~~~~~~~~~~~~~~~~~~~~~~~~~~~~在正方形区域内产生均匀分布的随机拓扑~~~~~~~~~~~~~~~~~~~~
C=-50+BorderLength.*rand(2,NodeAmount);%生成随机坐标
%带逻辑号的节点坐标
Sxy=[[1:NodeAmount];C];
Beacon=[Sxy(2,1:BeaconAmount);Sxy(3,1:BeaconAmount)];%信标节点坐标
UN=[Sxy(2,(BeaconAmount+1):NodeAmount);Sxy(3,(BeaconAmount+1):NodeAmount)];%未知节点坐标
%画出节点分布图
plot(Sxy(2,1:BeaconAmount),Sxy(3,1:BeaconAmount),'r*',Sxy(2,(BeaconAmount+1):NodeAmount),Sxy(3,(BeaconAmount+1):NodeAmount),'k.')
xlim([5-0,BorderLength/2]);
ylim([-50,BorderLength/2]);
title('* 红色信标节点 . 黑色未知节点')
%~~~~~~~~~~~~~~~~~~~~~~~~~~~~初始化节点间距离、跳数矩阵~~~~~~~~~~~~~~~~~~~~~~
for i=1:NodeAmount
for j=1:NodeAmount
Dall(i,j)=((Sxy(2,i)-Sxy(2,j))^2+(Sxy(3,i)-Sxy(3,j))^2)^0.5;%所有节点间相互距离
if (Dall(i,j)<=R)&(Dall(i,j)>0)
h(i,j)=1;%初始跳数矩阵
elseif i==j
h(i,j)=0;
else h(i,j)=inf;
end
end
end
%~~~~~~~~~~~~~~~~~~~~~~~~~最短路经算法计算节点间跳数~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
for k=1:NodeAmount
for i=1:NodeAmount
for j=1:NodeAmount
if h(i,k)+h(k,j)<h(i,j)%min(h(i,j),h(i,k)+h(k,j))
h(i,j)=h(i,k)+h(k,j);
end
end
end
end
%~~~~~~~~~~~~~~~~~~~~~~~~~求每个信标节点的校正值~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
h1=h(1:BeaconAmount,1:BeaconAmount); %信标节点的跳数因为在现实中我们只知道信标节点之间的距离
D1=Dall(1:BeaconAmount,1:BeaconAmount);%距离，信标节点的跳数因为在现实中我们只知道信标节点之间的距离
for i=1:BeaconAmount
dhop(i,1)=sum(D1(i,:))/sum(h1(i,:));%每个信标节点的平均每跳距离
end
%dhop=(sum(D1,2)./sum(h1,2));每行相加求和，去掉for循环效不率比较高
D2=Dall(1:BeaconAmount,(BeaconAmount+1):NodeAmount);%BeaconAmount行UNAmount列
for i=1:BeaconAmount
for j=1:UNAmount
if min(D2(:,j))==D2(i,j)
Dhop(1,j)=dhop(i,1);%未知节点从最近的信标获得校正值
end
end
end
%Dhop = min(D2,[],1);求每一列的最小值，也就是说每个未知点到信标的最短距离
Dhop
%~~~~~~~~~~~~~~~~~~~~~~~~~~~用跳数估计距离~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
hop1=h(1:BeaconAmount,(BeaconAmount+1):NodeAmount)%未知节点到信标跳数，BeaconAmount行UNAmount列
for i=1:UNAmount
hop=Dhop(1,i);%hop为从最近信标获得的校正值
Distance(:,i)=hop*hop1(:,i);%%Beacon行UN列；
end
%Distance2 = repmat(Dhop,8,1).*hop1;
% %~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~最小二乘法求未知点坐标~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
d=Distance;
for i=1:2
for j=1:(BeaconAmount-1)
a(i,j)=Beacon(i,j)-Beacon(i,BeaconAmount);
end
end
A=-2*(a');
% d=d1';
%polyfit();最小二乘法的matlab实现
%最小二乘法也就是最小平方法
for m=1:UNAmount
for i=1:(BeaconAmount-1)%根据八个信标点计算出最优位置
B(i,1)=d(i,m)^2-d(BeaconAmount,m)^2-Beacon(1,i)^2+Beacon(1,BeaconAmount)^2-Beacon(2,i)^2+Beacon(2,BeaconAmount)^2;
end
X1=inv(A'*A)*A'*B;%pseudo inverse
X(1,m)=X1(1,1);
X(2,m)=X1(2,1);
end
%pinv()求pseudo inverse.
% for i=1:UNAmount
% error(1,i)=(((X(1,i)-UN(1,i))^2+(X(2,i)-UN(2,i))^2)^0.5);
% end
for UNAmount=1:1:91
load topo
BeaconAmoun=9;
node=rand(2,BeaconAmoun)
x=node(1,:)*maxx;
y=node(2,:)*maxy;
E_0n=E_0+E_b*node(1,:);
R_nc=60;
R_cs=110;
E_men=k*E_elec+E_amp*k*R_nc^2;
R=0;
BeaconAmoun=BeaconAmoun+UNAmount;
E_ch=k*E_elec*(BeaconAmoun-1)+k*E_elec+E_amp*k*R_cs^2;
a=zeros(1,BeaconAmoun);
a(floor(rand*1)+1,floor(rand*BeaconAmoun)+1)=1;
for m=1:1:BeaconAmoun
if(a(1,m)==1)
s=m;
end
end
E_cho=E_0n(1,s);
while(E_cho>E_ch)
R=R+1;
R_sj(1,BeaconAmoun)=R
E_0n(1,s)= E_0n(1,s)-E_ch+E_men;
for j=1:1:BeaconAmoun
E_0n(1,j)= E_0n(1,j)-E_men;
end
E_mins=E_0n(1,1);
for l=1:1:BeaconAmoun
if(E_mins<E_0n(1,l))
E_mins=E_0n(1,l);
end
end
if (E_mins>E_men)
a=zeros(1,BeaconAmoun);
a(floor(rand*1)+1,floor(rand*BeaconAmoun)+1)=1;
for m=1:1:num1
if(a(1,m)==1)
s=m;
E_cho=E_0n(1,s);
end
end
if(E_mins<E_men)
E_cho=0;
end
end
end
end
