【路由優(yōu)化】基于隨機(jī)搜索算法優(yōu)化帶有速度的路由網(wǎng)絡(luò)附matlab代碼
【路由優(yōu)化】基于隨機(jī)搜索算法優(yōu)化帶有速度的路由網(wǎng)絡(luò)附matlab代碼
TT_Matlab
博主簡介：擅長智能優(yōu)化算法、神經(jīng)網(wǎng)絡(luò)預(yù)測、信號處理、元胞自動機(jī)、圖像處理、路徑規(guī)劃、無人機(jī)等多種領(lǐng)域的Matlab仿真，完整matlab代碼或者程序定制加qq1575304183。
1 內(nèi)容介紹
物聯(lián)網(wǎng)實現(xiàn)人和物的連接,感知和采集數(shù)據(jù)的無線傳感器網(wǎng)絡(luò)(Wireless Sensor Networks, WSNs)是其感知事物的核心技術(shù).隨著無線傳感器網(wǎng)絡(luò)的普及和研究的深入,基于確定性環(huán)境的假設(shè)前提,開展資源有限網(wǎng)絡(luò)路由問題的研究難以滿足實際應(yīng)用的需求.無線傳感器網(wǎng)絡(luò)在部署環(huán)境,無線通信,服務(wù)質(zhì)量和網(wǎng)絡(luò)拓?fù)涞确矫嫱瑫r存在眾多不確定性,既有刻畫事件發(fā)生與否的隨機(jī)性,也有刻畫對事件主觀認(rèn)識的模糊性.在不確定網(wǎng)絡(luò)環(huán)境下研究無線傳感器網(wǎng)絡(luò)的路由問題,需要相應(yīng)的不確定性理論和優(yōu)化理論來刻畫路由過程中的各種不確定因素,包括干擾模型,傳播模型,移動模型和服務(wù)質(zhì)量等,從而為不確定環(huán)境下WSN路由算法的研究提供理論基礎(chǔ)和保證. 為了刻畫WSN路由的不確定性,論文基于概率論,模糊集理論,模糊隨機(jī)理論和優(yōu)化理論進(jìn)行路由模型的建模,重點(diǎn)對無線干擾,網(wǎng)絡(luò)模型,服務(wù)質(zhì)量和路由優(yōu)化模型的不確定性進(jìn)行分析和表示,同時設(shè)計相應(yīng)的路由算法和仿真實驗開展不確定環(huán)境下WSN路由算法的研究.
2 仿真代碼
clc;
clear all;
close all;
global Qk ax ay Dik tou beta indA indB
%% Intialization
Nnodes=1;
Emax=1000;
Emin=5;
nk=8;
V=20;
%Z =[1 0.1 0.6 0.8 0.6 0 0.1 1 1 1];
%plotting network topology
%i2=1;
for i2 = 1:noOfNodes
plot(X(i2),Y(i2),’o’,’LineWidth’,1,...
’MarkerEdgeColor’,’k’,...
’MarkerFaceColor’,’y’,...
’MarkerSize’,8’);
xlabel(’X in m’)
ylabel(’Y in m’)
text(X(i2), Y(i2), num2str(i2),’FontSize’,10);
%% Destination
plot(X2,Y2,’^’,’LineWidth’,1,...
’MarkerEdgeColor’,’k’,...
’MarkerFaceColor’,’r’,...
’MarkerSize’,14’);
hold on
end
axis([0 40 -5 40])
M_max = 14; %// number of cells in vertical direction
N_max = 10; %// number of cells in horizontal direction
trans = 1; %// hexagon orientation (0 or 1)
%// Do the plotting:
hold on
C11={};
C={};
ab=1;
ik=1;
for x=0:7%:2;
ik=x;
for y=0:5
if(mod(ik,2))
x0=3+4.3*x;
y0=3+3*2*y;
hexagon(2,x0,y0);
C11{x+1,y+1}=[x0;y0];
% C{ab}=[x0;y0];
hold on
plot(x0,y0,’ok’,’MarkerFaceColor’,’k’)
cote=2;
x1=cote*sqrt(2)*[-1 -0.5 0.5 1 0.5 -0.5 -1]+x0;
y1=cote*sqrt(9)*[0 -0.5 -0.5 0 0.5 0.5 0]+y0;
else
x0=3+4.3*x;
y0=3*2*y;
hexagon(2,x0,y0);
C11{x+1,y+1}=[3+4.3*x;3*2*y];
hold on
plot(3+4.3*x,3*2*y,’ok’, ’MarkerFaceColor’,’k’)
cote=2;
x1=cote*sqrt(2)*[-1 -0.5 0.5 1 0.5 -0.5 -1]+x0;
y1=cote*sqrt(9)*[0 -0.5 -0.5 0 0.5 0.5 0]+y0;
end
C{ab}=[x0,y0];
%% Inside the polygon or not
[in,on] = inpolygon(X,Y,x1,y1);
Nk(ab)=numel(find(in==1));% set of sensor node
ind=[];
ind=find(in==1);
if(isempty(ind))
Dik{ab}=0;
Qk(ab)=0;
else
Dik{ab}=sqrt((X(ind)-x0).^2 +(Y(ind)-y0).^2 ); % distance froom node i to its cell center
Qk(ab)=1;
end
%Tk --> Time stays of WCV
ab=ab+1;
end
end
% axis([0 30 0 30])
% grid
%% Travelling path Model
k=ab-1;
Z=ones(1,k); %% important
aa=cell2mat(C.’);
Xa=aa(:,1);
Ya=aa(:,2);
%% WCV
plot(Xa(1)+2,Ya(1)+2,’o’,’LineWidth’,1,...
’MarkerEdgeColor’,’k’,...
’MarkerFaceColor’,’g’,...
’MarkerSize’,12’);
plot(Xa(1),Ya(1),’o’,’LineWidth’,1,...
’MarkerEdgeColor’,’k’,...
’MarkerFaceColor’,’k’,...
’MarkerSize’,14’);
xlabel(’X in m’)
ylabel(’Y in m’)
hold on
text(Xa(1), Ya(1),’Vechile Stay Room’,’FontSize’,10);
hold on
saveas(gcf,’fileint.fig’,’fig’)
% %% Existing Routing
s=cell2mat(C.’);
ax=s(:,1);
ay=s(:,2);
%% Routing
indA=find(Qk==1);
indB=find(Qk~=1);
G=randperm(numel(indA));
path1 = indA(G);
%% OPTIMIZATION
% %% Problem Definition
CostFunction=@(x) Sphere(x); % Cost Function
ik=1;
%cost1=1000;
eff1=inf;
while(ik<=4000)
T1=1000.*rand(1);
T=CostFunction(T1);
eff=(T);
if(eff<=eff1)
eff1=eff;
TT=T1;
end
costh1(ik)=eff1;
costh(ik)=eff;
ik=ik+1;
end
figure,
plot(1:ik-1,costh1,’-r’)
hold on
plot(1:ik-1,costh,’-b’)
xlabel(’Iteration’)
ylabel(’Objective Function’)
legend(’Optimal’,’Current’)
%%%%%%%%%%%%%%%%%%%%%%%
s=cell2mat(C.’);
ax=s(:,1);
ay=s(:,2);
%% Routing
%rand(’seed’,round(TT))
s = RandStream(’mt19937ar’,’Seed’,round(TT));
G=randperm(s,numel(indA));
% G=randperm(numel(indA));
path1 = indA(G);
if(min(path1)==1)
path1=path1(path1~=1);
path = [1 path1 1];
else
path = [1 path1 1];
end
%saveas(gcf,’fileint.fig’,’fig’)
h3=openfig(’fileint.fig’,’new’,’visible’)
%figure(10),
hold on
for p =1:(length(path)-1)
%line([ax(sor) ax(path(1))],[ay(sor) ay(path(1))],’Color’,’r’,’LineWidth’, 1, ’LineStyle’, ’-’)
line([ax(path(p)) ax(path(p+1))], [ay(path(p)) ay(path(p+1))], ’Color’,’m’,’LineWidth’,2.5, ’LineStyle’,’-’)
arrow([ax(path(p)) ay(path(p)) ], [ax(path(p+1)) ay(path(p+1)) ])
end
title(’Routing’)
%% Mathematical calc
for ab=1:numel(path)-1
dist(ab)=sqrt(((ay(path(ab))-ay(path(ab+1)))^2)+(ax(path(ab))-ax(path(ab+1)))^2);
end
L=sum(dist) % length of path
fprintf(’Length is--->%3.2f ’,L)
NumJunction=numel(path);
% Distance4 bw node and chaerger
for ic=1:numel(Dik)
d(ic)=mean(Dik{ic});
end
% d=cell2mat(Dik); %--- Modified
Prx=@(d)(tou./(d+beta).^2)
Prx(d)
% Tx power of Charger
%% eqn 2
HArE=Prx;
%rand(’seed’,1)
deplrate=0.000001.*rand(1,numel(C)); % Nearest to sink have high depletion rate
alpha=4; % max accel/deaccel
%% single node arbitary
% Temporal Disrtize
delT=2;
fmax=1./delT;
%% Spatial node
Pmax=4;
Pmin=1;
eta1=0.5;
Cq=abs(log(Pmax/Pmin)./(log(1+eta1)))
g=1+(Cq-1).*rand(1,3); % 3 discretize space value
Pq=Pmax.*(1+eta1).^-g;
saveas(gcf,’filePr.fig’,’fig’)
%%%%%%%%%%%%%%%%%%%%
ik=1;
%cost1=1000;
eff=0;
Ui=1000;
while(ik<=500)
[Vel,v] = randfixedsum((NumJunction).*3,1,L,0,10);
dist1=[0 dist];
for ib=1:numel(path)
rx1{ib}=Prx(d(path(ib))).*(dist1(ib)./Vel(((ib-1)*3+1):(ib*3)))-deplrate(path(ib)).*T ;% Temporal
Pqrx2{ib}=(Pq./d(path(ib)));% spatial
end
%end
Ac=cell2mat(Pqrx2);
Ac(isinf(Ac))=0;
Ad=cell2mat(rx1);
eff1=sum(Ac(:)+Ad(:));
costh21(ik)=eff1;
costh2(ik)=eff;
if(eff1>=eff)
eff=eff1;
Velb=Vel;
Ps=Pqrx2;
end
ik=ik+1;
end
figure,
plot(1:ik-1,costh21,’-r’)
hold on
plot(1:ik-1,costh2,’-b’)
xlabel(’Iteration’)
ylabel(’Objective Function’)
legend(’Current’,’Optimal’)
p1=cell2mat(Ps);
p1(isinf(p1))=0;
h2=openfig(’filePr.fig’,’new’,’visible’)
hold on
for p =1:(length(path))
if(p>=2)
text(ax(path(p)),ay(path(p)),num2str([Velb(((p-1)*3+1):(p*3))]),’FontSize’,10);
%plot(ax(path(p)),ay(path(p)),’yo’,’MarkerSize’,20)
end
if(p>=2)
plot(ax(path(p)),ay(path(p)),’yo’,’MarkerSize’,20)
hold on
plot(ax(path(p)),ay(path(p)),’co’,’MarkerSize’,50)
hold on
plot(ax(path(p)),ay(path(p)),’ro’,’MarkerSize’,90)
end
hold on
end
clear path1;
clear Qk C Dik
pause(1)
end
3 運(yùn)行結(jié)果
4 參考文獻(xiàn)
[1]金勁. 群集智能算法在網(wǎng)絡(luò)策略中的研究及其應(yīng)用[D]. 蘭州理工大學(xué).
[2]王永恒, WANG, Yong-heng,等. 基于改進(jìn)蟻群算法的計算機(jī)網(wǎng)絡(luò)路由優(yōu)化研究[J]. 電子設(shè)計工程, 2017(20):4.