why are there noise signals with increasing time?

The following is a program for simulation of a laser beam with dither modulation.

clear all
  clc
  freq=1000000;        %samples per second
  tmax=1;                %time
  Kac=0.1555;           %slope constant
  c=3*10^8;                %light velocity
  lam=632.8*10^-9;        %wavelength of beam
  wd=2*pi*400;            %dither angular frequency 
  M=250;                   %constant
  k=2*pi/lam;             
  wc=2*pi*c/lam;
  n=tmax*freq;    
  I1=zeros([n 1]);
  L=.28;                     %optical path length in meters
  Vhfo=6.5;                  %constant
  dell=0;
  omega=pi/240;                %rotation rate
  for ii=2:n,
  t(ii)=(ii-1)/freq;                                         
  L=.2800000784+(1e-7*sin(2*pi*250*t(ii)));  %optical path length with dither modulation                
  dell=(4*((L/4)^2)/c)*omega;  %change in lengths l1 and l2                              
  l1(ii)=L+dell;                                              
  l2(ii)=L-dell;
  q=round(L/lam);
  v1=q*(c/l1(ii));
  w1=2*pi*v1;
  v2=q*(c/l2(ii));
  w2=2*pi*v2;    
  I1(ii)=cos((w2-w1)*t(ii)+M*sin(2*pi*wd*t(ii)));      %modulated  output 
end
plot(I1);

I'm modulating l1 and l2 and from l1 and l2 the modulation is being passed to E01 and E02 and from there on to I1 . In addition to modulation, with increasing time there are spikes being built up on the signal in I1 which is not supposed to happen. The same thing when simulated in simulink doesnt have spikes. I'm unable to find the mistake. Any help anyone!! Thanks!!!

[Image Analyst]

Well it should be explained right there in the comments. Oh wait, there aren't any!

[Jan]

Please, Sampath, explain what you are doing and ask a specific question.

[Sampath reddy]

I have tried to explain the question above.

I think this is some computational error, but i'm unable to point the cause!!

[Jan]

@Sampath reddy: Your program does not contain any comment, but several unexplained magic numbers like 0.1555, 632.8, 250, 0.28, 6.5, .2800000784. It is not mentioned which algorithm your are using. Some computations are strange, e.g. "var=((1/l2)-(1/l1));" - but this variable is never used.

Finally you do not mention, why you expect, that there are no spikes. Perhaps the theory does expect spikes. We cannot check this easily, but you should be an expert, because you have implemented this code.

So please post the relevant details.

[Sampath reddy]

Now the comments are given. According to the equations there should be no spikes. But there are spikes. can you help now?