diff --git a/quantization/Readme.md b/quantization/Readme.md
new file mode 100644
index 0000000..5112fbc
--- /dev/null
+++ b/quantization/Readme.md
@@ -0,0 +1,13 @@
+Quantization
+
+It is a process which makes the input into discrete and maps to some limited discrete values
+The method goes like this
+Function Y = linear_quantation(x,B,a)
+    x - input
+    B - number of bits used to quantization levels
+    a - poistive real number [-a,+a) forms the range of qunatization
+
+let's have a example and it is linear quantization
+For the range [0,a): To implement this above function, divide the interval [0,1] into L = 2^(B-1)
+0 = r0 < r1 < r3 .... < rL = 1 Then we compare with [a*ri, a*r(i+1) ) input at a point then 
+If exists then we take midpoint of above interval.
\ No newline at end of file
diff --git a/quantization/linear_quantation.m b/quantization/linear_quantation.m
new file mode 100644
index 0000000..69b711c
--- /dev/null
+++ b/quantization/linear_quantation.m
@@ -0,0 +1,24 @@
+function Y = linear_quantation(x,B,a) 
+
+len = length(x) ;  %% calucates the length of input
+Y = zeros(1,len);  %% output of quantized input-intialized here
+r = linspace(0,a,B+1) ;  %% creates equally spaced intervals of (B+1) in [0,a) 
+
+for v = 1:B             %% it is for checking for each level
+  for h = 1:len         %% it is for checking every input
+      if ( x(h)<((-a)*r(v)) && x(h)>=((-a)*r(v+1)) ) %% cheks the condition
+          Y(h) = ((-a)*r(v)+(-a)*r(v+1))/2 ;         %% takes the mean
+      end
+  end
+    
+    for h = 1:len        %% it is for checking every input
+      if ( x(h)>=(a*r(v)) && x(h)<(a*r(v+1)) )       %% cheks the condition
+          Y(h) = (a*r(v)+a*r(v+1))/2 ;      %% takes the mean
+      end
+    end
+  
+end
+
+ 
+  
+end
\ No newline at end of file