📂Number Theory

Square-and-Multiply Algorithm Proof

Algorithm

Given a natural number $a,k,m$, $b \equiv a^{k} \pmod{m}$ can be computed as follows.

---

Step 1. Binary Expansion of $k$

Represent $u_{i} = 0$ or $u_{i} = 1$ as follows. $$ k = \sum_{i=0}^{r} u_{i} 2^{i} = u_{0} + 2 u_{1} + \cdots + 2^r u_{r} $$

---

Step 2.

Calculate $a^{2^{r}} \equiv ( a^{2^{r-1}} )^2 \equiv A_{r-1}^2 \equiv A_{r} \pmod{m}$ as follows: $$ \begin{align*} a & & & \equiv A_{0} \pmod{m} \\ a^{2} & \equiv ( a^1 )^2 & \equiv A_{0}^2 & \equiv A_{1} \pmod{m} \\ a^{4} & \equiv ( a^2 )^2 & \equiv A_{1}^2 & \equiv A_{2} \pmod{m} \\ a^{8} & \equiv ( a^4 )^2 & \equiv A_{2}^2 & \equiv A_{3} \pmod{m} \\ & & & \vdots \end{align*} $$

---

Step 3.

Calculating $A_{0}^{u_{0}} \cdot A_{1}^{u_{1}} \cdot \cdots A_{r}^{u_{r}}$ yields: $$ a^{k} \equiv A_{0}^{u_{0}} \cdot A_{1}^{u_{1}} \cdot \cdots A_{r}^{u_{r}} \pmod{m} $$

Explanation

Although the advancement of computers has made it possible to compute powers quickly, humans are never satisfied. The method of repeated squaring is a way to get the answer without directly computing immensely large numbers in modular operations. By dividing by $\pmod{m}$ after each square, the number falls below $m$, thus reducing the amount of computation.

Example

For instance, let’s calculate $7^{327} \pmod{853}$. Taking the common logarithm of $7^{327}$ gives $327 \log 7 \approx 327 \cdot 0.8450 = 276.315$, which is a large number with a digit count of $277$. It’s too big to multiply honestly and then divide, so let’s use the above algorithm. If you do the binary expansion of $327$ as described in the algorithm, it looks like this: $$ 327 = 256 + 64 + 4 +2 +1 = 2^{8} + 2^{6} + 2^2 + 2^1 + 2^0 $$ If you calculate according to Step 2, $$ \begin{align*} 7 & & \equiv 7 & \equiv 7 \pmod{853} \\ 7^{2} & \equiv ( 7^1 )^2 & \equiv 7^2 & \equiv 49 \pmod{853} \\ 7^{4} & \equiv ( 7^2 )^2 & \equiv 49^2 & \equiv 2401 \equiv 695 \pmod{853} \\ 7^{8} & \equiv ( 7^4 )^2 & \equiv 695^2 & \equiv 227 \pmod{853} \\ 7^{16} & \equiv ( 7^8 )^2 & \equiv 227^2 & \equiv 349 \pmod{853} \\ 7^{32} & \equiv ( 7^{16} )^2 & \equiv 349^2 & \equiv 675 \pmod{853} \\ 7^{64} & \equiv ( 7^{32} )^2 & \equiv 675^2 & \equiv 123 \pmod{853} \\ 7^{128} & \equiv ( 7^{64} )^2 & \equiv 123^2 & \equiv 628 \pmod{853} \\ 7^{256} & \equiv ( 7^{128} )^2 & \equiv 628^2 & \equiv 298 \pmod{853} \end{align*} $$ Thus, $$ 7^{327} = 7^{256} \cdot 7^{64} \cdot 7^{4} \cdot 7^{2} \cdot 7^{1} = 298 \cdot 123 \cdot 695 \cdot 49 \cdot 7 \equiv 286 \pmod{853} $$ This method might still feel hard, but the amount of computation is quite reasonable compared to multiplying $327$ times individually.

Proof

$$ \begin{align*} a^{k} =& a^{u_{0} + 2u_{1} + \cdots + 2^r u_{r} } \\ =& a^{u_{0}} a^{ 2u_{1} } \cdots a^{ 2^r u_{r} } \\ \equiv & A_{0}^{u_{0}} A_{1}^{u_{1}} \cdots A_{r}^{u_{r}} \pmod{m} \end{align*} $$

■

Code

Below is an example code for calculating $7^{327} \pmod{853}$ using the repeated squaring method, written in R language. Note that the power option accommodates negative integers, and if mod is a prime number, entering power=-1 allows you to find the multiplicative inverse of the given base.

FPM<-function(base,power,mod) #It is equal to (base^power)%%mod
{
  i<-0
  if (power<0) {
    while((base*i)%%mod != 1) {i=i+1}
    base<-i
    power<-(-power)}
  
  if (power==0) {return(1)}
  if (power==1) {return(base%%mod)}
  
  n<-0
  while(power>=2^n) {n=n+1}
  
  A<-rep(1,n)
  A[1]=base
  
  for(i in 1:(n-1)) {A[i+1]=(A[i]^2)%%mod}
  
  for(i in n:1) {
    if(power>=2^(i-1)) {power=power-2^(i-1)}
    else {A[i]=1} }
  
  for(i in 2:n) {A[1]=(A[1]*A[i])%%mod}
  
  return(A[1])
}
 
FPM(7,327,853)
FPM(7,-1,11)