Number of Solutions of a Functional Equation

Algebra Level 4

What can be said about the number of continuous real valued functions f ( x ) , f(x), defined on the set of real numbers, that are solutions of following functional equation? f ( x + y ) + 2 f ( x y ) = 2 f ( x ) f ( y ) . f(x+y)+2f(x-y)=2f(x)f(y).

A finite number of solutions greater than 4. Two solutions. No solution. Four solutions. Three solutions. Infinite number of solutions.

This section requires Javascript.
You are seeing this because something didn't load right. We suggest you, (a) try refreshing the page, (b) enabling javascript if it is disabled on your browser and, finally, (c) loading the non-javascript version of this page . We're sorry about the hassle.

3 solutions

Arturo Presa
May 17, 2019

If we consider the original functional equation together with the equation that results by interchanging x x and y y into that equation, we get f ( x + y ) + 2 f ( x y ) = 2 f ( x ) f ( y ) f(x+y)+2f(x-y)=2f(x)f(y) f ( y + x ) + 2 f ( y x ) = 2 f ( y ) f ( x ) f(y+x)+2f(y-x)=2f(y)f(x) Subtracting the second from the first, we get that f ( x y ) f ( y x ) = 0 , f(x-y)-f(y-x)=0, and making y = 0 , y=0, the equation f ( x ) = f ( x ) . f(-x)=f(x). So any solution of the given original functional equation is an even function.

Making x = t 2 x=\frac{t}{2} and y = t 2 y=\frac{t}{2} into the given equation again, we get f ( t ) + 2 f ( 0 ) = 2 f ( t 2 ) f ( t 2 ) ( 1 ) f(t)+2f(0)=2f(\frac{t}{2})f(\frac{t}{2})\quad\quad\quad(1) Making x = t 2 x=\frac{t}{2} and y = t 2 , y=-\frac{t}{2}, we get f ( 0 ) + 2 f ( t ) = 2 f ( t 2 ) f ( t 2 ) = 2 f ( t 2 ) f ( t 2 ) ( 2 ) f(0)+2f(t)=2f(\frac{t}{2})f(-\frac{t}{2})=2f(\frac{t}{2})f(\frac{t}{2})\quad\quad\quad(2)

Therefore, f ( t ) + 2 f ( 0 ) = f ( 0 ) + 2 f ( t ) , f(t)+2f(0)=f(0)+2f(t), that implies that f ( t ) = f ( 0 ) . f(t)=f(0). So, the only possible solutions of the given functional equation are constant function of the form f ( x ) = f ( 0 ) f(x)=f(0) . Making the substitution x = y = 0 x=y=0 into the given functional equation, it results that 3 f ( 0 ) = 2 f ( 0 ) 2 . 3f(0)=2f(0)^2. Therefore, the only possible values of f ( 0 ) f(0) are 0 or 3 2 . \frac{3}{2}. Then the only possible continuous solutions of the given functional equation are f ( x ) = 0 f(x)=0 for all real values of x x and the function f ( x ) = 3 2 . f(x)=\frac{3}{2}. for all real values of x . x. Then the number of possible solutions is 2 . \boxed{2}.

Mark Hennings
May 21, 2019

We do not need continuity here. Putting y = 0 y=0 gives us that 3 f ( x ) = 2 f ( x ) f ( 0 ) 3f(x) = 2f(x)f(0) for all x x , so either f ( x ) = 0 f(x) = 0 for all real x x or else f ( 0 ) = 3 2 f(0) = \tfrac32 .

Suppose that f ( 0 ) = 3 2 f(0) = \tfrac32 . Putting x = 0 x=0 gives us that f ( y ) + 2 f ( y ) = 3 f ( y ) f(y) + 2f(-y) = 3f(y) for all real y y , so that f ( y ) = f ( y ) f(-y) = f(y) for all real y y . Putting x = y x=y gives f ( 2 x ) + 3 = 2 f ( x ) 2 f(2x) + 3 \; = \; 2f(x)^2 for all real x x . Putting y = x y=-x gives 3 2 + 2 f ( 2 x ) = 2 f ( x ) f ( x ) = 2 f ( x ) 2 \tfrac32 + 2f(2x) \; = \; 2f(x)f(-x) \; = \; 2f(x)^2 for all real x x . Hence f ( 2 x ) + 3 = 3 2 + 2 f ( 2 x ) f(2x) + 3 \; = \; \tfrac32 + 2f(2x) for all real x x , so that f ( 2 x ) = 3 2 f(2x) = \tfrac32 for all real x x .

Thus, without assuming that f f is continuous, we deduce that either f 0 f \equiv 0 or f 3 2 f \equiv \tfrac32 , and so there are exactly 2 \boxed{2} solutions.

Tom Engelsman
May 18, 2019

Let's examine this functional equation for two cases:

CASE 1 ( f ( x ) f(x) is constant): This case yields C + 2 C = 2 C 2 0 = C ( 2 C 3 ) C = 0 , 3 2 C + 2C = 2C^2 \Rightarrow 0 = C(2C-3) \Rightarrow C = 0, \frac{3}{2}

So f ( x ) = 0 , 3 2 \boxed{f(x) = 0, \frac{3}{2}} are acceptable solutions.

CASE 2 ( f ( x ) f(x) is non-constant): Assuming the continuous function f : R R f: \mathbb{R} \rightarrow \mathbb{R} is diffentiable over all real numbers, let us first differentiate the above functional equation with respect to x x and y y :

f ( x + y ) + 2 f ( x y ) = 2 f ( x ) f ( y ) f'(x+y) + 2f'(x-y) = 2f'(x)f(y) (i) AND f ( x + y ) 2 f ( x y ) = 2 f ( x ) f ( y ) f'(x+y) - 2f'(x-y) = 2f(x)f'(y) (ii)

and adding (ii) to (i) gives: 2 f ( x + y ) = 2 [ f ( x ) f ( y ) + f ( x ) f ( y ) ] 2f'(x+y) = 2[f'(x)f(y) + f(x)f'(y)] (iii). Now, we differentiate (iii) same as before to obtain:

2 f ( x + y ) = 2 [ f ( x ) f ( y ) + f ( x ) f ( y ) ] 2f''(x+y) = 2[f''(x)f(y) + f'(x)f'(y)] (iv) AND 2 f ( x + y ) = 2 [ f ( x ) f ( y ) + f ( x ) f ( y ) ] 2f''(x+y) = 2[f'(x)f'(y) + f(x)f''(y)] (v)

which ultimately yields the differential equation: f ( x ) f ( y ) = f ( x ) f ( y ) f ( x ) = K f ( x ) f ( x ) = a e K x + b e K x f''(x)f(y) = f(x)f''(y) \Rightarrow f''(x) = K \cdot f(x) \Rightarrow \boxed{f(x) = ae^{\sqrt{K}x} + be^{-\sqrt{K}x}} (vi) (where a , b , K R a,b,K \in \mathbb{R} ).

Substitution of (vi) back into our original functional equation produces:

a e K ( x + y ) + b e K ( x + y ) + 2 [ a e K ( x y ) + b e K ( x y ) ] = 2 [ a e K x + b e K x ] [ a e K y + b e K y ] ae^{\sqrt{K}(x+y)} + be^{-\sqrt{K}(x+y)} + 2[ae^{\sqrt{K}(x-y)} + be^{-\sqrt{K}(x-y)}] = 2[ae^{\sqrt{K}x} + be^{-\sqrt{K}x}][ae^{\sqrt{K}y} + be^{-\sqrt{K}y}] ;

or a e K ( x + y ) + b e K ( x + y ) + 2 [ a e K ( x y ) + b e K ( x y ) ] = 2 [ a 2 e K ( x + y ) + a b e K ( x y ) + a b e K ( x y ) + b 2 e K ( x + y ) ] ae^{\sqrt{K}(x+y)} + be^{-\sqrt{K}(x+y)} + 2[ae^{\sqrt{K}(x-y)} + be^{-\sqrt{K}(x-y)}] = 2[a^{2}e^{\sqrt{K}(x+y)} + abe^{-\sqrt{K}(x-y)} + abe^{\sqrt{K}(x-y)} + b^{2}e^{-\sqrt{K}(x+y)}] ; (vii)

By matching the coefficients in each exponential term, we obtain:

a = 2 a 2 a = 0 , 1 2 ; a = 2a^2 \Rightarrow a = 0, \frac{1}{2}; ,

b = 2 b 2 b = 0 , 1 2 b = 2b^2 \Rightarrow b = 0, \frac{1}{2} ,

2 a = 2 a b b = 1 2a = 2ab \Rightarrow b = 1 ,

2 b = 2 a b a = 1 2b = 2ab \Rightarrow a = 1

The only solution that satisfies all four of these coefficient conditions is a = b = 0 a = b = 0 , thus no non-constant function f ( x ) f(x) satisfies the original functional equation above.

Conclusion: f ( x ) = 0 , 3 2 \boxed{f(x) = 0, \frac{3}{2}} are the only solutions.

This is a good - but partial - solution. We can't assume f f is differentiable (let alone twice differentiable); we're only told it's continuous (although, as pointed out elsewhere, we don't even need to use that in this case). I wonder if it's possible to prove f f is twice differentiable from the functional equation? (Without just solving the functional equation, of course!)

Chris Lewis - 2 years ago

Log in to reply

I think the method used in this article can be used in this case also to proved that any continuous solution has to be also infinitely differentiable.

Arturo Presa - 2 years ago

Log in to reply

Thanks for the article, Arturo!

tom engelsman - 2 years ago

Thanks for the feedback, Chris. This was a quick -n- dirty solution I sloughed through late that night, and it's my preferred approach to most functional equation problems (aside from the Plug -n- Chug -n- Match method for values of x and y).

tom engelsman - 2 years ago

Log in to reply

I wish I knew a more standardised approach for these - I think your solution is very useful as it gives a constructive method (albeit under the condition of differentiability). I've definitely come across cases where a "nice" looking functional equation has a solution that's continuous but not differentiable, though. Perhaps I'll post one as a problem if I'm feeling particularly mean...

Chris Lewis - 2 years ago

Log in to reply

Look forward to solving such a problem once it's posted....alert me when ready!

tom engelsman - 2 years ago

0 pending reports

×

Problem Loading...

Note Loading...

Set Loading...