# Thread: Minimizing probability function based on constraint

1. ## Minimizing probability function based on constraint

Minimize the probability:

$P_b=\left(2^{R-B_1log(1+\frac{P_1}{N_1})}-1\right)\left(\frac{N_2B_2}{P_2}\right)$

with respect to the constraint $P=P_1+P_2$

I am trying to solve this using lagrange multipliers.
Even though i got the $\lambda$ value i am not able to find the $P_1$ and $P_2$ values .

Can anyone please help me to get the $P_1$ and $P_2$ values .

Thanks

2. Originally Posted by bjkrishna
Minimize the probability:

$P_b=\left(2^{R-B_1log(1+\frac{P_1}{N_1})}-1\right)\left(\frac{N_2B_2}{P_2}\right)$

with respect to the constraint $P=P_1+P_2$

I am trying to solve this using lagrange multipliers.
Even though i got the $\lambda$ value i am not able to find the $P_1$ and $P_2$ values .

Can anyone please help me to get the $P_1$ and $P_2$ values .

Thanks
Why are you using Lagrange multiplers? If you are required to do so then fine, but other wise just substitute $P_2=P-P_1$ into the expression from $P_b$ and minimise as usual subject only to $0 \le P_1\le 1$

Also clarify if the log is natural or common (or base 2).

CB

3. Thanks Bruno

the logarithm is respect to base 2

It is not compulsory to use Lagrange multipliers.
Can you please explain the minimizing procedure in detail.

4. Originally Posted by bjkrishna
Thanks Bruno

the logarithm is respect to base 2

It is not compulsory to use Lagrange multipliers.
Can you please explain the minimizing procedure in detail.
You will need to check the algebra but I have:

$P_b=2^k \left[1+\frac{P_1}{N_1}\right]^{-B} \frac{N_2}{P-P_1}$

Now differentiate this with respect to $P_1$ and set the derivative to zero and solve the resulting equation for $P_1$.

Now compare $P_b$ for $P_1=0,1$ and the solution (if any in $[0,1]$) found above the minimum for $P_b$ is the smallest of these.

CB

5. Thanks a lot for the clue
I am able to get the value of $P_1.$