Bounds for the loss probability in large loss queueing systems

Abstract: Let

m a t h c a l G (f r a k g_{1}, f r a k g_{2})

be the class of all probability distribution functions of positive random variables having the given first two moments

f r a k g_{1}

and

f r a k g_{2}

. Let

G_{1} (x)

and

G_{2} (x)

be two probability distribution functions of this class satisfying the condition

| G_{1} (x) - G_{2} (x) | < e p s i l o n

for some small positive value

e p s i l o n

and let

w i d e h a t G_{1} (s)

and, respectively,

w i d e h a t G_{2} (s)

denote their Laplace-Stieltjes transforms. For real

m u

satisfying

m u f r a k g_{1} > 1

let us denote by

g a m m a_{G_{1}}

and

g a m m a_{G_{2}}

the least positive roots of the equations

z = w i d e h a t G_{1} (m u - m u z)

and

z = w i d e h a t G_{2} (m u - m u z)

respectively. In the paper, the upper bound for

| g a m m a_{G_{1}} - g a m m a_{G_{2}} |

is derived. This upper bound is then used to find lower and upper bounds for the loss probabilities in different large loss queueing systems.

This page was built for publication: Bounds for the loss probability in large loss queueing systems