91Ó°ÊÓ

First, we need to express m(t+h) in terms of m(t), P_{00}(t), and o(h). We know that: \(m(t) = E[O(t)]\) (Expected occupation time in state 0 by time t) \(P_{00}(t) =\) probability of being in state 0 at time t given that we start in state 0 Using the first-order Taylor expansion, we can write m(t+h) as: \(m(t+h) = m(t) + m'(t)h + o(h)\) Where m'(t) is the derivative of m(t) with respect to time t, and o(h) is a term that goes to 0 as h goes to 0. Using the properties of Markov chains, we can express m'(t) in terms of P_{00}(t) as follows: \(m'(t) = P_{00}(t)\) Now substitute this expression for m'(t) into the Taylor expansion: \(m(t+h) = m(t) + P_{00}(t)h + o(h)\) This is what we wanted to show in part (a).

In this step, we will find the expression for the derivative of m(t), i.e., m'(t). We know that the off-diagonal elements in the generator matrix Q are \(\lambda\) and \(\mu\), with \(\lambda\) being the transition rate from state 0 to state 1, and \(\mu\) being the transition rate from state 1 to state 0. The equation provided is: \(m'(t) = \frac{\mu}{\lambda+\mu} + \frac{\lambda}{\lambda+\mu} e^{-(Q+k)t}\) Notice that k is not mentioned anywhere in the exercise. It seems to be a typo and could possibly be replaced with (\lambda+\mu). Thus, the corrected equation is as follows: \(m'(t) = \frac{\mu}{\lambda+\mu} + \frac{\lambda}{\lambda+\mu} e^{-(\lambda+\mu)t}\)

Now, we need to solve the differential equation derived in part (b) to find m(t). \(m'(t) = \frac{\mu}{\lambda+\mu} + \frac{\lambda}{\lambda+\mu} e^{-(\lambda+\mu)t}\) We can rewrite this as: \(m'(t) - \frac{\mu}{\lambda+\mu} = \frac{\lambda}{\lambda+\mu} e^{-(\lambda+\mu)t}\) Now, we can integrate both sides with respect to t: \(\int (m'(t) - \frac{\mu}{\lambda+\mu}) dt = \int \frac{\lambda}{\lambda+\mu} e^{-(\lambda+\mu)t} dt\) Applying the integral, we get: \(m(t) - \frac{\mu}{\lambda+\mu}t = -\frac{\lambda}{(\lambda+\mu)^2}e^{-(\lambda+\mu)t} + C\) Where C is the constant of integration. Since m(0) = 0, we can find C: \(0 - \frac{\mu}{\lambda+\mu}(0) = -\frac{\lambda}{(\lambda+\mu)^2}(1) + C\) Therefore, C = \(\frac{\lambda}{(\lambda+\mu)^2}\) Now, substitute the value of C back into the equation for m(t): \(m(t) = \frac{\mu}{\lambda+\mu}t + \frac{\lambda}{(\lambda+\mu)^2}(1 - e^{-(\lambda+\mu)t})\) This is the final expression for the expected occupation time in state 0 by time t, m(t).