Abstract:
We investigate optimal call admission control (CAC)
policy for multicast streaming services (MSS) in 3rd generation
(3G) and beyond wireless mobile networks. Several MSS sessions
are supported simultaneously in a bandwidth-limited network.
Active sessions are those that are currently serving some users,
and inactive sessions are those that are currently not serving any
users. An admission decision in MSS is required only when an
inactive session is requested, unlike in unicasting. For this reason,
if a user request for an inactive MSS session arrives, we should
make an admission decision in anticipation of (i) the possible
reward earned based on users served during a session active time
generated by accepting it, and (ii) the influence of the session active
time upon the future status of network bandwidth and admission
decisions. Our objective is to determine when to admit or block a
user asking an inactive MSS session to achieve the optimality in
rewards. We formulate this problem as a semi-Markov decision
process (SMDP), and a value iteration algorithm is used to obtain
an optimal stationary deterministic policy. We also derive the
user blocking probability of the optimal policy by analyzing an
embedded Markov chain induced by it.
