failure. There has also been a growing body of literature for
backup provisioning to meet availability guarantees [8–14]. In
all of these, primary and backup flows are allocated such that
the connection is disrupted for at most a specified fraction of
time or probability. During these down-states, the service is
completely disrupted. A version of availability guarantees is
considered in [15], where an end-to-end flow having a certain
expected capacity, based on link availabilities, is found. In our
paper, a flow is guaranteed to be at least a fraction q of the full
demand at all times, which is known as “partial protection”. Our
novel approach is the first to combine the traditional availability
guarantee and partial protection guarantee to allow the user to
specify flows with different availability guarantees. Moreover,
it is particularly applicable to IP-over-WDM networks where
MPLS tunnels are used to provision resources.
failure. There has also been a growing body of literature for
backup provisioning to meet availability guarantees [8–14]. In
all of these, primary and backup flows are allocated such that
the connection is disrupted for at most a specified fraction of
time or probability. During these down-states, the service is
completely disrupted. A version of availability guarantees is
considered in [15], where an end-to-end flow having a certain
expected capacity, based on link availabilities, is found. In our
paper, a flow is guaranteed to be at least a fraction q of the full
demand at all times, which is known as “partial protection”. Our
novel approach is the first to combine the traditional availability
guarantee and partial protection guarantee to allow the user to
specify flows with different availability guarantees. Moreover,
it is particularly applicable to IP-over-WDM networks where
MPLS tunnels are used to provision resources.
การแปล กรุณารอสักครู่..
