As data rates continue to rise, a network failure can cause
catastrophic service disruptions. To protect against such failures,
networks typically use full protection schemes, often
doubling the cost of resources needed to route a connection. An
alternative approach is to provide a guarantee on the maximum
time a connection can be disrupted. This is known as an
“availability guarantee”, and it is a bound on the fraction of
time or probability that a connection can be disrupted. However,
these disruptions (downtimes) may be unacceptably long; thus,
many service providers opt for the more resource intensive
full protection. In this paper, we propose a novel protection
scheme with multiple availability guarantees. In addition to the
traditional availability guaranteed protection, which maintains
the full demand for at least a guaranteed fraction of time, we
guarantee partial connectivity at all times. Thus, our approach
is a hybrid between the traditional availability guarantees and
full protection schemes