Class-based bandwidth allocation and admission control for virtual private networks with differentiated service
View Patent ↗A method and apparatus for establishing a virtual private network (VPN) session for a priority request. The method and apparatus includes identifying a priority request from a plurality of VPN session request classes. Reserved bandwidth is allocated to the priority request if a nominally allocated bandwidth is insufficient to satisfy the priority request. Once the allocated bandwidth is assigned, admission control is applied to the priority request.
1. A method for reallocating bandwidth for use in establishing virtual private network (VPN) sessions in a network supporting a plurality of VPN session request classes, comprising:
identifying termination of a VPN session having an allocated bandwidth associated therewith, wherein the identified VPN session is associated with one of the VPN session request classes;
computing, for each of at least one of the VPN session request classes, a respective weighted load factor indicative of whether a current traffic load of the VPN session request class is approaching a load bound parameter associated with the VPN session request class, wherein
the weighted load factor associated with the VPN session request class is computed using the current traffic load of the VPN session request class, the load bound parameter associated with the VPN session request class; and a weight associated with the VPN session request class; and
the current traffic load of the VPN session request class is computed using a measured bandwidth load of the VPN session request class, and an amount of bandwidth allocated for the VPN session request class; and
determining, based on the at least one weighted load factor, whether to reallocate the allocated bandwidth of the terminated VPN session from the one of the VPN session request classes.
2. The method of claim 1 , wherein the at least one of the VPN session request classes for which a respective weighted load factor is computed comprises each of the VPN session request classes identified as being oversubscribed.
3. The method of claim 1 , wherein the at least one of the VPN session request classes for which a respective weighted load factor is computed comprises each of the plurality of VPN session request classes.
4. The method of claim 1 , wherein, for each of the at least one VPN session request class for which a respective weighted load factor is computed, the weight associated with the VPN session request class depends on a priority level of the VPN session request class.
5. The method of claim 1 , wherein, for each of the at least one VPN session request class (i) for which a respective weighted load factor is computed, the weighted load factor (L i ) associated with the VPN session request class is computed as:
L i =( w i )( u i )/(μ i ).
where
w i is the weight;
u i is the current traffic load; and
μ i is the traffic (load) bound.
6. The method of claim 1 , wherein, for each of the at least one VPN session request class for which a respective weighted load factor is computed, the current traffic load (u i ) of the VPN session request class is computed as:
u i =U i /B i .
where: U i is the measured load bandwidth; and
B i is the allocated bandwidth.
7. The method of claim 1 , further comprising:
in response to a determination to reallocate the allocated bandwidth of the terminated VPN session from the one of the VPN session request classes:
selecting at least one of the VPN session request classes based on the at least one weighted load factor computed for the respective at least one of the VPN session request classes; and
reallocating the allocated bandwidth of the terminated VPN session from the one of the VPN session request classes to the selected at least one of the VPN session request classes.
8. The method of claim 7 , wherein the selected at least one of the VPN session request classes comprises the VPN session request class having the largest associated weighted load factor.
9. The method of claim 7 , wherein the selected at least one of the VPN session request classes comprises at least one VPN session request class having a weighted load factor greater than a threshold.
10. An apparatus for reallocating bandwidth for use in establishing virtual private network (VPN) sessions in a network supporting a plurality of VPN session request classes, comprising:
means for identifying termination of a VPN session having an allocated bandwidth associated therewith, wherein the identified VPN session is associated with one of the VPN session request classes;
means for computing, for each of at least one of the VPN session request classes, a respective weighted load factor indicative of whether a current traffic load of the VPN session request class is approaching a load bound parameter associated with the VPN session request class; wherein
the weighted load factor associated with the VPN session request class is computed using the current traffic load of the VPN session request class, the load bound parameter associated with the VPN session request class; and a weight associated with the VPN session request class; and
the current traffic load of the VPN session request class is computed using a measured bandwidth load of the VPN session request class, and an amount of bandwidth allocated for the VPN session request class; and
means for determining, based on the at least one weighted load factor, whether to reallocate the allocated bandwidth of the terminated VPN session from the one of the VPN session request classes.