Dimensioning granular multi-timescale fairness
A boost is provided in an overloaded system by distinguishing nodes with a “bad” traffic history from nodes with a “good” traffic history. In so doing, a core network node is able to apply additional resources to the node(s) having a “good” history in the form of a boost factor. Based on a system capacity and a working point, e.g., a critical number of active nodes with a “bad” traffic history, the core network node may determine a throughput history limit belonging to the “bad” traffic history. Responsive to expected requirements for a newly active node (i.e., a node having a “good” traffic history), the core network node determines a boost factor for the newly active node, applies the boost factor to the average resources allocated to the nodes with the “bad” traffic history to determine boosted resources, and allocates the boosted resources to the newly active node.
1. A method of adjusting resource allocation to provide multi-timescale fairness for a communication network at a bottleneck defining shared resource availability for a plurality of first nodes and a second node, the method implemented by a core network node and comprising:
determining a boost factor responsive to an average communication resource and a planned data size for download by the second node during an initial boost time of a second time period within a first time period, wherein the average communication resource is required by each of the first nodes over the first time period and the second node intermittently requires communication resources during the first time period;
applying the boost factor to the average communication resource to determine a boosted communication resource; and
allocating the boosted communication resource to the second node to configure the second node to download data using the boosted communication resource during the initial boost time of the second time period.
2. The method of claim 1 , wherein the average communication resource comprises a bandwidth.
3. The method of claim 1 , further comprising allocating the average communication resource to the second node to configure the second node to download data using the average communication resource during a remainder of the second time period after expiration of the initial boost time.
4. The method of claim 1 , further comprising:
determining a second boost factor for a second boost time of the second time period after expiration of the initial boost time responsive to the average communication resource and a second planned data size for download by the second node during the second boost time;
applying the second boost factor to the average communication resource to determine a second boosted communication resource; and
allocating the second boosted communication resource to the second node to configure the second node to download data using the second boosted communication resource during the second boost time of the second time period.
5. The method of claim 1 , wherein a total load of the second node relative to a common link communication resource summed with a product of the average communication resource and the number of the plurality of first nodes is less than the common link communication resource.
6. The method of claim 1 , further comprising allocating a communication resource to each of the plurality of first nodes and allocating the boosted communication resource to the second node according to a Multi-TimeScale Throughput Value Function (MTS-TVF) comprising a TVF for each of two or more different timescales.
7. The method of claim 6 , wherein each of the TVFs of the MTS-TVF further comprise a different TVF for each of one or more service classes.
8. The method of claim 6 , wherein the allocating the communication resource to each of the plurality of first nodes comprises allocating the communication resource to each of the plurality of first nodes responsive to one or more quality of experience requirements.
9. The method of claim 1 , wherein the core network node comprises a gateway node.
10. A core network node configured to adjust resource allocation for a plurality of first nodes and provide multi-timescale fairness for a communication network at a bottleneck defining shared resource availability for the plurality of first nodes and a second node, the core network node comprising:
processing circuitry; and
memory storing instructions executable by the processing circuitry whereby the core network node is configured to:
determine a boost factor responsive to an average communication resource and a planned data size for download by the second node during an initial boost time of a second time period within a first time period, wherein the average communication resource is required by each of the first nodes over the first time period and the second node intermittently requires communication resources during the first time period; and
apply the boost factor to the average communication resource to determine a boosted communication resource; and
allocate the boosted communication resource to the second node to configure the second node to download data using the boosted communication resource during the initial boost time of the second time period.
11. The core network node of claim 10 , wherein the average communication resource comprises a bandwidth.
12. The core network node of claim 10 , wherein the core network node is further configured to allocate the average communication resource to the second node to configure the second node to download data using the average communication resource during a remainder of the second time period after expiration of the initial boost time.
13. The core network node of claim 10 , wherein the core network node is further configured to:
determine a second boost factor for a second boost time of the second time period after expiration of the initial boost time responsive to the average communication resource and a second planned data size for download by the second node during the second boost time; and
apply the second boost factor to the average communication resource to determine a second boosted communication resource; and
allocate the second boosted communication resource to the second node to configure the second node to download data using the second boosted communication resource during the second boost time of the second time period.
14. The core network node of claim 10 , wherein a total load of the second node relative to a common link communication resource summed with a product of the average communication resource and the number of the plurality of first nodes is less than the common link communication resource.
15. The core network node of claim 10 , wherein the core network node is further configured to allocate a communication resource to each of the plurality of first nodes and allocating the boosted communication resource to the second node according to a Multi-TimeScale Throughput Value Function (MTS-TVF) comprising a TVF for each of two or more different timescales.
16. The core network node of claim 15 , wherein each of the TVFs of the MTS-TVF further comprise a different TVF for each of one or more service classes.
17. The core network node of claim 15 , wherein the core network node allocates the communication resource to each of the plurality of first nodes by allocating the communication resource to each of the plurality of first nodes responsive to one or more quality of experience requirements.
18. The core network node of claim 10 , wherein the core network node comprises a gateway node.
19. The core network node of claim 10 , wherein the plurality of first nodes and the second node each comprise different communication devices in the communication network.
20. A non-transitory computer readable medium storing a computer program product for controlling a core network node to provide multi-timescale fairness for a communication network at a bottleneck defining shared resource availability for a plurality of first nodes and a second node, the computer program product comprising software instructions that, when run on the core network node, cause the core network node to:
determine a boost factor responsive to an average communication resource and a planned data size for download by the second node during an initial boost time of a second time period within a first time period, wherein the average communication resource is required by each of the first nodes over the first time period and the second node intermittently requires communication resources during the first time period;
apply the boost factor to the average communication resource to determine a boosted communication resource; and
allocate the boosted communication resource to the second node to configure the second node to download data using the boosted communication resource during the initial boost time of the second time period.