User assignment strategies for hierarchical and other overlay networks
View Patent ↗A hierarchical assignment manager (HAM) improves system throughput in a hierarchical overlay network by taking into account the requested data rates in addition to or instead of the velocities of mobile users within the network. The user population is characterized by the statistical description of their user profiles. The entire set of profiles is partitioned along assignment curves into two or more subsets, where each subset corresponds to the assignment to a particular layer of a hierarchical network. The HAM determines assignment curves to optimize the total effective system throughput in a systematic way and can be adapted to time-varying user behavior and profiles. The HAM can be used to decide upon assignment strategies between different layers of an overlay architecture, between different available network technologies, or between competing service providers.
1. A computer-implemented method for assigning users in an overlay network having two or more layers, comprising:
receiving a request from a user for assignment within the network, wherein the request identifies (1) a desired data transmission parameter for the user's assignment and (2) velocity of the user, wherein the velocity is independent of the desired data transmission parameter; and
selecting a layer of the network for the user's assignment based on both the desired data transmission parameter and the velocity, wherein the network layer for the user's assignment is selected by comparing the desired data transmission parameter and the velocity to an assignment curve that is dependent on both data transmission parameter and velocity.
2. The invention of claim 1 , wherein the overlay network is a wireless, hierarchical overlay network comprising at least first and second layers, each layer comprising one or more cells.
3. The invention of claim 1 , wherein:
the user is a real-time-data user;
the desired data transmission parameter is a desired data rate; and
the network layer for the user's assignment is selected based on the desired data rate and the velocity.
4. The invention of claim 1 , wherein the assignment curve is a piecewise linear curve connecting two or more points in a two-dimensional space having data transmission parameter and velocity as its two dimensions.
5. The invention of claim 1 , wherein:
the user is a non-real-time-data user;
the desired data transmission parameter is a desired amount of data associated with the user's assignment; and
the network layer for the user's assignment is selected based on the desired amount of data and the velocity.
6. The invention of claim 1 , further comprising modeling the network to determine criteria used in selecting the network layer for the user's assignment.
7. The invention of claim 6 , wherein the modeling comprises using a Markov chain.
8. The invention of claim 6 , wherein the modeling is used to estimate system throughput as a function of two or more assignment curves, wherein an assignment curve that yields the highest system throughput is used to select the network layer for the user's assignment.
9. The invention of claim 1 , wherein:
the desired data transmission parameter is one of desired data rate and desired amount of data; and
the greater the value of the desired data transmission parameter, the higher the probability that the method will select a lower layer in the overlay network for the user's assignment.
10. An assignment manager for assigning users in an overlay network having two or more layers, wherein the assignment manager:
receives a request from a user for assignment within the network, wherein the request identifies (1) a desired data transmission parameter for the user's assignment and (2) velocity of the user, wherein the velocity is independent of the desired data transmission parameter; and
selects a layer of the network for the user's assignment based on both the desired data transmission parameter and the velocity, wherein the network layer for the user's assignment is determined by comparing the desired data transmission parameter and the velocity to an assignment curve that is dependent on both data transmission parameter and velocity.
11. The invention of claim 10 , wherein the overlay network is a wireless, hierarchical overlay network comprising at least first and second layers, each layer comprising one or more cells.
12. The invention of claim 10 , wherein:
the user is a real-time-data user;
the desired data transmission parameter is a desired data rate; and
the network layer for the user's assignment is based on the desired data rate and the velocity.
13. The invention of claim 10 , wherein the assignment curve is a piecewise linear curve connecting two or more points in a two-dimensional space having data transmission parameter and velocity as its two dimensions.
14. The invention of claim 10 , wherein:
the user is a non-real-time-data user;
the desired data transmission parameter is a desired amount of data associated with the user's assignment; and
the network layer for the user's assignment is based on the desired amount of data and the velocity.
15. The invention of claim 10 , further comprising a model of the network used to determine criteria used to determine the layer of the network for the user's assignment.
16. The invention of claim 15 , wherein the model is a Markov chain.
17. The invention of claim 15 , wherein the model is used to estimate system throughput as a function of two or more assignment curves.
18. The invention of claim 17 , wherein an assignment curve that yields the maximum system throughput is selected as the criteria used to determine the layer of the network for the user's assignment.
19. The invention of claim 10 , wherein:
the desired data transmission parameter is one of desired data rate and desired amount of data; and
the greater the value of the desired data transmission parameter, the higher the probability that the assignment manager will select a lower layer in the overlay network for the user's assignment.
20. A computer-implemented method for assigning users in an overlay network having two or more layers, comprising:
receiving a request from a user for assignment within the network, wherein the request identifies a desired data transmission parameter for the user's assignment; and
selecting a layer of the network for the user's assignment based on the desired data transmission parameter, wherein:
the user is a real-time-data user;
the desired data transmission parameter is a desired data rate;
the network layer for the user's assignment is selected based on the desired data rate;
the request further identifies velocity of the user;
the network layer for the user's assignment is selected based on the velocity; and
the network layer for the user's assignment is selected by comparing the desired data rate and the velocity to an assignment curve that is dependent on both data rate and velocity.
21. The invention of claim 20 , wherein the assignment curve is a piecewise linear curve connecting two or more points in a two-dimensional space having data rate and velocity as its two dimensions.
22. A computer-implemented method for assigning users in an overlay network having two or more layers, comprising:
receiving a request from a user for assignment within the network, wherein the request identifies a desired data transmission parameter for the user's assignment;
selecting a layer of the network for the user's assignment based on the desired data transmission parameter; and
modeling the network to determine criteria used in selecting the network layer for the user's assignment, wherein:
the modeling is used to estimate system throughput as a function of two or more assignment curves; and
an assignment curve that yields the highest system throughput is used to select the network layer for the user's assignment.
23. The invention of claim 22 , wherein the modeling comprises using a Markov chain.
24. An assignment manager for assigning users in an overlay network having two or more layers, wherein the assignment manager:
receives a request from a user for assignment within the network, wherein the request identifies a desired data transmission parameter for the user's assignment; and
selects a layer of the network for the user's assignment based on the desired data transmission parameter, wherein:
the user is a real-time-data user;
the desired data transmission parameter is a desired data rate;
the network layer for the user's assignment is based on the desired data rate;
the request further comprises information about velocity of the user;
the network layer for the user's assignment is determined based on the velocity; and
the network layer for the user's assignment is determined by comparing the desired data rate and the velocity to an assignment curve that is dependent on both data rate and velocity.
25. The invention of claim 24 , wherein the assignment curve is a piecewise linear curve connecting two or more points in a two-dimensional space having data rate and velocity as its two dimensions.
26. An assignment manager for assigning users in an overlay network having two or more layers, wherein the assignment manager:
receives a request from a user for assignment within the network, wherein the request identifies a desired data transmission parameter for the user's assignment;
selects a layer of the network for the user's assignment based on the desired data transmission parameter; and
uses a model of the network to determine criteria used to determine the layer of the network for the user's assignment, wherein:
the model is used to estimate system throughput as a function of two or more assignment curves; and
an assignment curve that yields the maximum system throughput is selected as the criteria used to determine the layer of the network for the user's assignment.
27. The invention of claim 26 , wherein the model is a Markov chain.
28. A computer-implemented method for assigning users in an overlay network having two or more layers, comprising:
receiving a request from a user for assignment within the network, wherein the request identifies (1) a desired data transmission parameter for the user's assignment and (2) velocity of the user, wherein the velocity is independent of the desired data transmission parameter;
selecting a layer of the network for the user's assignment based on both the desired data transmission parameter and the velocity; and
modeling the network to determine criteria used in selecting the network layer for the user's assignment, wherein the modeling is used to estimate system throughput as a function of two or more assignment curves, wherein an assignment curve that yields the highest system throughput is used to select the network layer for the user's assignment.
29. The invention of claim 28 , wherein the modeling comprises using a Markov chain.
30. An assignment manager for assigning users in an overlay network having two or more layers, wherein the assignment manager:
receive a request from a user for assignment within the network, wherein the request identifies (1) a desired data transmission parameter for the user's assignment and (2) velocity of the user, wherein the velocity is independent of the desired data transmission parameter;
select a layer of the network for the user's assignment based on both the desired data transmission parameter and the velocity; and
use a model of the network to determine criteria used to determine the layer of the network for the user's assignment, wherein the model is used to estimate system throughput as a function of two or more assignment curves, wherein an assignment curve that yields the maximum system throughput is selected as the criteria used to determine a layer of the network for the user's assignment.
31. The invention of claim 30 , wherein the model is a Markov chain.