Connection acceptance control
View Patent ↗A method of bandwidth allocation for delivery of stored digital content from at least one server device to at least one client device by way of a network. The method includes the steps of: describing a control variable which represents a target flow rate of content from the server device to each client device; determining time-varying constraints on the target flow rate of the content; determining a cost function of the control variables for all clients, wherein the cost function represents an aggregate flow rate and is a sum of all flow rates for all clients; and prescribing bandwidth to all clients based upon a value of the control variables that maximize the cost function comprising performing periodic computations to update the value of the control variable such that the bandwidth can be continuously allocated to each client.
1. A method of bandwidth allocation for delivery of stored digital content from a server to one or more client devices in a session, the method comprising:
repeatedly calculating a control variable for each of the one or more client devices, wherein the control variable represents a minimum flow rate of content from the server to a client device by faster-than-real-time transmission, wherein the minimum flow rate of content is guaranteed not to increase during the session;
determining time-varying constraints for each minimum flow rate of content based on scheduled changes in network capacity;
determining a sum of all minimum flow rates of content for all client devices in the session; and
allocating excess bandwidth to all client devices in the session based on a cost function that maximizes revenue for a service provider, wherein excess bandwidth is based on a difference between network capacity and the sum of all minimum flow rates of content for all client devices in a session, and further based on the time-varying constraints.
2. The method of claim 1 , wherein the control variable further represents a maximum flow rate of content from the server to a client device.
3. The method of claim 1 , further comprising establishing a service prioritization of all client devices based upon at least one variable.
4. The method of claim 3 , wherein establishing a service prioritization of all client devices based upon at least one variable includes prioritizing all client devices based upon respective minimum flow rates.
5. The method of claim 3 , wherein establishing a service prioritization of all client devices based upon at least one variable includes prioritizing all clients based upon respective costs of service.
6. A method of bandwidth allocation for delivery of stored digital content from a server to one or more active clients in a session, comprising:
repeatedly determining a control variable which represents a minimum flow rate of the delivery of stored digital content from the server to an active client by faster-than-real-time transmission, wherein the minimum flow rate is guaranteed not to increase during the session;
determining time-varying constraints on the minimum flow rate based on scheduled changes in network capacity;
determining a sum of all minimum flow rates for all client devices in the session and
allocating bandwidth to all active clients based upon a cost function that maximized revenue for a service provider, wherein excess bandwidth is based on a difference between network capacity and the sum of all minimum flow rates for all client devices in a session, and further based on the time-varying constraints such that each active client is guaranteed bandwidth for successful delivery of the stored digital content from the server to the active client during the session.
7. The method of claim 6 , wherein the control variable further represents a maximum flow rate of the delivery of stored digital content from the server to an active client.
8. The method of claim 6 , wherein repeatedly determining a control variable occurs at least whenever a new client is activated during a session.
9. The method of claim 6 , further comprising establishing a service ordering of all active clients.
10. The method of claim 9 , wherein the service ordering is based on respective costs of service of all active clients.
11. A non-transitory computer-readable storage medium having embodied thereon instructions executable by a computer processor to perform a method of bandwidth allocation for delivery of stored digital content from a server to one or more active clients in a session, the method comprising:
repeatedly determining a control variable which represents a minimum flow rate of the delivery of stored digital content from the server to an active client by faster-than-real-time transmission, wherein the minimum flow rate is guaranteed not to increase during the session;
determining time-varying constraints on the minimum flow rate based on scheduled changes in network capacity;
determining a sum of all minimum flow rates for all client devices in the session; and
allocating bandwidth to all active clients based upon a cost function that maximized revenue for a service provider, wherein excess bandwidth is based on a difference between network capacity and the sum of all minimum flow rates for all client devices in a session, and further based on the time-varying constraints such that each active client is guaranteed bandwidth for successful delivery of the stored digital content from the server to the active client during the session.
12. The non-transitory computer-readable storage medium of claim 11 , wherein the control variable further represents a maximum flow rate of the delivery of stored digital content from the server to an active client.
13. The non-transitory computer-readable storage medium of claim 11 , wherein repeatedly determining a control variable occurs at least whenever a new client is activated during a session.
14. The non-transitory computer-readable storage medium of claim 11 , the method further comprising establishing a service ordering of all active clients.