Method for allocating communication network resources using adaptive demand prediction
View Patent ↗A method for allocating network resources of a communication network ( 100 ) to a plurality of users is disclosed. The allocation method includes predicting figure access demand from users and making resource allocation decisions based, at least in part, on the predicted demand.
1. In a communications network having a plurality of devices competing for network resources, a method for allocating the network resources comprising the steps of:
receiving at a network resource manager, from each of said plurality of devices, a request for network access, a first set of distribution parameters for the distribution of time when the device is active and a second set of distribution parameters for the distribution of the periods of time the device is inactive;
predicting whether sufficient network resources exist to accommodate said request based on the first set of distribution parameters and the second set of distribution parameters for each of said plurality of devices; and
allocating the network resources in accordance with said prediction.
2. The method of claim 1 , further comprising transmitting, from said network resource manager to said plurality of devices, information indicative of the allocation of the network resources.
3. The method of claim 1 , wherein said step of allocating network resources comprises allocating network resources in a cellular telephone network and further wherein said plurality of devices comprise a corresponding plurality of cellular telephones.
4. The method of claim 1 , further comprising providing a preemption process to allow a high-priority device to preempt service from a low-priority device.
5. A network resource manager for allocating network resources comprising:
a demand prediction processor operable to store for each of a plurality of devices coupled to the network resource manager a first set of distribution parameters associated with the distribution of the period of time when the device is active and a second set of distribution parameters associated with the distribution of the periods of time when the device is inactive, the demand prediction processor further operable to calculate, upon receiving a request for network access, an estimated probability of whether each of the plurality of devices will be active or inactive; and
a network allocator coupled to the demand processor, the network allocator operable to receive the estimated probability and to generate network resource allocation decisions based on the estimated probability.
6. The network resource manager of claim 5 , wherein the network resource allocation decisions are sent to each of the plurality of devices.
7. The network resource manager of claim 5 wherein the network resource allocation decisions are stored in a decision history database coupled to the network allocator.