Back to results
US 11,962,668 B2Granted
Caching method, device, and computer program product for edge server
AssigneeDell Products L.P.
InventorsSi Chen, Zijia Wang, Zhenzhen Lin
Life of this patent
Estimating the term…
Abstract
Embodiments of the present disclosure relate to a caching method, device, and computer program product for an edge server. The method includes acquiring a historical distribution of a set of variables associated with a network environment of the edge server, and generating one or more time series of the set of variables based on the historical distribution. The method further includes determining a caching strategy associated with the edge server based on the one or more time series of the set of variables. With the solution of the present disclosure, in an edge computing scenario, unnecessary overhead caused by ineffective caching can be reduced while ensuring smooth user experience.
Claims20
A caching method for an edge server, comprising:
- acquiring a historical distribution of a set of variables associated with a network environment of the edge server;
- generating one or more time series of the set of variables based on the historical distribution;
- and
- determining a caching strategy associated with the edge server based on the one or more time series of the set of variables, the caching strategy being configured to:
- determine, based on values of the set of variables at a certain time and values of a set of states of the edge server at the time, a change to a cache of the edge server that should occur at the time and is related to online content, wherein the online content is associated with one or more user devices, and the one or more user devices are located within a service area of the edge server at the time;
- each of at least a subset of a plurality of content items of the online content being encoded as a combination of a first vector representing a type of the content item and a second vector representing a remaining lifetime of the content item to form a semantic embedding;
- the one or more time series of the set of variables characterizing at least movement of the one or more user devices within the network environment of the edge server, the network environment comprising a plurality of edge servers including a target edge server to which a connection of a given one of the user devices will be switched as the given user device moves within the network environment;
- wherein the caching strategy is determined in accordance with a reinforcement learning model in which the change to the cache comprises one of a plurality of actions of an action time series characterizing caching actions of the edge server, each of the actions of the action time series being selected based at least in part on (i) one or more states of a state time series characterizing corresponding states of the edge server, and (ii) a reward function of the reinforcement learning model.
5 more elements
An electronic device, comprising:
- a processor;
- and
- a memory coupled to the processor, wherein the memory has instructions stored therein which, when executed by the processor, cause the electronic device to execute actions comprising:
- acquiring a historical distribution of a set of variables associated with a network environment of an edge server;
- generating one or more time series of the set of variables based on the historical distribution;
- and
- determining a caching strategy associated with the edge server based on the one or more time series of the set of variables, the caching strategy being configured to:
- determine, based on values of the set of variables at a certain time and values of a set of states associated with the edge server at the time, a change to a cache of the edge server that should occur at the time and is related to online content, wherein the online content is associated with one or more user devices, and the one or more user devices are located within a service area of the edge server at the time;
- each of at least a subset of a plurality of content items of the online content being encoded as a combination of a first vector representing a type of the content item and a second vector representing a remaining lifetime of the content item to form a semantic embedding;
- the one or more time series of the set of variables characterizing at least movement of the one or more user devices within the network environment of the edge server, the network environment comprising a plurality of edge servers including a target edge server to which a connection of a given one of the user devices will be switched as the given user device moves within the network environment;
- wherein the caching strategy is determined in accordance with a reinforcement learning model in which the change to the cache comprises one of a plurality of actions of an action time series characterizing caching actions of the edge server, each of the actions of the action time series being selected based at least in part on (i) one or more states of a state time series characterizing corresponding states of the edge server, and (ii) a reward function of the reinforcement learning model.
8 more elements
A computer program product tangibly stored on a non-transitory computer-readable medium and comprising machine-executable instructions, wherein the machine-executable instructions, when executed by a machine, cause the machine to perform a caching method for an edge server, the method comprising:
- acquiring a historical distribution of a set of variables associated with a network environment of the edge server;
- generating one or more time series of the set of variables based on the historical distribution;
- and
- determining a caching strategy associated with the edge server based on the one or more time series of the set of variables, the caching strategy being configured to:
- determine, based on values of the set of variables at a certain time and values of a set of states of the edge server at the time, a change to a cache of the edge server that should occur at the time and is related to online content, wherein the online content is associated with one or more user devices, and the one or more user devices are located within a service area of the edge server at the time;
- each of at least a subset of a plurality of content items of the online content being encoded as a combination of a first vector representing a type of the content item and a second vector representing a remaining lifetime of the content item to form a semantic embedding;
- the one or more time series of the set of variables characterizing at least movement of the one or more user devices within the network environment of the edge server, the network environment comprising a plurality of edge servers including a target edge server to which a connection of a given one of the user devices will be switched as the given user device moves within the network environment;
- wherein the caching strategy is determined in accordance with a reinforcement learning model in which the change to the cache comprises one of a plurality of actions of an action time series characterizing caching actions of the edge server, each of the actions of the action time series being selected based at least in part on (i) one or more states of a state time series characterizing corresponding states of the edge server, and (ii) a reward function of the reinforcement learning model.
5 more elements
Dependencies are inferred from the claim text.