Abstract: Aspects of the subject disclosure may include, for example, identifying a flow of data packets between first and second network addresses of a network, with each packet including respective header and payload portions. The identified flow of data packets is monitored over a number of sample periods to obtain a number of monitored results. A data-flow activity record is generated, having a number of symbols corresponding to the number of monitored results, the symbols including an active symbol value indicative of a presence of an exchange of data and an idle symbol value indicative of an absence of an exchange of data. A suitability of the identified data flow is inferred for estimating a throughput of the network according to the data-flow activity record without interpreting contents of each respective packet payload portion. Other embodiments are disclosed.

Abstract: A method for pre-fetching and storing media locally on customer premises equipment includes predicting, using a machine learning model that analyzes historical video consumption behavior in a home broadband network, a video content that is likely to be consumed by a user of the home broadband network during a next period of peak demand on a communication service provider network, scheduling a download of the video content to a local storage during a next period of predicted non-peak demand in a cell of the communication service provider network that serves the home broadband network, and downloading, in response to determining that the next period of predicted non-peak demand has occurred, the video content to the local storage.

Abstract: Aspects of the subject disclosure may include, for example, identifying a flow of data packets between first and second network addresses of a network, with each packet including respective header and payload portions. The identified flow of data packets is monitored over a number of sample periods to obtain a number of monitored results. A data-flow activity record is generated, having a number of symbols corresponding to the number of monitored results, the symbols including an active symbol value indicative of a presence of an exchange of data and an idle symbol value indicative of an absence of an exchange of data. A suitability of the identified data flow is inferred for estimating a throughput of the network according to the data-flow activity record without interpreting contents of each respective packet payload portion. Other embodiments are disclosed.