Abstract: Example implementations relate to the processing of refresh token requests at an API gateway. The API gateway determines a first time associated with receipt of the refresh token request and a second time associated with the generation of a current access token. The current access token and a refresh token in the refresh token request are provided by the API gateway to the client device for accessing a backend service. The API gateway determines whether a difference between the first time and the second time is within a pre-defined threshold duration. When the difference between the first time and the second time is within the pre-defined threshold, the API gateway denies the refresh token request for generating the new access token and transmits the current access token back to the client device.

Abstract: Example implementations relate to the processing of refresh token requests at an API gateway. The API gateway determines a first time associated with receipt of the refresh token request and a second time associated with the generation of a current access token. The current access token and a refresh token in the refresh token request are provided by the API gateway to the client device for accessing a backend service. The API gateway determines whether a difference between the first time and the second time is within a pre-defined threshold duration. When the difference between the first time and the second time is within the pre-defined threshold, the API gateway denies the refresh token request for generating the new access token and transmits the current access token back to the client device.

Abstract: A method for classifying devices may include determining that a device is located at a location for more than a threshold amount of time and identifying that the device is a moving candidate based on a percentage of times a plurality of signal strength values for the device fall inside a baseline interval. The method may also include determining a most frequent pattern set of signal strength values for the device, classifying the device as fixed if the most frequent pattern set of signal strength values for the device includes at least one overlapping interval, and classifying the device as moving if the most frequent pattern set of signal strength values for the device includes at least one non-overlapping interval.

Abstract: Aspects of categorizing MAC address randomization include obtaining probe data for a plurality of time intervals. A likelihood measure of a specified OS type and version belonging to each of a plurality of MAC address randomization categories is determined and the specified OS type and version is categorized based on the likelihood measure.

Abstract: A method for classifying devices may include determining that a device is located at a location for more than a threshold amount of time and identifying that the device is a moving candidate based on a percentage of times a plurality of signal strength values for the device fall inside a baseline interval. The method may also include determining a most frequent pattern set of signal strength values for the device, classifying the device as fixed if the most frequent pattern set of signal strength values for the device includes at least one overlapping interval, and classifying the device as moving if the most frequent pattern set of signal strength values for the device includes at least one non-overlapping interval.

Abstract: Aspects of categorizing MAC address randomization include obtaining probe data for a plurality of time intervals. A likelihood measure of a specified OS type and version belonging to each of a plurality of MAC address randomization categories is determined and the specified OS type and version is categorized based on the likelihood measure.