US 12,407,705 B1
Prediction of network vulnerability of a user's network-connected smart device using crowdsourced vulnerability profiles
Michael D. Melnick, Brighton, NY (US); David L Knudsen, Saint Helena, CA (US); and Alyssa J. Kersey, Chicago, IL (US)
Assigned to EVERYTHING SET INC., Berkeley, CA (US)
Filed by EVERYTHING SET INC., Berkeley, CA (US)
Filed on Feb. 22, 2022, as Appl. No. 17/677,240.
Int. Cl. H04L 9/40 (2022.01)
CPC H04L 63/1433 (2013.01) [H04L 63/1425 (2013.01)] 7 Claims
OG exemplary drawing
 
1. An automated method of predicting network vulnerability of a user's network-connected smart device, wherein the user's network-connected smart device has one or more device classifications, the method comprising:
(a) storing in memory:
(i) a plurality of crowdsourced vulnerability profiles generated from individual vulnerability profiles of a plurality of network-connected smart devices, each crowdsourced vulnerability profile being generated from individual vulnerability profiles of a plurality of network-connected smart devices having one of the same device classifications, each of the individual vulnerability profiles being created from network scans for a respective network-connected smart device, and
(ii) anomalous behavior associated with each of a respective crowdsourced vulnerability profile, wherein a database in the memory stores anomalous behavior for each crowdsourced vulnerability profile in a one-to-one correlation, and wherein the anomalous behavior is separate and distinct from the crowdsourced vulnerability profile;
(b) generating, by a processor, a vulnerability profile of the user's network-connected smart device;
(c) identifying, by the processor, the crowdsourced vulnerability profiles in the memory that match the vulnerability profile of the user's network-connected smart device by matching the one or more device classifications of the user's network-connected smart device with the device classifications of the crowdsourced vulnerability profiles in the memory; and
(d) identifying, using the matched crowdsourced vulnerability profiles and the database that stores anomalous behavior for each crowdsourced vulnerability profile in a one-to-one correlation, a percentage of each anomalous behavior associated with the crowdsourced vulnerability profiles that the user's network-connected smart device is at risk of exhibiting.