Proximity in privacy and security enhanced internet geolocation
The present innovation extends the innovations inherent in our prior work on security and privacy enhanced geolocation to address specific problems of determining the proximity of two Internet connected devices. This could be used for applications such as determining proximity of a user's smartphone to an ATM machine they are using. Or to determine the physical proximity of a shopper and a cashier at a check out line. A straightforward application of the prior work will not suffice, as devices may have very different types of connectivity to the Internet. We solve this problem by placing reciprocal slave servers on each of the devices and giving high weight to the time taken for messages to travel from each device to the slave on the other device. The system can be extended to more than two entities by performing the calculation in a pairwise fashion and using further optimizations.
1. A method for determining the proximity of two entities with Internet connected devices, without using IP geolocation or GPS data, and by using the Time Taken to Travel at that particular Time for a message of size k (4Tk) statistics obtained by sending multiple characters from or through the Internet connected devices to multiple slaves by:
a. Sending messages including multiple characters from or through the Internet connected devices to multiple slaves;
b. Placing N geographically distributed servers across the Internet, including one on each of the Internet connected devices;
c. Calculating a statistic on the mean time taken for messages of size k to travel from or through the Internet connected devices to the N servers at various times of day;
d. Determining an average time taken for the data sent between the Internet connected devices, wherein the data sent between the Internet connected devices is of size k and each Internet connected device transmits and receives data;
e. Identifying a particular one of the various times of day matching a time of transmission of the data sent between the Internet connected devices; and
f. Determining the proximity of the two entities based on the statistic of the mean time, the average time taken for the data sent between the Internet connected devices, and the particular one of the various times of day.
2. A method according to claim 1 wherein the messages are formed by users typing characters including usernames or CAPTCHAs on the devices, with optional padding of each character.
3. A method according to claim 1 wherein the messages are automatically generated based on a signal from the user.
4. A method for determining the proximity of at least three entities with Internet connected devices in a pairwise fashion, without using IP geolocation or GPS data, and by using the Time Taken to Travel at that particular Time for a message of size k (4Tk) statistics obtained by sending multiple characters from or through the Internet connected devices to multiple slaves by:
sending messages including multiple characters from or through pairs of the Internet connected devices to multiple slaves;
placing N geographically distributed servers across the Internet, including one on each of the Internet connected devices;
calculating a statistic on the mean time taken for messages of size k to travel from or through the Internet connected devices to the N servers at various times of day;
determining an average time taken for the data sent between pairs of the Internet connected devices, wherein the data sent between the pairs of the Internet connected devices is of size k and each Internet connected device transmits and receives data;
identifying a particular one of the various times of day matching a time of transmission of the data sent between the pairs of the Internet connected devices; and
determining the proximity of the at least three entities based on the statistic of the mean time, the average time taken for the data sent between pairs of the Internet connected devices, and the particular one of the various times of day.
5. A method according to claim 4 wherein determining the proximity of the at least three entities utilizes optimizations including deriving from the knowledge that entity A is proximate to entity B, and that entity B is proximate to entity C, the information that entity A and entity C are also proximate without having to perform any pairwise calculation.
6. A method for determining the proximity of two entities with Internet connected devices, without using IP geolocation or GPS data, and by using the Time Taken to Travel at that particular Time for a message of size k (4Tk) statistics obtained by sending multiple characters from or through the Internet connected devices to multiple slaves by:
detecting messages including multiple characters sent from or through the Internet connected devices to multiple slaves;
placing N geographically distributed servers across the Internet, including one on each of the Internet connected devices;
making the Internet connected devices reciprocal slaves with one another such that data sent between the Internet connected devices travels a same distance over equivalent networks in each direction;
calculating a mean time value for messages of size k to travel from or through the Internet connected devices to the N servers at various times of day;
determining an average time taken for the data sent between the Internet connected devices to travel the same distance over the equivalent networks, wherein the data sent between the Internet connected devices is of size k and each Internet connected device transmits and receives data;
identifying a particular one of the various times of day matching a time of transmission of the data sent between the Internet connected devices; and
determining a value of the same distance of the two Internet connected devices based on the average time, the particular one of the various times of day, and the mean time value, wherein the same distance is indicative of the proximity of the two entities.
7. A method according to claim 6 wherein, the data sent between the Internet connected devices is intermediately routed through at least one other of the N geographically distributed servers.