System and methods for smart intrusion detection using wireless signals and artificial intelligence
An intelligent entrance detection system is provided where, primarily, any authorized or unauthorized entry to an area of a residential (or small industrial) unit covered by a wireless network is automatically detected from active off-the-shelf devices in that area. After the detection, an identification algorithm is employed to verify if this entry is a legal or illegal action. Based on this verification, either the routine smart home system is activated or a hazardous monitoring period begins to further investigate the suspicious event. If the illegal entrance is confirmed during the hazardous monitoring period, the owner of the property is informed through an intruder alarm protocol. In this invention, all of the analytic and processing steps, including entrance detection, owner identification, and device-free authority verification are designed based on monitoring and quantification of changes in surrounding wireless signals originated by human or object movements within the sensing area.
1. A method comprising:
providing a monitoring system comprising:
a receiving circuit for receiving metrics from each wireless device of a plurality of wireless devices associated with a sensing region operating according to a common wireless standard where each wireless device of the plurality of wireless devices generates metrics in dependence upon wireless signals received by the wireless device of the plurality of devices from other wireless devices of the plurality of wireless devices;
a processing block comprising a pre-processing system, a plurality of detection blocks and a decision making block comprising a plurality of decision making modules; wherein
the pre-processing system comprises a bank of digital filters according to a predetermined architecture to perform signal conditioning upon the received metrics from the plurality of wireless devices to generate a plurality of signal paths of filtered signals;
each detection block of the plurality of detection blocks receives a predetermined subset of the plurality of signal paths; and
the predetermined subset is established in dependence upon an activity the detection block of the plurality of detections blocks establishes a classification for; wherein
the monitoring system executes a sequence comprising:
determining that an entry has been made into the sensing region;
determining whether a legal entry into the sensing area has been made;
upon a positive determination passing a subset of the plurality of signal paths to one or more home automation applications; and
upon a negative determination passing a subset of the plurality of signal paths to one or more security applications;
determination as to whether an entry has been made into the sensing area is established in dependence upon the processing block establishing with a template matching technique whether the sensing region is empty or not;
the template matching technique employs a template for a static empty profile established from a histogram of the pre-processed received wireless signals; and
the static empty profile of the sensed region from wireless measurements was established without a presence of a person or vehicle within the sensing region.
2. The method according to claim 1 , wherein
the outputs from the plurality of detection blocks are only coupled to the decision making block when all outputs from the plurality of detection blocks are established.
3. The method according to claim 1 , wherein
the processing block comprises a plurality of processing units; and
each processing unit executes a semantic and logical strategy relating to an action of a predetermined set of actions.
4. The method according to claim 1 , wherein
the wireless devices forming the plurality of wireless devices varies with time as at least one of portable electronic devices and mobile electronic devices the wireless devices are associated with at least one of enter and leave the sensing region.
5. The method according to claim 1 , wherein
the monitoring system also receives complimentary measurements which are independent of the wireless signals;
the pre-processing system applies one or more first signal processing techniques of a plurality of signal processing techniques to the received metrics where the one of more first signal processing techniques are established in dependence upon a type of the received metrics;
the pre-processing system applies one or more second signal processing techniques of the plurality of signal processing techniques to the complimentary measurements where the one of more second signal processing techniques are established in dependence upon a type of the complimentary measurements;
the pre-processing system applies one or more specialized processing units of a plurality of specialized processing units to the received metrics where the one or more specialized processing units are established in dependence upon the type of the received metrics; and
the pre-processing system applies one or more other specialized processing units of the plurality of specialized processing units to the complimentary measurements where the one or more other specialized processing units are established in dependence upon the type of the complimentary measurements.
6. The method according to claim 1 , wherein
the system operates upon a sequence of time frames;
at the end of each time frame the decision making block determines what information is needed and establishes enquiries to a subset of the plurality of decision making blocks to collect data from the subset of the plurality of decision making blocks; and
the determination of what information is needed is established in dependence upon a past status of the system in a previous time frame and current status of the system in the time frame just ended.
7. The method according to claim 1 , wherein
the static empty profile and template are at least one of periodically updated and verified by establishing updates when the sensing region is determined to be empty.
8. A method comprising:
providing a monitoring system comprising:
a receiving circuit for receiving metrics from each wireless device of a plurality of wireless devices associated with a sensing region operating according to a common wireless standard where each wireless device of the plurality of wireless devices generates metrics in dependence upon wireless signals received by the wireless device of the plurality of devices from other wireless devices of the plurality of wireless devices;
a processing block comprising a pre-processing system, a plurality of detection blocks and a decision making block comprising a plurality of decision making modules; wherein
the pre-processing system comprises a bank of digital filters according to a predetermined architecture to perform signal conditioning upon the received metrics from the plurality of wireless devices to generate a plurality of signal paths of filtered signals;
each detection block of the plurality of detection blocks receives a predetermined subset of the plurality of signal paths; and
the predetermined subset is established in dependence upon an activity the detection block of the plurality of detections blocks establishes a classification for; wherein
the monitoring system executes a sequence comprising:
determining that an entry has been made into the sensing region;
establishing a monitoring phase of predetermined duration employing a predefined subset of the plurality of detection blocks and a subset of the plurality of decision making modules;
determining with a subset of the plurality of processing blocks whether during the monitoring phase either that no anomaly within the sensing region was detected or that the entry relates to entry of a subject identified as being a safe or legal entry;
upon a positive determination during the monitoring phase that either no anomaly within the sensing region was detected or that the entry relates to entry of a subject was identified as being a safe or legal entry terminating the monitoring period passing another subset of the plurality of signal paths to another subset of the plurality of decision making modules relating to one or more smart home applications; and
determination that the subject is identified as being safe or legal is established in dependence upon analysis of the perturbations within the wireless signals within the sensing area to identify the user based upon a walking pattern of the user and a body shape of the user.
9. A method comprising:
providing a monitoring system comprising:
a receiving circuit for receiving metrics from each wireless device of a plurality of wireless devices associated with a sensing region operating according to a common wireless standard where each wireless device of the plurality of wireless devices generates metrics in dependence upon wireless signals received by the wireless device of the plurality of devices from other wireless devices of the plurality of wireless devices;
a processing block comprising a pre-processing system, a plurality of detection blocks and a decision making block comprising a plurality of decision making modules; wherein
the pre-processing system comprises a bank of digital filters according to a predetermined architecture to perform signal conditioning upon the received metrics from the plurality of wireless devices to generate a plurality of signal paths of filtered signals;
each detection block of the plurality of detection blocks receives a predetermined subset of the plurality of signal paths; and
the predetermined subset is established in dependence upon an activity the detection block of the plurality of detections blocks establishes a classification for; wherein
the monitoring system executes a sequence comprising:
determining that an entry has been made into the sensing region;
establishing a monitoring phase of predetermined duration employing a predefined subset of the plurality of detection blocks and a subset of the plurality of decision making modules;
determining with a subset of the plurality of processing blocks whether during the monitoring phase either that no anomaly within the sensing region was detected or that the entry relates to entry of a subject identified as being a safe or legal entry; and
upon a positive determination during the monitoring phase that either no anomaly within the sensing region was detected or that the entry relates to entry of a subject was identified as being a safe or legal entry terminating the monitoring period passing another subset of the plurality of signal paths to another subset of the plurality of decision making modules relating to one or more smart home applications; and
determination that the subject is identified as being safe or legal is established in dependence upon analysis of the perturbations within the wireless signals within the sensing area with respect to pre-defined user profiles established for subjects identified as being safe or legal.
10. A method comprising:
providing a monitoring system comprising:
a receiving circuit for receiving metrics from each wireless device of a plurality of wireless devices associated with a sensing region operating according to a common wireless standard where each wireless device of the plurality of wireless devices generates metrics in dependence upon wireless signals received by the wireless device of the plurality of devices from other wireless devices of the plurality of wireless devices;
a processing block comprising a pre-processing system, a plurality of detection blocks and a decision making block comprising a plurality of decision making modules; wherein
the pre-processing system comprises a bank of digital filters according to a predetermined architecture to perform signal conditioning upon the received metrics from the plurality of wireless devices to generate a plurality of signal paths of filtered signals;
each detection block of the plurality of detection blocks receives a predetermined subset of the plurality of signal paths; and
the predetermined subset is established in dependence upon an activity the detection block of the plurality of detections blocks establishes a classification for; wherein
the monitoring system executes a sequence comprising:
determining that an entry has been made into the sensing region;
establishing a monitoring phase of predetermined duration employing a predefined subset of the plurality of detection blocks and a subset of the plurality of decision making modules;
determining with a subset of the plurality of processing blocks whether during the monitoring phase either that no anomaly within the sensing region was detected or that the entry relates to entry of a subject identified as being a safe or legal entry; and
upon a positive determination during the monitoring phase that either no anomaly within the sensing region was detected or that the entry relates to entry of a subject was identified as being a safe or legal entry terminating the monitoring period passing another subset of the plurality of signal paths to another subset of the plurality of decision making modules relating to one or more smart home applications;
determination that the subject is identified as being safe or legal is established in dependence upon analysis of the perturbations within the wireless signals within the sensing area to identify the user based upon a walking pattern of the user; and
where the entry into the area is made with a vehicle user identification is enabled to detect the identity of the subject through gait recognition as soon as the subject exists the vehicle and begins walking.