# US 10,313,470 B2 · Hierarchical caching and analytics > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 10,313,470 B2 - Document ID: US10313470B2 - Status: Granted - Kind: B2 - Application number: 15/257,026 - Filed: 2016-09-06 - Published: 2018-03-08 - Granted: 2019-06-04 - Earliest claimed: 2016-09-06 - Assignees: Integrated Device Technology, Inc. - Applicants: Integrated Device Technology, Inc. - Inventors: Mohammad S. Akhter - Web page: https://patentgrep.com/patent/US10313470B2 ## CPC - H04L 41/147: for predicting network behaviour - H04L 67/568: Storing data temporarily at an intermediate stage, e.g. caching - H04L 49/25: Routing or path finding in a switch fabric - H04L 67/12: specially adapted for proprietary or special-purpose networking environments, e.g. medical networks, sensor networks, networks in vehicles or remote metering networks - H04W 4/44: for communication between vehicles and infrastructures, e.g. vehicle-to-cloud [V2C] or vehicle-to-home [V2H] - H04W 4/46: for vehicle-to-vehicle communication [V2V] - H04L: TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION (arrangements common to telegraphic and telephonic communication H04M) - H04W: WIRELESS COMMUNICATION NETWORKS (broadcast communication H04H; communication systems using wireless links for non-selective communication, e.g. wireless extensions H04M1/72) ## Abstract A system includes at least one end-node, at least one edge node, and an edge cloud video headend. The at least one end node generally implements a first stage of a multi-stage hierarchical analytics and caching technique. The at least one edge node generally implements a second stage of the multi-stage hierarchical analytics and caching technique. The edge cloud video headend generally implements a third stage of the multi-stage hierarchical analytics and caching technique. ## Claims ### Claim 1 A system comprising: at least one end-node implementing a first stage of a multi-stage hierarchical analytics and caching technique, wherein said end-node is configured to communicate directly with one or more other end-nodes via one or more wireless links; at least one edge node implementing a second stage of the multi-stage hierarchical analytics and caching technique, wherein said edge node is configured to communicate directly with one or more end nodes and one or more other edge nodes via wireless links; and an edge cloud video headend implementing a third stage of the multi-stage hierarchical analytics and caching technique and configured to communicate with one or more of said end-nodes and one or more of said edge nodes, wherein said edge cloud video headend is configured to communicate with said one or more edge nodes using cable interconnections, optical interconnections, and wireless interconnections, the at least one end-node, the at least one edge node, and the edge cloud video headend each comprise an analytics and caching module/server that includes a flexible computing engine configured to implement low latency remote memory and storage access acceleration and predictive traffic shaping, and a low latency switching fabric performs predictive routing of packets using a destination ID assigned to each node and maps the destination ID to individual wavelengths of the optical interconnections. ### Claim 2 (depends on 1) The system according to claim 1, wherein the at least one edge node comprises one or more of a camera node and a sensor node. ### Claim 3 (depends on 1) The system according to claim 1, wherein each flexible computing engine is further configured to implement predictive analytics and caching of video and image data based on pre-defined and learned metrics. ### Claim 4 (depends on 3) The system according to claim 3, wherein the learned metrics are based on use. ### Claim 5 (depends on 1) The system according to claim 1, wherein each analytics and caching module/server includes a time and node metadata synchronization engine. ### Claim 6 (depends on 1) The system according to claim 1, wherein each analytics and caching module/server implements a caching and storage function directly attached to said low latency switching fabric. ### Claim 7 (depends on 6) The system according to claim 6, wherein attachment of the caching and storage function to the low latency switching fabric is implemented through both a wireless link and an integrated photonics link. ### Claim 8 (depends on 1) The system according to claim 1, wherein respective analytics and caching module/servers of two or more end-nodes are configured to communicate via a wireless link with each other and with a respective edge node. ### Claim 9 An apparatus comprising: a low latency switch fabric enabled to communicate via one or more wireless links and one or more optical links; a first flexible computing engine directly attached to the low latency switch fabric and configured to implement predictive analytics and caching; a second flexible computing engine directly attached to the low latency switch fabric and configured to implement low latency remote memory and storage access acceleration and predictive fabric traffic shaping, wherein the low latency switch fabric performs predictive routing of packets using a destination ID assigned to each node and maps the destination ID to individual wavelengths of the optical links; a synchronization engine directly attached to the low latency switch fabric and configured to synchronize time and node metadata; and a caching and storage module directly attached to the low latency switch fabric by one of said wireless links and one of said optical links. ### Claim 10 (depends on 9) The apparatus according to claim 9, wherein said apparatus implements an analytics and caching server platform. ### Claim 11 (depends on 9) The apparatus according to claim 9, further comprising a video encoding and decoding module directly attached to the low latency switch fabric and configured to receive at least one camera feed. ### Claim 12 (depends on 9) The apparatus according to claim 9, wherein said low latency switch fabric further comprises a wireless link processing module and an integrated photonics link processing module. ### Claim 13 (depends on 12) The apparatus according to claim 12, wherein said integrated photonics link processing module is configured to communicate data using different protocols on different wavelengths. ### Claim 14 (depends on 12) The apparatus according to claim 12, wherein said wireless link processing module and said integrated photonics link processing module are configured to distribute a data transmission across the one or more wireless links and the one or more optical links. ### Claim 15 (depends on 9) The apparatus according to claim 9, wherein at least one of said first flexible computing engine and said second flexible computing engine comprise one or more of a field-programmable gate array, a graphics processing unit, and a digital signal processor. ### Claim 16 (depends on 15) The apparatus according to claim 15, wherein said at least one of said first flexible computing engine and said second flexible computing engine further comprise a local low latency switch fabric. ### Claim 17 A method of real-time decision making comprising: implementing a first stage of a multi-stage hierarchical analytics and caching technique in at least one end-node, wherein said end-node is configured to communicate directly with one or more other end-nodes via one or more wireless links; implementing a second stage of the multi-stage hierarchical analytics and caching technique in an edge node comprising at least one of a camera node and a sensor node, wherein said edge node is configured to communicate with one or more other edge nodes via one or more wireless links and with one or more end-nodes via one or more wireless links; and implementing a third stage of the multi-stage hierarchical analytics and caching technique in an edge cloud video headend, wherein said edge cloud video headend is configured to communicate with one or more of said end-nodes and one or more of said edge nodes, and said edge cloud video headend is configured to communicate with said one or more edge nodes using cable interconnections, optical interconnections, and wireless interconnections, the at least one end-node, the at least one edge node, and the edge cloud video headend each comprise an analytics and caching module/server that includes a flexible computing engine configured to implement low latency remote memory and storage access acceleration and predictive traffic shaping, and a low latency switching fabric performs predictive routing of packets using a destination ID assigned to each node and maps the destination ID to individual wavelengths of the optical interconnections. ### Claim 18 (depends on 17) The method according to claim 17, wherein respective analytics and caching module/servers of two or more end-nodes are configured to communicate via a wireless link with each other and via a wireless link with a respective edge node.