US 11,743,351 B2
Service allocation selection method and service allocation selection program
Misao Kataoka, Musashino (JP); Hirofumi Noguchi, Musashino (JP); Takuma Isoda, Musashino (JP); and Yoji Yamato, Musashino (JP)
Assigned to Nippon Telegraph and Telephone Corporation, Tokyo (JP)
Appl. No. 17/429,530
Filed by Nippon Telegraph and Telephone Corporation, Tokyo (JP)
PCT Filed Feb. 14, 2020, PCT No. PCT/JP2020/005688
§ 371(c)(1), (2) Date Aug. 9, 2021,
PCT Pub. No. WO2020/170952, PCT Pub. Date Aug. 27, 2020.
Claims priority of application No. 2019-028789 (JP), filed on Feb. 20, 2019.
Prior Publication US 2022/0109732 A1, Apr. 7, 2022
Int. Cl. H04L 41/0806 (2022.01); H04L 67/51 (2022.01); H04L 67/10 (2022.01)
CPC H04L 67/51 (2022.05) [H04L 67/10 (2013.01)] 17 Claims
OG exemplary drawing
 
1. A service allocation selection method, under an environment where at least a plurality of IoT devices capable of outputting data and a plurality of computers capable of executing a plurality of pieces of application software for processing the output data of each of the plurality of IoT devices are individually connected to one another at different positions via a predetermined communication network, for determining, when a piece of new application software is activated as a specific application x, an optimized combination of the specific application x, one specific computer y selected from among the plurality of computers, and one or more specific devices z selected from among the plurality of IoT devices, the service allocation selection method comprising:
calculating data effectiveness expectation value J(x, z) regarding a combination of the specific application x and a specific device z of the one or more specific devices z, wherein the one or more specific devices z comprise a physical sensor configured to collect data from a surrounding environment in which the one or more specific devices z is located;
calculating a processing time increase amount expectation value G(x, y, z) for the specific application x with a combination with the specific computer y and the specific device z, wherein G(x, y, z) depends on a ratio computed as I(x,y)/J(x, z), wherein I(x,y) is an estimation of processing time for processing the specific application x on the specific computer y while taking into account a computation load of the specific computer y;
calculating an evaluation value F(x, y, z) including at least the processing time increase amount expectation value G(x, y, z); and
selecting a combination of the specific application x, the one specific computer y, and the specific device z by using the evaluation value F(x, y, z).