We have the States table - which records the current state of that sensor/platform.
We have the Contacts table - which records a measurement of another platform.
So, the location in the Contacts table is the estimated location of the other platform, as produced by a radar, for example.
But, a sensor may include its' own location. When a sonar is towed 1km behind a ship, it will have it's own location. So, I think that sensor is producing a State and a Contact. So, it may be providing two locations: it's own, and the location of the platform it's detecting.
So, when we're parsing the line of data for that platform, we may be storing it's own state in the States table, and the measurements on the other platform in the Contacts table. Actually, there could even be a comment associated with the detection, which would go in the Comments table. That's one line of text data that is being broken up and put into 3 tables.
The entries in the 3 tables are linked by having exactly the same time. But, I wonder if we need some other linking if we're going to allow the data to be de-normalised.
We have the
Statestable - which records the current state of that sensor/platform. We have theContactstable - which records a measurement of another platform.So, the location in the
Contactstable is the estimated location of the other platform, as produced by a radar, for example.But, a sensor may include its' own location. When a sonar is towed 1km behind a ship, it will have it's own location. So, I think that sensor is producing a
Stateand aContact. So, it may be providing two locations: it's own, and the location of the platform it's detecting.So, when we're parsing the line of data for that platform, we may be storing it's own state in the
Statestable, and the measurements on the other platform in theContactstable. Actually, there could even be a comment associated with the detection, which would go in theCommentstable. That's one line of text data that is being broken up and put into 3 tables.The entries in the 3 tables are linked by having exactly the same time. But, I wonder if we need some other linking if we're going to allow the data to be de-normalised.