SUBSIM Radio Room Forums - View Single Post

Fidd · 01-31-24, 01:37 AM

115. Thoughts on escort radar implementation.......

There are a number of players who are somewhat nervous at the prospect of the gameplay ramifications of escorts receiving radar. As I see it, this is something not appropriate for use by AI, but rather one better suited to human interpretation, and the full range of deficiencies early radar possessed.

Early radar was tempramental, and when it did work, suffered problems in rougher sea-states where the escort was pitching and rolling, as any time the radar-head was tilted down, it caused back-scatter from the surface - effectively false returns, and viewed over a period of time, it meant a considerable skill was needed on the part of the operator, to spot an actual consistent return on a given bearing amongst the false returns, if the escort was in rougher water. It is forecastable that AI interpreting such returns is either going to be far too capable and recognising a return as a u-boat, or, will be hopeless at it. A happy medium in AI's ability in almost any situation is usually not realised in game. So, conclusion one is that escorts should be playable before radar is implemented, and it be usable only by human players.

Early radar had an oscilloscope presentation, rather than the PPI (Plan position indicator) we know today. In other words as the radar-head rotated through a bearing, assuming the radar head was not pointing at the sky or the sea surface around the escort, then an intermittent return of the u-boat would occurr, usually with some amount of clutter from the sea surface. The amplitude of the spike in the return would be a function of the strength of the return, with no range indication, other than that amplitude. So, a surfaced u-boat side-on to the radar-head, close to the escort would be a large amplitude, but one at AOB zero would be less in amplitude, and a distant one still less. That of a schnorkel viewed on the radar at range would be very difficult to make out amongst all the other spurious returns by a player.

Conclusion 2 is that player-skill on the escort should be a major factor as to whether - or not - a correct interpretation is made from an early iteration of radar.

Later on in the time-line the PPI should replace the oscilloscope presentation, but according to the pitch and roll of the escort, the ability to spot a contact amongst other spurious returns from the radar-head being tilted too far down, fairly difficult, especially at range. Similarly. as the radar head faces the the convoy, many strong returns are caused, however a ship or u-boat BEHIND a ship generating a return nearest the escort, will create a blank area behind that return, in which a u-boat can remain undetectable by radar, assuming it is surfaced. This would make spotting a u-boat's return the other side of the convoy, very difficult indeed.

As the date becomes later war, more and more escorts should have radar. Radar was preferentially fitted to destroyers, then sloops, then eventually corvettes. The height of the radar-head raising the range considerably. Similarly, as the date becomes later, sets with PPI should become the norm. Failure rates - or periods where the radar has to be turned off, should reduce as the timeline continues. Counter-measures such as Metox should likewise alert the radio-operator that the u-boat is being struck by radar.

My estimate is that AI escorts should not be able to "spot" u-boats on radar unless the u-boat is both surfaced an close - under circa 6km for an extended period. If AI is able to use radar at all! A trail of returns caused by life-boats, carly-floats and suchlike would further complicate use.

01-31-24, 01:37 AM	#139
Fidd XO Join Date: Apr 2005 Location: Blighty! Posts: 439 Downloads: 10 Uploads: 0	115. Thoughts on escort radar implementation....... There are a number of players who are somewhat nervous at the prospect of the gameplay ramifications of escorts receiving radar. As I see it, this is something not appropriate for use by AI, but rather one better suited to human interpretation, and the full range of deficiencies early radar possessed. Early radar was tempramental, and when it did work, suffered problems in rougher sea-states where the escort was pitching and rolling, as any time the radar-head was tilted down, it caused back-scatter from the surface - effectively false returns, and viewed over a period of time, it meant a considerable skill was needed on the part of the operator, to spot an actual consistent return on a given bearing amongst the false returns, if the escort was in rougher water. It is forecastable that AI interpreting such returns is either going to be far too capable and recognising a return as a u-boat, or, will be hopeless at it. A happy medium in AI's ability in almost any situation is usually not realised in game. So, conclusion one is that escorts should be playable before radar is implemented, and it be usable only by human players. Early radar had an oscilloscope presentation, rather than the PPI (Plan position indicator) we know today. In other words as the radar-head rotated through a bearing, assuming the radar head was not pointing at the sky or the sea surface around the escort, then an intermittent return of the u-boat would occurr, usually with some amount of clutter from the sea surface. The amplitude of the spike in the return would be a function of the strength of the return, with no range indication, other than that amplitude. So, a surfaced u-boat side-on to the radar-head, close to the escort would be a large amplitude, but one at AOB zero would be less in amplitude, and a distant one still less. That of a schnorkel viewed on the radar at range would be very difficult to make out amongst all the other spurious returns by a player. Conclusion 2 is that player-skill on the escort should be a major factor as to whether - or not - a correct interpretation is made from an early iteration of radar. Later on in the time-line the PPI should replace the oscilloscope presentation, but according to the pitch and roll of the escort, the ability to spot a contact amongst other spurious returns from the radar-head being tilted too far down, fairly difficult, especially at range. Similarly. as the radar head faces the the convoy, many strong returns are caused, however a ship or u-boat BEHIND a ship generating a return nearest the escort, will create a blank area behind that return, in which a u-boat can remain undetectable by radar, assuming it is surfaced. This would make spotting a u-boat's return the other side of the convoy, very difficult indeed. As the date becomes later war, more and more escorts should have radar. Radar was preferentially fitted to destroyers, then sloops, then eventually corvettes. The height of the radar-head raising the range considerably. Similarly, as the date becomes later, sets with PPI should become the norm. Failure rates - or periods where the radar has to be turned off, should reduce as the timeline continues. Counter-measures such as Metox should likewise alert the radio-operator that the u-boat is being struck by radar. My estimate is that AI escorts should not be able to "spot" u-boats on radar unless the u-boat is both surfaced an close - under circa 6km for an extended period. If AI is able to use radar at all! A trail of returns caused by life-boats, carly-floats and suchlike would further complicate use. Last edited by Fidd; 02-01-24 at 05:20 AM.