Yeah, so there's no point in tilting a PESA radar, since the lack of Low Probability of Intercept ability pretty much already renders it much more observable. I've never seen a configuration where the PESA is tilted.
Incorrect...There is
NOTHING that said only an AESA system can perform low probability of intercept (LPI) mode. This is a gross misunderstanding.
Radar detection is essentially a stochastical process, fancy word for statistics, it mean the system must detect <something> consistently over an arbitrarily set of timespan before declaring that there is a 'valid' target. Most radar systems operate at what could be considered 'high probability of intercept', meaning the transmitted signals are above a certain level and with characteristics that are predictable.
What low probability of intercept (LPI) does is to alter those characteristics, such as power level or pulse repetition, to deny the receiver that statistical certainty that <something> is out there. Absent that statistical certainty, the system can only alert the operator that it suspect that <something> is present and that it effectively passed the decision ball to the humans to investigate and declare if that <something> is a true target. Not a 'valid' target but a true target. The system can make a 'valid' call but investigation can reveal that it is a 'false positive'. Hence it is important that we distinguish the differences between a 'valid' declaration versus a physically 'true' target. The more the human operator is involved handling that decision ball, be he a radar operator or a pilot, the more time is consumed in the decision making process.
That said...The 'low probability of intercept' (LPI) radar is more an operating mode than a hardware related issue. What it mean is that
ANY radar, in theory, can be an LPI radar. The misunderstanding come from the fact that a hardware/software combination make that operating mode more efficient, as in a thousand times more efficient than if the human operator can attempt.
A passive ESA system can be an LPI radar if the power and array software management is good enough. However, an active ESA system is a hundred times more efficient at being deceitful to any receiver that it practically make no sense, financially or otherwise, to install the LPI mode into a PESA system. A non-ESA system can also be LPI but it would be a hundred times less efficient than a PESA can. And the AESA is superior to all.
In certain situations, such as marine navigation through an EM dense environment such as population centers, the radar's signals can interfere with other navigation radars, or television, or communication radios, or even cellular towers' operations, so the word 'intercept' is sometimes replaced by some engineers to be 'interference', as in 'low probability of interference'. The difference here is the type of response by those who 'intercept' such signals. The military would respond harshly. The civilian sector would respond with annoyance that their music and television entertainment is being adversely affected. So the words 'intercept' and 'interference' is quite philosophically synonymous.
So to get back to the current topic, if the J-10B is discovered to benefit from a tilted PESA, as far as RCS reduction goes, then tilting the antenna would make practical sense.