Who has the advantage here? In general whar will perform better in this scenario- 3 pawns vs. 3 pawns, or 4 pawns vs. 2 pawns?
Black is better here, but this is much more due to piece activity and coordination than to the pawn structure.
Once upon a time (i.e. a century ago), it was thought that such structures gave a big advantage in endgames for the camp with 3 pawns on the queenside - because in the long term they can create an outer passed pawn. However, this has been debunked in practice since the times of Alekhine (who would enter such endgames on either side as long as he had more activity) and in theory at least since the works of Shereshevsky (who notably analysed endgames by Capablanca and Alekhine). Many examples in (online available) chapter V of this book by Srodovsky. In a 2001 book, Glenn Flear showed that even in pawn endgames the queenside argument is not void but generally not the most important.
Here Black controls the only open file, enjoys centralised and coordinated minor pieces, is on move and has no pawn weaknesses while a2, f2 and e4 might become useful targets. wNa3 is offside, wBf3 is unprotected and restricted by its own pawn on e4, and white rooks are undeveloped.
After, say 1...Rd8 2.Rad1 Ne5 3.Be2 (or 3.Rxd6 Nxf3+ 4.Kg2 Nh4+ 5.gh4 Rxd6 -/+) Rd2, Black is clearly for choice. 1...Rd2 2.Rfd1 Ne5 3.Bg2 Rfd8 is similar.
Assuming black to move, black is definitely better, but he still must make sure that his Be6 cannot be driven to the unfavorable c8 square, or that too many trades occur along the d-file. I concur with Evargalo that the slight pull is primarily due to activity, and due to control of the only open file. In addition, white's queenside can be attacked soon.
In addition, white's pieces do not have the best squares, especially the Na3, which even given a few reasonable moves, like Nc3-e3, it still has not improved much. The best square would be c5, or e4 after e5, but neither will happen anytime soon. In addition the black N will soon get to b4, and a2 will be very touchy, or have to move, weakening the q-side that much more.
The best way to make use of black' slight advantage is: 1...Ne5; 2.Bg2 f6! (this secures the B retreat to f7, which will keep the Na3 off c4. Normally, white would not want to play f4-f5 there, but if it secured the c4 square for the N, and drove the B back to c8, then the trade off might be OK. That would be a similar idea to the famous Lasker-Capablanca, St. Petersburg 1914 game.); 3.f4 (or 3.Rfd1 Rfd8; 4.Rd6 Rd6 and Rd2 will be unavoidable) 3...Nd3 (and this is a bit akin to the famous Karpov-Unzicker, Nice 1974 game where Karpov blocked the a-file with Ba7, then tripled on the file, and although he never actually moved the Ba7, had he done so, he would have had control of the file); 4.Rad1! (White will defend a2 with Rf2, and try to keep the black rook off the 7th.) Rfd8; 5.Nc2 a5, and black's pressure is growing. He will continue with Nc5 and a4.
I like Evargalo's answer although I think the position is closer to equal than he credits.
Shereshevsky's book is great for late middle-game to early endgame study.
Concerning the opposing pawn majorities, it is easier to play the endgame with the side having the majority furthest from the kings. As play usually has both sides castling king-side, the queen-side majority is desirable. Your hope is to get an "outside passed pawn" to distract the king away. But, as Evargalo points out, it's the pieces that decide what is going on. Better would be to consider such positions with king an pawns only.
Black is slightly better though a draw is still the most likely outcome. White's a3-knight is poor, and Black has the superior bishop (White's f3-bishop is somewhat "bad" due to the e4-pawn). Also, the 3v2 queenside majority is better than White's central majority in this specific case. The reason is that Black can more easily create an outside passed pawn, while I don't see White starting anything soon in the centre.
The position is roughly equal but I would give a slight edge to black based on piece placement and the queenside majority. All 4 of white's pieces seem to be misplaced where black is a little better coordinated.
You don't say whose move but black can play to control the d file with rooks or Nc5 attacking the e pawn and then look to create a passed pawn on the queenside. White is just responding to black. His knight isn't doing much. His bishop is vulnerable to Ne5. He needs to take control of the d-file but the e pawn could quickly become a target and if white loses that he's playing for a draw.