Aircraft are the backbone of the US military's offense, with unmanned aircraft steadily moving into the lead role. Given US reliance on air power, it makes sense to think about better air defense. I will examine this in terms of radar, UAVs, and the pros and cons of AAA vs. SAMs. I will then describe a hypothetical plan of defense against a US bombing campaign.
First, a brief trip back to the first Gulf War. Iraqi air defense was defeated early on and very few US aircraft were lost. This was accomplished by a clever gambit. The first wave of aircraft were drone decoys whose purpose was to get the Iraqis to turn on all their radar systems. The second wave was strike aircraft that launched radar-homing missiles which destroyed most of the Iraqi radar. Without radar, the Iraqis could only fire blindly. The US quickly gained air supremacy.
The lesson here is that while radar is a key part of air defense, it is also its key weakness. Optical and infrared targeting is more difficult but it has the advantage of not being vulnerable to radar-homing weapons.
It would be interesting to see if a distributed radar system based on patrolling drones would be better than ground-based radar networks.
UAVs like the Predator drone are less expensive to operate and can loiter far longer than manned aircraft.
There's also no risk of losing the pilot. On the flip side, they are relatively easy to shoot down. 15 were destroyed by the Serbs by both SAMs and AAA during the 1999 war with NATO. Since they communicate by radio, they are also vulnerable to jamming, although there have been no records of that as yet. The satellites they rely on are also vulnerable to attack. The Chinese successfully destroyed a satellite with a missile in 2007. My conclusion is that drones are best used against enemies with minimal air defense, though even against strong defense they make good decoys.
Low and slow flying aircraft can be shot down with small arms, but for everything else, the only options are AAA, SAMs, and other aircraft. Ground-based weapons have historically shot down far more aircraft than fighters/interceptors, and are far cheaper. During the Vietnam War, almost all US air losses were from AAA, even though the SAMs and fighters also made kills. AAA is cheap and reliable, but less accurate and have less range. SAMs are more accurate and longer range, but are more expensive and can be foiled by chaff, flares, and ECM.
If I was tasked with designing the air defense of a small country, my strategy would revolve around, dispersion, deception & camouflage. I would construct many decoy targets to tempt the US into wasting bombs. The Serbs protected most of their armored vehicles from being destroyed by NATO through decoys. I would construct decoy air base, planes, SAM sites, etc. The fake defenses would be more numerous than the real ones.
Rather than concentrate defenses in vital areas, I'd spread them out, which would help confuse them about where the real targets are. The decoys would be defended as strongly as any other potential target. The US is very good at hitting what they want to hit, but they only have so many planes and so many bombs and most importantly, only so much time to win the war.
As for the actual weapons, I'd put a little of my money into MANPADs. I would also have a large store of unguided rocket like the old Z-battery to launch at aircraft to get them to waste their chaff and flares. The bulk would go to top-of-line missile systems like the Tor, the Buk, and the Pantsir.