Possibly - but you'll need to pull the enemy down to the effective altitude of cheap drones. They are already doing that.
Deny the high sky, part 1: high altitude.
The preferable method is to use long-range, long-reach, very large SAM systems with 200-300km range, well in the rear. The extant systems of this type are huge, "4 missiles per semitrailer + separate RADAR trucks + separate command post" type deals; the missiles are 0.8m diameter and 10m long, basically flying telephone poles. This includes the Soviet S-300 and successor S-400 systems (no NATO equivalent), and warship-based S-300 and AEGIS/SM-2 system. These control the high sky, but cannot control the middle and lower sky because they are positioned well in the rear, and lower altitudes are below the system's horizon.
In the Ukraine theater, Ukraine and Russia both possess the S-300 truck mounted system, Russia has the evolved S-400, and there are no warships with such systems anymore.
The less preferable method is to exhaust your adversary's supply of precision guided bombs and missiles by making it necessary, either by absorbing the hits to valuable infrastructure, or by using disinformation tactics to make them waste precious precision-guided weapons on bogus targets. I'm figuring on quite a lot of the latter, given Ukraine's demonstrated history of doing a lot with a little. Once they are down to World War II tier dumb bombs, to hit a given target their bombers must come well below 20,000 feet (6000m).
Deny the high sky, part 2: medium altitude.
Next, the objective is to deny the medium altitudes below 20,000 feet (6000m). This can be done several ways. First, use medium SAMs such as Buk, Tor or Pantsir. By putting everything - RADAR, command and missiles on the same truck, there's a downside in missile size, which limits range to medium - but a massive gain in agility, so they can "keep moving" to avoid counterattack. As such, they can be much farther forward with less of a horizon.
Second, use MANPADs like Stinger or SA-7 which have far shorter range (under 10,000'/3000m altitude) but are yet more portable, and more importantly use infrared seekers, so anti-RADAR missiles don't work on them, and it won't light up the threat-warning receiver on aircraft. The aircrafts' only warning is seeing the puff of smoke. And flares don't work if you don't deploy them. The MANPAD shooter has an effective field-of-view of "the tree line", buildings or other obstacles, so their access to the sky is somewhat less than 180 degrees. As such, the aircraft flight path is in that arc only for a limited time. The higher the aircraft flies, the longer it is in that arc and thus subject to attack. This amount of time matters. because MANPAD shooters take a moment to grab the weapon and get ready.
An aircraft operating in these zones is now denied the lower "good bombing" altitudes of 300-20,000 feet (100-6000m). So to manage both threats, the aircraft must fly very low. It defeats the medium range SAMs by dropping into the RADAR "ground clutter", literally below the RADAR. It defeats MANPADS by shortening the time it is in the shooter's field of view to where there is not time to alert, get to the MANPAD, shoulder up, power up, aim and shoot. But, this has the effect of denying the enemy medium and low altitudes, and forces them to hug the tree line. If you don't want to die, you better like the taste of leaves in your teeth.
In the Ukraine theater, everybody has these systems. On March 29, 2022, Ukraine claims not to have held a clinic in Belogrod on penetrating a modern, networked SAM system.
To see these systems at play in simulation, with discussion, check out the "Grim Reapers team" video on Youtube where they attempted to "fact check" the Belgorod raid. The conclusion that if you fly "not perfect" the SAMs get you, but a more experienced player (who really liked the taste of leaves) made it in and out (even overflew a SAM site too low and close for it to engage).
Let's change the meaning of "drone", though.
I know "quadcopter" is synonymous with "drone". Forget that. Why?
- Any 'copter type aircraft is hideously fuel-inefficient compared to a fixed-wing because it's "suspending with thrust" instead of actually flying. For many uses, that's a worthy trade-off, i.e. the ability to hover. But the efficiency is bad - loitering surveillence aircraft would fly like gliders and probably look like gliders. Does the glider community have any rotary-wing gliders? No.
- The quad/multi copter needs many engines and thus electric drive, which locks it out of using fuel for longer duration flight. A fixed-wing can go either way. And R/C airplane sized engines are plentiful.
- Payload capacity is much higher on a fixed-wing, so a meaningful warhead or sensor package could be carried.
- Speed - which matters both for intercept, and a second thing.
The second thing is the ability to have a simple basic airframe with swappable mission modules. Some drones might carry a warhead. Others might carry a RADAR or other feature that needs a lot of electricity. Rather than complicate the basic drone design by adding a generator to some engines, do as the military does and build the module with a generator called a RAT (Ram Air Turbine). Instead of belt drive, airflow drive. Less to go wrong. But the RAT requires airspeed beyond what is possible for a quadcopter, but is easy for a fixed-wing. In fact, I would even prefer a RAT for the basic electrical needs of the drone. A RAT is just a generator attached to a prop, in fact every quadcopter comes with 4 of them lol. BLDC motors work as generators.
And they could fly high but they don't have to because the above measures have pressured the enemy into the narrow zone below 400' AGL. In fact, having the interceptors loiter high would give them the ability to dive to gain energy/speed.
Nothing against a quadcopter - they're a good fit for the task size and attention span of hobby and task-focused drone flyers. But a sustained air surveillance or combat air patrol needs much more endurance than that, so for your task I'd absolutely say fixed-wing is the way to go.
Now, network them.
The general idea I see is to have some drones tasked with surveillance and others tasked with intercept. Surveillance could be as easy as listening - using the sound of aircraft to locate them - the drones have accurate position sensing and accurate clocks, and report contacts to each other via radio. That makes it easy to triangulate targets. Digital signal processing of that sound could identify right down to model of aircraft. You can't outrun Motorola, so the intercept drones are getting a picture of the track of the enemy even while it's a long way off. They have time to position to "get in its way".
There would be technical tricks to resolve regarding sensors, geolocation (since the enemy can be counted on to mess with GPS/GLONASS), and telecommunication - you might actually have to read Hedy Lamarr's work on spread-spectrum... not using a single frequency subject to jamming or RDF. See also Qualcomm's work with CDMA cellular, or Metricom's lovely-for-the-time "Ricochet" - a wireless mesh network that did not need the backbone -- Ricochet modems could talk peer-peer. It might be as simple as 25-year-old Ricochet modems in very drone.
Another thing this network could do is dull the senses of pilots to MANPAD launches by having $20 decoy rockets all over the place designed to mimic a MANPAD launch. They would listen to the network and fake out just as the aircraft is overhead. At worst the pilots use up their flares on decoys so they're Winchester when they meet a real one. At best the pilots treat everything as a decoy and are blindsided by the real McCoy.
Once the hardware is in place, the performance of the system is mostly down to software.
So yes, possible. Very possible. Because you've reduced the battlespace to 2-D, or really a line of a particular width.