Not that these problems are impossible to solve, but they are nontrivial in practice, I'm afraid. It would take quite some effort to turn the fantasy into a reliable theory, and then quite some more, into a practical swarm.
Will be cool though.
What would be even more fun (i.e. stretching the limits) would be to include a 30-300 mhz radar in the mix, which would resolve just fine with dispersed nodes, and penetrate most stealth coatings and bounce off the airframes of the enemy.
Then, there should be a super-node in a larger aircraft that would receive the data from the swarm and turn it into a coherent picture, run object detection and tracking, etc.