There isn't necessarily any need for entangled particles to communicate at all, that requirement stems from a way of thinking called "local realism" (that each particle has its own attributes that it carries around with it, and require some kind of exchange of communication to respond to the attributes of other particles). We use local realism in classical situations because it seems to work, not because it has to work. In the quantum domain, one simple resolution of the entanglement quandary is simply to drop local realism. Then there's no need for "communication" between the particles at all, the system is thought of a single thing.
Incidentally, there is plenty of precedence for having to think of systems holistically, without communication between particles. A stark example is a white dwarf star, in which the structure of some 10^57 electrons is governed by the Pauli exclusion principle. This says no two electrons are allowed in the same state, and this is essentially an entanglement phenomenon. What "communication" allows fermions to obey the PEP? Probably none-- it's likely a holistic feature of the system, not requiring communication between distinguishable parts because the whole reason you have a PEP is that the parts are indistinguishable. Not all entanglement is between indistinguishable particles, but if you have one type of entanglement for which communication would not make much sense, it isn't a stretch to extend that to all forms of entanglement.
