Since the 2000s, a plethora of wireless technologies appeared to satisfy a continuously growing set of requirements, ranging from best effort to strict constraints on reliability or delays. It is very difficult for a service provider to choose the protocol suite tailored for one application among this variety of technologies and protocols. In addition, applications deployed over existing wireless networks can evolve or shift their requirements, making the protocol suite obsolete or requiring the replacement of the existing infrastructure with new devices to support new wireless communication technologies and protocols. This contributes to the premature obsolescence of electronic devices and increases the generation of e-waste. Last but not least, it is envisioned that the next generation of wireless standards will shift from static transceiver specifications towards end-to-end learning capable to adapt to its environment. In PERENNE, we propose to develop a fully programmable communication stack for wireless networks by leveraging the software-defined paradigm (namely, Software Defined Networking - SDN, and Software Defined Radio - SDR). We target the ultimate flexibility, from the reconfiguration of the protocol suite at runtime to the deployment of any future wireless technologies on the same infrastructure. The ultimate goal is to sustain the deployed infrastructure in the very long term, fighting against premature obsolescence without endangering innovation. We will transform each element of the communication stack into a computer program, i.e. a set of instructions in a specific language, that will be transmitted and executed by the end devices. Such an approach raises major scientific challenges such as securing the control plane, optimizing the system efficiency (in terms of energy consumption for example), or guaranteeing system convergence.