With the proliferation of Internet of Things (IoT) based applications in fields such as manufacturing, energy, transportation, healthcare, and emergency/disaster response, autonomous deployments of large scale IoT networks will ace the burden of hauling large volume of produced and consumed data. Since most of the IoT devices are expected to be connected wirelessly, there will be an unprecedented need for higher capacity wireless networks. Naturally, the current wireless networks that operate on the Industrial, Scientific and Medical (ISM) or licensed bands will fall short. As a remedy, dynamic spectrum access (DSA) and sharing have been proposed as a high-throughput, cost-effective solution for the growing demands [1]. Using DSA, unlicensed (i.e., secondary) IoT devices will opportunistically use the underutilized or unused channels for licensed or primary users/network (PU). In recent times, DSA based solutions are being proposed and pursued for the growing demands of commercial networks [2], smart cities with smart vehicular networks (VANET) [3], smart grids [4], and military communications [5], to name a few. Since the spectrum availability is space and time variant, selection of the best among the available channels between a given pair of IoT devices becomes very crucial. This is especially true for Device-to-Device (D2D) communications, such as [6,7] where data needs to be routed over multiple hops. Sophisticated multi-hop routing algorithms are needed that can find the best end-to-end route in terms of quality of service (QoS) metrics and the best channels at each hop. The primary challenges for such multi-hop D2D IoT communications in using DSA over dedicated spectrum are: (a) the need for spatio-temporal spectrum-awareness in terms of finding unused or underused channels at different locations along an end-to-end route, (b) protecting licensed primary transmission on channels when and where they arrive from harmful interference caused by secondary IoT communication, and (c) ensuring power controlled IoT communication to maintain the strict energy preservation requirements of the IoT devices. Due to these reasons, the state-ofthe-art Internet routing protocols, such as, LSR [8], and OSPF [9], or traditional wireless mesh network routing protocols, such as, DSR and AODV cannot be seamlessly applied to DSA based IoT communications. Fig. 1 shows one such scenario where an inefficient multi-channel route from source IoT device (D1) to the destination (D6) yields harmful interference to primary receiver (PRX ) when an alternate multi-channel route (D1 → D7 → D8 → D9 → D6) is available that ensures primary protection as well as uses low power transmission (leading to multiple hops) ensuring energy preservation.