Resource-sharing cognitive radio networks in commercial and military environments
Date of Issue2017
School of Electrical and Electronic Engineering
Positioning and Wireless Technology Centre
Cognitive radio (CR) provides the platform for dynamic spectrum sensing, spectrum allocation and intuitive decision making in a very rigid spectrum framework architecture. Accessing white spaces in licensed spectrum through spectrum hopping and managing interference in free-to-use spectrum are means by which CR can help improve the overall spectrum efficiency. But utilizing these spectrum resources generally comes at a price, whether that be a monetary cost or interference to transmission. Resource-sharing therefore demands a two-way exchange for users to e ffectively transmit simultaneously across any spectrum. We investigate how this cost changes given the application of CRs in a commercial or military setting. With commercial or licensed spectrum, monetary incentives are important, and therefore spectrum leasing or spectrum trading is often sought as the avenue for resource-sharing. While the bulk of spectrum sharing research focuses on interference mitigation or rate maximization for the cognitive or secondary users (SU), the work in this Thesis deals with incentivizing the licensed or primary users (PU) that lease out their spectrum. The resource-sharing protocol incorporates three opportunities for spectrum exchange in the form of spectrum access, spectrum sharing and relaying, that each have a market-driven cost. With the highest SU bid gaining transmission rights to the leased PU spectrum, this model perfectly showcases the demand-and-supply economics involved in a commercial setting. From a military standpoint, time is a commodity that is precious. Time-efficient relays in particular, help play an important role in reestablishing non-LOS (line-of-sight) communications between military users stranded in the battlefield. With link quality-of-service (QOS) and cost measures to be considered as well, our design features a partial-relay node that serves to connect the stranded users while carrying out its own transmission needs. The relay node features a zero-forcing equalizer (ZFE) and space-time block coding (STBC) to mimic properties of physical-layer network coding in a practical sense. The result is a half-duplex communication setup that requires only 4 timeslots for three way non-LOS bi-directional communication, and bit error rates (BER) that are comparable to the direct link QOS. We also see a great demand for blind and decentralized decision making amongst resource-sharing setups both in the military and commercial communication domains. Incognizance or anonymity is the requisite for the SUs that sense and utilize the spatial-temporal gaps in the spectrum. Moreover, for a network-blind and decentralized SU, this translates to using only statistical sensing information to carry out all spectrum allocation and transmission decisions. When sensing licensed subspace, the optimization goal is outage minimization at the primary receiver (PR), while for transmission over unlicensed bands, the goal is to meet a network-wide QOS for SU transmission. The work discussed in this Thesis demonstrates how outage-constrained threshold design using energy detection (ED) and adaptive power control can result in both harmonious and anonymous spectrum sharing across both platforms. The results justify our contributions to resource-sharing research in the CR domain, and its general applicability to both commercial and military markets.
DRNTU::Engineering::Electrical and electronic engineering::Wireless communication systems