Many infrastructure industries have made the transition from private ownership to utility, often through intermediate stages. It is believed that computing is in the midst of this transition today. The PIs believe that wireless-based systems will follow the same route in time. This path promises to alleviate the spectrum bottlenecks that plague emergent wireless applications. For this to occur, technological advances must take place that elicit cooperative behavior from firms that also compete with each other. These advances must leverage advances in radio system design, including cognitive radios. Relying heavily on system design methodologies and agent-based simulation modeling, this research explores the relationships between cooperation, competition and the implementation mechanisms, including system architectures and protocol design, especially where agents optimize locally. Particular attention is paid to the scalability of different approaches to large numbers of devices and heterogeneous uses. In the application to wireless systems, the research pays particular attention to ways in which transmission opportunities (called spectrum holes) may be created. These outcomes will enable easier access to wireless connectivity to emergent applications, thus enabling application designers to build services with explicit quality expectations for their users. Realistic models are constructed so that the mechanisms for spectrum hole generation and their use through protocols can be evaluated. The results will be published in major communications conferences and journals, and the development of pilot applications will be encouraged.
The main focus of this research is on the general problems that emerge when radios must cooperate with each other while competing for a common resource in cognitive radio networks. These problems become more challenging with the recognition that a hierarchy of rights exists and different rights may be attached to different radios. License holders have usage rights that are superior to opportunistic users. Usage opportunities of radios with subordinate rights are sometimes referred to as "spectrum holes" and can occur in time, space and/or frequency. Thus, it is reasonable to imagine that the existence and persistence of spectrum holes may depend on the cooperation of radios with superior usage rights; these radios may benefit from resource sharing when their own usage rights are not degraded. The main goals of the project, then, were to examine how this more complex spectrum hole ecosystem could be exploited. Doing so required an understanding that users with superior rights might change their behavior upon learning that users with inferior rights are making opportunistic use of spectrum holes. We refer to this recognition as endogenous spectrum holes, because the spectrum hole statistics must include explicit actions and reactions on the part of all users. Exogenous spectrum hole statistics can be quantified and compared to the endogenous spectrum hole statistics. Mechanisms can then be developed for explicit cooperation, such as spectrum markets or tradeable interference rights, and model the subsequent behavior. The research agenda investigated three principal research thrusts. The first thrust focused on collaboration and competition in cognitive radio networks. As part of this effort, we investigated novel spectrum trading markets to encourage cooperation among primary and secondary users and allow competition among secondary users to enhanced spectrum efficiency. This is achieved through (i) incentives to primary users to "create" spectrum holes, (ii) markets to allow primary users to buy back leased spectrum and compete for shared bands, and (iii) penalties to deter primary users from "forcing" secondary users to vacate their current operating channels. Our analysis of the cooperative, competitive spectrum market shows that the creation of endogenous spectrum holes incentivizes spectrum trading and increases spectrum utilization; it offers PUs the opportunity to explore tradeoffs between spectrum leasing and spectrum usage to increase their spectrum utility, while mitigating the risk of trading spectrum. The second investigated barriers to adoption of DSA techniques in practical systems. To this end, we developed a first order model of the secondary user’s decision of DSA choices, taking into consideration the risks these choices entail. The focus is on spectrum entrants who intend to operate a direct infrastructure-based system in one of four ways. The results show that exclusive use produces the highest NPV, while cooperative sharing ranks as the second best choice. Due to the level of uncertainty it entails, opportunistic sharing is the third best choice. The results also show that unlicensed spectrum sharing is generally not profitable. The third thrust focused on spectrum rights and enforcement and on tradeable interference rights. In this research effort, we illustrate how tradeable interference rights could work and developed some cases to illustrate how this can be achieved. While there is still considerable detailed work to be done before systems could be implemented using this technology, the development of the concept is a major step forward, since it allows primary users (i.e., users with superior access rights) to control the manner and extent of sharing with secondary users (i.e., users with inferior access rights). A number of papers detailings the outcomes and contributions of this research projects are published in major conferences and journal . The project has also produced a Ph.D. dissertation, which explores spectrum sharing choices and provides new insights on risk and decision analysis of spectrum usage. The project recruited and trained students, providing them with a forum and a unique opportunity to gain better understanding of the emerging cognitive radio technology, interact and exchange ideas with members of the research community and participate in major conferences related to the field of cognitive radio networks, where they presented their research findings.
