Due to the emergence of ever increasing diversified applications provided by the smart devices such as smart phones, traditional telecommunications systems such as the wireless cellular systems no longer meet the ever exploding traffic demand, and cannot effectively deal with the shortage of available spectrum or congestion over wireless systems. On the other hand, tremendous temporal and spatial network resources, such as spectrum and computational capability, are severely under-utilized. Obviously, how to proactively harvest such residual resources and utilize them opportunistically to support diversified user traffic is an important yet challenging research direction. Although cognitive radio networks are to address this pressing issue, there is lack of viable network architecture in taking full advantage of the opportunistic spectrum access and there exist many practical design issues to be resolved. In this project, the PIs propose a flexible Cognitive Capacity Harvesting (CCH) network architecture to intelligently harvest network resources in both time and space and develop the corresponding technologies to support users' services effectively. Moreover, the CCH network along with the newly developed networking technologies can enable non-cognitive devices to significantly gain benefits from cognitive radio networks and provides innovative approaches to the cognitive radio networks design. Furthermore, this project research opens a new school of thoughts in better utilizing the residual network resources and potentially changes the design approaches for next-generation telecommunications systems. Finally, this project involves the international partners and can enhance the international components in the educational program and prepare more competitive workforce.
Due to the surge of mobile applications provided by the smart devices such as smart phones, mobile traffic has been exploding at an exponential rate and traditional telecommunications systems such as the wireless cellular systems no longer meet such traffic demands, and hence cannot effectively handle the congestion problem effectively over wireless systems. By dissecting such mobile traffic, it has been observed that traditional real-time traffic such as voice traffic remains relatively steady, but exponential traffic growth comes from the ever increasing data traffic, which tends to be delay-tolerant. From Shannon’s capacity theorem, the allocated cellular spectrum, no matter how much is allocated, will soon or later run short to handle such a traffic growth, and obviously extra spectrum will be needed. How to find additional spectrum to solve the spectrum shortage will be a major challenge. On the other hand, experiments by FCC concluded that tremendous temporal and spatial spectrum resources (either licensed or unlicensed spectrum bands) are severely under-utilized, and thus it will be a good idea to proactively harvest such residual spectrum resources and utilize them opportunistically to support such mobile traffic. Moreover, there are plenty of unlicensed bands from 30 GHz to 300 GHz which have not been efficiently utilized and have been intensively investigated for LTE use (LTE-Unlicensed standard). Although cognitive radio networks (CRNs) are designed to harvest idle licensed spectrum, currently there is lack of viable network architecture to take full advantage of the opportunistic spectrum access and there still exist many practical design challenges to be resolved. In this project, we have proposed a novel flexible Cognitive Capacity Harvesting (CCH) network architecture to intelligently harvest spectrum resources in both time and space and develop corresponding technologies to handle ever increasing delay-tolerant data traffic effectively. In the proposed architecture, new entities called cognitive radio routers (CR-routers) are deployed to form short-range backhaul networks, which can be used to collect information on both mobile traffic demands and (licensed or unlicensed) spectrum availability in both space and time, be self-organized or reconfigured to transport delay-tolerant mobile traffic around to be closer to the end users for high efficient transmissions while lowering interference and conserving energy. Various stochastic optimization problems have been formulated and solutions have been found to best utilize the random available spectrum resource to handle variable mobile traffic. More importantly, the proposed CCH network together with the newly developed cognitive radio technologies can enable non-cognitive devices such as normal smartphones without cognitive radios to access CCH and significantly gain benefits. This project research opens a new school of thoughts in utilizing the residual network resources and potentially changes the design approaches for next-generation telecommunications systems. This research project effort has yielded tremendous impacts in many aspects. First, the proposed network architecture was originally proposed to handle more efficiently the time-varying mobile traffic with uncertain spectrum availability and it relies more on collaboration among CR routers to smooth out the random variations of traffic and uncertain spectrum availability. It turns out that the proposed collaborative architecture may provide the technology for future cellular systems where collaborative small base stations can work together for high rate transmissions using emerging technologies such as (massive) MIMO. Besides, LTE-Unlicensed standard targets at the efficient use of unlicensed bands (e.g., mmWave) for short-range transmissions and our developed network technologies in harvesting spectrum resource will play an important role. Thus, the project research will have tremendous impact on the development of future generation telecommunications systems. Second, based on our proposed network architecture with more target-oriented mobile services, we formulate appropriate stochastic optimization problems and have successfully found solutions by developing efficient approximation algorithms. This has significantly advanced the state of the art in network research. Third, this project is an international collaboration among researchers, and the research efforts tap all brains from different cultures and efficiently utilize the crowdsourcing techniques by adding more diversity in the think tank, which has really boosted the productivity of the project. Fourth, the project team mixes the junior faculty and senior faculty, which provides a platform for junior faculty to grow. It is obvious that the junior faculty (Dr. Pan Li) has grown into a well-respected excellent researcher who has developed a strong research program in his institution. Last, but probably the most important, this project has trained a group of excellent graduate students. PI Fang has sent three PhD students to academia while PI Li has sent two PhD students to academia, including one minority student, who have been partially supported by this project. The project team consisted of multiple graduate students including three minority students, who have joined or will join the work force in telecommunications industries.
