期刊
IEEE JOURNAL ON SELECTED AREAS IN COMMUNICATIONS
卷 31, 期 11, 页码 2465-2482出版社
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/JSAC.2013.131131
关键词
Cognitive Radio System; Game Theory; MIMO System; Successive Convex Approximation; Complex Variational Inequalities; Worst-Case Robust Design
资金
- Hong Kong RGC [617911]
- National Science Foundation [CNS-1218717]
- Division Of Computer and Network Systems
- Direct For Computer & Info Scie & Enginr [1218717] Funding Source: National Science Foundation
Cognitive Radio (CR) systems are built on the coexistence of primary users (PUs) and secondary users (SUs), the latter being allowed to share spectral resources with the PUs but under strict interference limitations. However, such limitations may easily be violated by SUs if perfect SU-to-PU channel state information (CSI) is not available at the secondary transmitters, which always happens in practice. In this paper, we propose a distributed design of MIMO CR networks under global interference temperature constraints that is robust (in the worst-case sense) against SU-to-PU channel uncertainties. More specifically, we consider two alternative formulations that are complementary to each other in terms of signaling and system performance, namely: a game-theoretical design and a social-oriented optimization. To study and solve the proposed formulations we hinge on the new theory of finite-dimensional variational inequalities (VI) in the complex domain and a novel parallel decomposition technique for nonconvex sum-utility problems with coupling constraints, respectively. A major contribution of this paper is to devise a new class of distributed best-response algorithms with provable convergence. The algorithms differ in computational complexity, convergence speed, communication overhead, and achievable performance; they are thus applicable to a variety of CR scenarios, either cooperative or non-cooperative, which allow the SUs to explore the trade-off between signaling and performance.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据