In this paper, we study two important transmission strategies in full-duplex two-way relaying in the presence of channel estimation errors. In analog network coding (ANC), the relay transmits the combined signals that were received from both sources, with the aim of achieving better spectral efficiency. However, due to imperfect channel-state information (CSI), sources cannot perfectly cancel their own data in the relayed signal. We derive an achievable information rate for ANC in imperfect-CSI conditions and show how the ANC performance can significantly be degraded as a result. Moreover, we derive cut-set bounds with channel estimation errors for traditional routing (TR), in which time sharing is used at the relay. Although it has been previously shown that ANC outperforms TR when the CSI is perfect, we find that it may not maintain its superiority in imperfect-CSI case at low signal-to-noise ratio (SNR) conditions. Next, we propose practical power allocation techniques that can be used in the sources and relay for both ANC and TR. The proposed power allocation schemes are relatively simple to compute and rely only on long-term channel statistics. Nevertheless, they are shown to be effective and close to optimal solutions for a wide range of SNRs, to different positions of the relay, and for both perfect-and imperfect-CSI conditions. By using the proposed power allocation techniques, it is possible to bring back advantages of ANC over TR for a wide range of SNRs in imperfect-CSI conditions.
Consistent growth in the volume and dynamic behavior of traffic mandates new requirements for fast and adaptive resource allocation in metro networks. We propose a dynamic resource allocation technique for adaptive minimization of spectrum usage in metro elastic optical networks. We consider optical transmission as a service specified by its bandwidth profile parameters, which are minimum, average, and maximum required transmission rates. To consider random traffic events, we use a stochastic optimization technique to develop a novel formulation for dynamic resource allocation in which service level specifications and network stability constraints are addressed. Next, we employ the elegant theory of Lyapunov optimization to solve the stochastic optimization problem and derive a fast integer linear program, which is periodically solved to create an adaptation between available resources and dynamic network state. To quantize the performance of the proposed technique, we report its spectral efficiency as a function of peak to average traffic ratio and Lyapunov penalty coefficient. Simulation results show that the dynamic resource allocation procedure can improve spectral efficiency by a factor of 3.3 for a peak to average traffic ratio of 1.37 and a Lyapunov penalty coefficient of 10 3 in comparison with fixed network planning. There is also a trade-off between transmission delay and spectrum utilization in the proposed technique, which can be adjusted by a Lyapunov penalty coefficient.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.