QoS-Aware Channel Selection in Cognitive Radio Networks: A Game-Theoretic Approach

Application-aware dynamic spectrum access

“…Certain recent works attempt to account for QoS and context-awareness in DSA [10,12,13] using either Reinforcement Learning or Game Theory. However, the QoS considered in these proposals is not the high level application QoS metrics, but link level metrics like SNR and BER [10] or number of successful link layer transmissions [13].…”

“…Consequently, CASA is able to provide effective, fine-grained, and dynamic adaptation to application QoS demands during DSA operation. Apart from narrow QoS context, the adaptive responses in [10,12,13,54] are also limited and just include channel selection and spectrum sharing schemes-typically for every packet transmission which incurs very high overhead. No adaptation of DSA's fundamental parameters (like sensing duration, etc) is considered.…”

“…While there have been recent work in accounting for QoS in DSA [10,12,13], only low level QoS metrics have been considered, e.g., link SNR and BER. Thus, they are unable to accurately capture specific application-layer QoS demands which have high-level semantics.…”

Optimization of a QoS-aware channel assignment for cognitive radio networks

“…A channel assignment scheme is proposed in [5], and evolutionary algorithms are used to solve the centralized allocation problem. Because of existing SUs with heterogeneous QoS requirements in CRNs, each channel should be assigned to a set of proper SUs according to their QoS requirements [8,9]. So in this work the centralized QoS-Aware channel assignment considered as an optimization problem subject to different SUs demands and spectrum levels of QoS channel availabilities for each SU.…”

Price War with Partial Spectrum Sharing for Competitive Wireless Service Providers

“…It is now proposed to share at least a part of the spectrum (unlicensed bands) to cope with that situation and provide a better average QoS [8]. That kind of principle somewhat also justifies the development of cognitive networks where end-to-end performance is dynamically optimized by providers sensing and opportunistically accessing the underutilized spectrum [9], [10].…”