Exploiting Multi-Channel Diversity for Throughput Maximization


Project Description

Recently, spectral agile network has been recognized as a solution for spectrum scarcity problem in wireless communications. In spectral agile network, secondary (unlicensed) users can opportunistically utilize licensed spectrum resources to improve spectrum efficiency. We restrict our attention to secondary devices equipped with a single transceiver, and study the spectrum access characteristics. We derive the channel accessibility of secondary user from primary users’ traffic load and spectrum sensing overhead. In addition, since secondary user can utilize any idle channel, wireless fading effect can be mitigated by periodically switching to the best channel and thus maximize the throughput. To solve this problem, we propose channel-aware switching algorithm for throughput maximization. We assume Rayleigh fading and each channel is modeled as finite-state Markov channel (FSMC) model. Based on periodic measurement of the received SNR on each channel, the channel states are predicted using the FSMC model. The expected channel holding time is also considered in channel switching decision. Our numerical results show that maximum channel accessibility can be achieved by sensing the optimal subset of channels. We also show that our proposed scheme achieves significant performance improvement over random channel switching.