Department of Information and Communication Engineering, Yeungnam University, Gyeongsan 38541, Korea.
Department of Electronics Engineering, Yeungnam University, Gyeongsan 38541, Korea.
Sensors (Basel). 2019 Apr 10;19(7):1703. doi: 10.3390/s19071703.
The widespread growth of the Internet-of-Things (IoT) and its dependence on the license-exempt Industrial, Scientific, and Medical (ISM) bands have made spectrum resources scarce. IoT can nonetheless get advantage from the Cognitive Radio (CR) technology to resolve the spectrum shortage issue. Since in CR networks the unlicensed Secondary Users (SUs) can exploit the white spaces in licensed channels of Primary Users (PUs) opportunistically. CR ad hoc networks are more useful in IoT due to ease of installation, low cost, and less complexity. However, CR ad hoc networks are prone to the rendezvous issue and hidden primary terminal problem. Moreover, the available channels in the CR system are not identical, PUs' and SUs' activities can diversify them as well. In this connection, channel selection by SUs is a complex balancing act since the transmission opportunities are space, frequency and time bounded. In this paper, we hence proposed a new Ranked Sense Multiple Access with Collision Avoidance (RSMA/CA) protocol for multichannel CR-based IoT networks. Our proposed RSMA/CA protocol not only resolves the hidden primary terminal problem but also avoids hidden and exposed terminal problems at the same time by mutual spectrum sensing. We suggest a new channel ranking mechanism to rank the available channels based on the long term qualities of the channels, PUs' return rate, and SUs' activities and tailor-made the algorithms in an existing scheme to make the rendezvous process more efficient. We analyze the performance of our proposed RSMA/CA in terms of normalized throughput through the Markov chain model and compared with that of the existing scheme. Simulation results show that our RSMA/CA protocol outperforms the existing scheme due to efficient rendezvous and access mechanisms.
物联网（IoT）的广泛发展及其对免授权工业、科学和医疗（ISM）频段的依赖使得频谱资源变得稀缺。然而，物联网可以利用认知无线电（CR）技术来解决频谱短缺问题。由于在 CR 网络中，未经授权的次用户（SUs）可以机会性地利用主用户（PUs）授权信道中的空闲频谱。CR 自组织网络在物联网中更有用，因为它们易于安装、成本低且复杂性低。然而，CR 自组织网络容易出现 rendezvous 问题和隐藏的主终端问题。此外，CR 系统中的可用信道并不相同，PUs 和 SUs 的活动也会使它们多样化。在这种情况下，SUs 的信道选择是一个复杂的平衡行为，因为传输机会是空间、频率和时间受限的。在本文中，我们提出了一种新的基于排序感知的多址接入与碰撞避免（RSMA/CA）协议，用于多信道基于 CR 的物联网网络。我们提出的 RSMA/CA 协议不仅解决了隐藏的主终端问题，而且通过互感避免了隐藏和暴露的终端问题。我们提出了一种新的信道排序机制，根据信道的长期质量、PUs 的返回率和 SUs 的活动对可用信道进行排序，并对现有方案中的算法进行定制，以使 rendezvous 过程更加高效。我们通过马尔可夫链模型分析了我们提出的 RSMA/CA 在归一化吞吐量方面的性能，并与现有方案进行了比较。仿真结果表明，由于高效的 rendezvous 和接入机制，我们提出的 RSMA/CA 协议优于现有方案。
