Department of Communications and Computer Engineering, Faculty of Science and Engineering, Waseda University, Shinjuku-ku, Tokyo 169-0051, Japan.
Sensors (Basel). 2018 Aug 31;18(9):2889. doi: 10.3390/s18092889.
Information-Centric Networking (ICN) is a new Internet architecture design, which is considered as the global-scale Future Internet (FI) paradigm. Though ICN offers considerable benefits over the existing IP-based Internet architecture, its practical deployment in real life still has many challenges, especially in the case of high congestion and limited power in a sensor enabled-network for the Internet of Things (IoT) era. In this paper, we propose a smart congestion control mechanism to diminish the network congestion rate, reduce sensor power consumptions, and enhance the network performance of ICN at the same time to realize a complete green and efficient ICN-based sensor networking model. The proposed network system uses the chunk-by-chunk aggregated packets according to the content popularity to diminish the number of exchanged packets needed for data transmission. We also design the sensor power-based cache management strategy, and an adaptive Markov-based sensor scheduling policy with selective sensing algorithm to further maximize power savings for the sensors. The evaluation results using ndnSIM (a widely-used ICN simulator) show that the proposed model can provide higher network performance efficiency with lower energy consumption for the future Internet by achieving higher throughput with higher cache hit rate and lower Interest packet drop rate as we increase the number of IoT sensors in ICN.
信息中心网络（ICN）是一种新的互联网架构设计，被认为是全球范围的未来互联网（FI）范例。虽然 ICN 相对于现有的基于 IP 的互联网架构提供了相当大的优势，但在物联网（IoT）时代的传感器支持网络中，其实际部署仍然存在许多挑战，特别是在高拥塞和有限电源的情况下。在本文中，我们提出了一种智能拥塞控制机制，以降低网络拥塞率，减少传感器功耗，并同时提高 ICN 的网络性能，以实现完整的绿色高效基于 ICN 的传感器网络模型。所提出的网络系统根据内容流行度使用逐块聚合数据包，以减少数据传输所需交换的数据包数量。我们还设计了基于传感器功率的缓存管理策略，以及具有选择性传感算法的自适应马尔可夫传感器调度策略，以进一步为传感器最大程度地节省功率。使用 ndnSIM（一种广泛使用的 ICN 模拟器）进行的评估结果表明，随着 ICN 中 IoT 传感器数量的增加，通过实现更高的吞吐量、更高的缓存命中率和更低的兴趣包丢弃率，所提出的模型可以通过更高的网络性能效率和更低的能耗为未来互联网提供更高的网络性能效率。
