Three-Dimensional Resource Matching for Internet of Things Underlaying Cognitive Capacity Harvesting Networks.

Baoshan Lu,Shijun Lin,Jianghong Shi

DOI: https://doi.org/10.1109/tccn.2022.3151940

2022-01-01

Abstract:In this paper, we propose a cognitive capacity harvesting network (CCHN) based Internet of Things (IoT) architecture, which allows the lightweight IoT devices without spectrum monitoring/sensing capabilities to enjoy the benefits of cognitive radio networks (CRNs). We investigate the sum-rate maximization of IoT links in this proposed architecture. In particular, we formulate the considered problem as a three-dimensional (3-D) resource matching between the IoT links, the CR links and the available CR spectrum blocks (CSBs). Then, two approaches, i.e., Hungarian based switching iteration (HBSI) approach and minimum interference clustering based Lagrange relaxation (MICBLR) approach, are proposed to obtain the near-optimal solution. In HBSI approach, the IoT and CR links are divided into a set of IoT and CR links clusters (ICCs). Based on the partition of ICCs, the considered problem can be simplified to a maximum weight bipartite-matching problem and solved by the Hungarian algorithm. Switching iteration is then used to improve the partition of ICCs. To achieve a better tradeoff between the performance and running time, we further propose the MICBLR approach, which contains IoT links clustering according to the minimum interference rule and a Lagrange relaxation (LR) algorithm used to solve the 3-D matching problem between the clusters of IoT links, the CR links, and the available CSBs. Simulations show that the performance of the proposed approaches is close to the exhaustive search (ES) method but with a much shorter running time. Compared with the Nearest sharing based Hungarian (NSBH), Furthest sharing based Hungarian (FSBH), and Random allocation (RA) policies, the proposed approaches can averagely improve the system performance by 33.68%-38.18%.