Abstract: Aiming at the problem of wireless resource scarcity and errors caused by imperfect channel in Internet of Things (IoT) sensor network, a method for allocating bandwidth and power to Secondary IoT Device (SID), under the condition of imperfect Channel State Information (CSI), was proposed. Firstly, considering the secondary and primary situations, the imperfect CSI of links between secondary transmitter side and primary receiving end, secondary transmitter side and secondary receiving end were modeled. Secondly, a mechanism for allocating bandwidth and power to secondary IoT devices was proposed. Meanwhile, the measurement and corresponding penalty policy were considered when current bandwidth is not enough to allocate. The idea is to maximize the secondary Energy Efficiency (EE) while fully using resources in extent. Finally, Particle Swarm Optimization (PSO) algorithm and weighted Chebyshev method were implemented to solve the problem, which made it possible to obtain the optimal allocation policy while lowering the problem complexity. The simulation results show that compared with equal allocation method and random allocation method, the proposed method improves the overall system transmission rate and minimizes the average secondary IoT base station transmit power by 75%, which can effectively enhance the energy efficiency of the entire network.

Key words: Internet of Things (IoT) sensor network, imperfect channel, bandwidth and power allocation, Particle Swarm Optimization (PSO) algorithm, Cognitive Radio Network (CRN), weighted Chebyshev method

摘要： 针对物联网（IoT）传感器网络无线资源的稀缺和信道不完美带来的误差问题，提出了一种在不完美信道状态信息（CSI）下为次级IoT设备（SID）用户分配带宽和功率的方法。首先，基于次级系统和主级系统的不同情况，分别建立了次级系统发送端与主级系统接收端、次级系统发送端与接收端之间的不完美信道模型。其次，提出了为次级IoT设备用户分配带宽和功率的机制，并考虑了当前带宽不够分配时应采取的措施和相应的惩罚机制，目的是在充分利用现有无线资源的情况下，最大化整个次级系统的能量效率（EE）。最后，使用粒子群优化（PSO）算法和加权切比雪夫法进行分步求解，在降低问题复杂度的同时得到最优的分配方案。仿真结果表明，与均等分配和随机分配方法相比，所提方法在最大化系统总发送速率和最小化次基站平均发送功率方面提升了约75%，能有效提升整个网络的能量效率。

关键词: 物联网传感器网络, 不完美信道, 功率和频谱分配, 粒子群优化算法, 认知无线电网络, 加权切比雪夫法