Secondary Users (SU) are guided by Cognitive Radio Device in identifying a channel licensed to Primary Users (PU) when it is free. Whenever PU arrives, SU shall vacate and search for the alternate band by a conditional handoff. Existing handoff methods reduce the network’s efficiency. The proposed stochastic approach under hybrid Spectrum Hand-off is (SHO) employing the Invasive Weed Optimization (IWO) algorithm thus increasing the SHO efficiency in CR Networks. The proposed Centralized Cognitive Device monitors load balancing, minimizes handoff delay as well as the SU’s service time, conserves energy, speeds up the data transmission, supports the Internet of Things, multimedia applications including audio, live video streaming and still images over resource constrained WSNs. This algorithm is compared with the existing Genetic algorithm and Particle Swarm Optimization. The channel selection accuracy of the proposed IWO method is found to be 97.8% and it outperforms conventional methods. Besides, the pre-emptive resume priority M/G/1 queuing model is also utilized. This paper presents a complete system model along with a detailed study of its parameters proving the technique’s effectiveness.

辅助用户（SU）由认知无线电设备指导，以识别免费的，授权给主要用户（PU）的频道。每当PU到达时，SU应退出并通过有条件的越区切换搜索备用频段。现有的切换方法降低了网络的效率。在混合频谱切换下，所提出的随机方法是采用侵入性杂草优化（IWO）算法，从而提高了CR网络中的SHO效率。拟议的集中式认知设备可监控负载平衡，最小化切换延迟以及SU的服务时间，节约能源，加快数据传输，支持物联网，多媒体应用程序（包括音频，实时视频流和资源受限的WSN上的静态图像） 。将该算法与现有的遗传算法和粒子群算法进行了比较。所提出的IWO方法的信道选择精度为97.8％，并且优于常规方法。此外，还采用了抢先恢复优先级M / G / 1排队模型。本文提出了一个完整的系统模型，并对其参数进行了详细研究，以证明该技术的有效性。