The development of 6G has enhanced the communication capabilities of Internet of Things (IoT) devices. However, in wireless system, due to cost constraints, only a few devices have the communication capability of 6G, and the rest can only communicate with the Internet through self-organized networks. Due to simple hardware, these IoT devices have low capacities of communication, computing and caching. And they are vulnerable to all kinds of attacks. One of harmful attacks is traffic-based attack such as on–off attack, Denial of Service (DoS) attack which consumes the limited energy of IoT devices and wreaks havoc on data-based applications. However, there is no effective way to obtain the truth traffic of IoT devices, which makes it difficult for the cloud to secure the communication of IoT devices and manage the state of network. To ensure reliable communication, a novel approach to detect traffic-based attack by Unmanned Aerial Vehicle (UAV) and Wavelet Packet Energy Entropy (UWPEE) is proposed. In UWPEE scheme, UAV is sent to collect the truth traffic from IoT devices. Then wavelet packet energy entropy is innovatively adopted to detect attacks. Finally, the trust of IoT devices is determined according to their entropy. The experimental results show that UWPEE scheme can effectively identify traffic-based attacks with an accuracy rate of 84.47% and an average recognition efficiency of 4.89 for malicious nodes. Meanwhile, compared with the greedy algorithm, the flight path of the UAVs is reduced by 15.44%.

6G的发展增强了物联网（IoT）设备的通信能力。但在无线系统中，由于成本限制，只有少数设备具备6G的通信能力，其余的只能通过自组织网络与互联网通信。由于硬件简单，这些物联网设备的通信、计算和缓存能力较低。他们很容易受到各种攻击。一种有害的攻击是基于流量的攻击，例如开关攻击、拒绝服务 (DoS) 攻击，它消耗物联网设备的有限能量并对基于数据的应用程序造成严重破坏。然而，目前还没有有效的方式来获取物联网设备的真实流量，这使得云端难以保护物联网设备的通信安全和管理网络状态。为了确保可靠的通信，提出了一种利用无人机（UAV）和小波包能量熵（UWPEE）检测基于流量的攻击的新方法。在 UWPEE 方案中，发送无人机以收集来自物联网设备的真实流量。然后创新性地采用小波包能量熵来检测攻击。最后，物联网设备的信任是根据它们的熵来确定的。实验结果表明，UWPEE方案能够有效识别基于流量的攻击，准确率为84.47%，对恶意节点的平均识别效率为4.89。同时，与贪心算法相比，无人机的飞行路径减少了15.44%。发送无人机以收集来自物联网设备的真实流量。然后创新性地采用小波包能量熵来检测攻击。最后，物联网设备的信任是根据它们的熵来确定的。实验结果表明，UWPEE方案能够有效识别基于流量的攻击，准确率为84.47%，对恶意节点的平均识别效率为4.89。同时，与贪心算法相比，无人机的飞行路径减少了15.44%。发送无人机以收集来自物联网设备的真实流量。然后创新性地采用小波包能量熵来检测攻击。最后，物联网设备的信任是根据它们的熵来确定的。实验结果表明，UWPEE方案能够有效识别基于流量的攻击，准确率为84.47%，对恶意节点的平均识别效率为4.89。同时，与贪心算法相比，无人机的飞行路径减少了15.44%。47%，恶意节点的平均识别效率为 4.89。同时，与贪心算法相比，无人机的飞行路径减少了15.44%。47%，恶意节点的平均识别效率为 4.89。同时，与贪心算法相比，无人机的飞行路径减少了15.44%。