Due to enormous amount of data in emerging Internet of Things (IoT) applications, congestion control of traffic flow is very important to achieve certain level of security and Quality of Service (QoS). In the case of a congested network, the data is vulnerable of packet drop and loss, and so that the data integrity, which is a vital security issue, is degraded though it is accidental and not intentional. A Proportional Integrator Differentiator or (PID) controller is proposed in this paper and tuned with a congestion control rate based scheme for the purpose of IoT data collection. The proposed fine-tuning approach is created based on an optimization problem with an appropriate compact fine-tuning function that reflects the control requirements. In this paper, the key contribution is to implement a superior hybrid hill climbing immune algorithm to be as tuning method to secure IoT data flow. The proposed method is cheap, simple and fast. The proposed hybrid algorithm is compared with the immune-algorithm alone and proved its superiority in term of the overall tuning time. The proposed algorithm works in a cascaded way; the immune algorithm is used for rough tuning and then the hill climbing algorithm is activated for fine tuning. A straightforward network model is constructed and simulated for the goal of comparison using two control approaches, namely the single-bit indicator scheme and control scheme. The experimental results show that the proposed tuned PID controller by immune hill climbing algorithm is superior in terms of the stability of both, buffer occupancy in the switch and the source rate. The additional delay needs by securing the data is solved by the tuned PID controller. The congestion controller needs a minimum number of iterations in the tuning process to guarantee the stability and some target-accepted efficiency. The minimum number of iterations highly depends on the current forward and backward delays. It is shown that as the round-trip delay is increased, the tuning process takes longer before it reaches a stabilized and efficient controlled network. The steady-state response and the transient behavior for the PID controller are shown to be much better than the single-bit indicator scheme. Subsequently, the overall computational cost of the proposed hybrid algorithm is less than using the immune-algorithm. This is an additional feature for securing IoT data flow, which is an important contribution in this paper. The proposed scheme is better than the single-bit indicator in buffer utilization, link utilization, response to any suddenly change in the network and packet drop, which provides a highly secure system.

由于新兴物联网（IoT）应用程序中的数据量很大，因此，对流量进行拥塞控制对于实现一定级别的安全性和服务质量（QoS）非常重要。在网络拥塞的情况下，数据很容易出现丢包和丢失的情况，因此，至关重要的安全问题数据完整性虽然是偶然的，也并非故意的，但仍会降低。本文提出了一种比例积分微分器（PID）控制器，并根据基于拥塞控制速率的方案进行了调整，以用于IoT数据收集。所提出的微调方法是基于具有适当小巧微调功能的优化问题而创建的，该功能可以反映控制要求。在本文中，关键的贡献是实现了一种卓越的混合爬坡免疫算法，作为确保物联网数据流安全的调整方法。所提出的方法便宜，简单且快速。将提出的混合算法与单独的免疫算法进行比较，并证明了其在总调谐时间方面的优越性。该算法以级联方式工作。使用免疫算法进行粗调，然后激活爬山算法进行细调。为了比较的目的，使用两种控制方法，即单位指示符方案和控制方案，构建和仿真了一个简单的网络模型。实验结果表明，提出的基于免疫爬山算法的PID调节控制器在稳定性，开关占用率和信号源速率方面均具有优越性。调谐的PID控制器解决了通过保护数据来增加延迟的需求。拥塞控制器在调整过程中需要最少的迭代次数，以确保稳定性和某些目标可接受的效率。最小迭代次数高度取决于当前的前向和后向延迟。结果表明，随着往返延迟的增加，调谐过程需要更长时间才能到达稳定且有效的受控网络。PID控制器的稳态响应和瞬态行为显示出比单位指示符方案好得多。随后，所提出的混合算法的总体计算成本低于使用免疫算法的成本。这是确保物联网数据流安全的附加功能，这是本文的重要贡献。