Remora Walruses Optimization Based Routing and Lung Cancer Disease Detection in the IoT Platform

With the development of cancer detection technologies, medical image analysis with image processing assists in diagnosing cancer even better. Most of the previously existing methods transmit a large quantity of information and it consumes more time for detecting the disease. In this paper, a newly developed optimization-based routing and detection of lung cancer disease in the Internet of Things (IoT) is demonstrated by reducing the time complexity. Here, the nodes in IoT are simulated and this node collects patient’s data. For transmitting the collected data to the base station and the optimal path is determined by using the proposed Walruses Remora Optimization Algorithm (WaROA) with measures like energy, delay, distance, and link quality. This approach is the integration of two optimizations, such as the Walruses Optimization Algorithm (WaOA) and Remora Optimization Algorithm (ROA). The nature of the WaOA approach mimics the walrus behavior and it is a new bio-inspired metaheuristic technique. The essential motivations engaged in the method of WaOA are the course of migrating, feeding, fighting, and escaping predators. The ROA approach is based on nature-inspired, new bionics and meta-heuristic. Due to the parasitic activities of remora, the ROA approach is motivated. By combining both approaches of WaOA and ROA, the best path is chosen and the performance of routing is enhanced. Then, the lung cancer detection at Base Station (BS) is carried out with the use of Computed Tomography (CT) images. The pre-processing is done by utilizing a Mean filter, which is used to eliminate the noise from the images. The lung lobe is segmented using Ethernet (E-Net) in which the features are, Gabor, Local Vector Pattern (LVP) and statistical features are extracted. Finally, lung cancer detection takes place with the help of SqueezeNet, where the weight of SqueezeNet is trained by the proposed WaROA. Finally, the WaROA approach for routing is compared with other existing techniques in terms of energy and delay, which is considered to obtain a maximum energy is 0.047J and a minimum delay is 0.107 ms in 50 nodes based on the Lung Image Dataset. Furthermore, the detection of lung cancer using WaROA-SqueezeNet is assessed by the metrics of accuracy, sensitivity, and specificity in which it attains the values of 93%, 93.5%, and 93.4% in 100 nodes based on the Lung Image Dataset.