In this research article, the advancements in arrhythmia analysis & detection with a comprehensive exploration of hybrid deep learning models integrated with farmland fertility algorithm for ECG signal segregation for different bio-medical engineering applications is presented along with simulation results carried out in the Python environment. We have summarized the deep learning-based research on detecting arrhythmia in the ECG signals. The research plan was to design an efficient & faster deep learning model for the classification of varying number of cardiac arrhythmias. The method was to obtain ECG signals which involved taking a multiple number of signals on the ECG tracings from a diverse class of arrhythmias & each beat in a variety of classes of arrhythmias 2D images obtained utilizing the Discrete Wavelet Transform. Finally, deep learning approaches using Genetic Algorithm (GA) based deep neural network for improved arrhythmia classification was introduced. The system that we have proposed in this paper is intended for real-time frameworks, which enables instantaneous evaluation of the conditions of a patient and immediate execution of therapy by a healthcare professional (say a doctor or a surgeon / consultant in the medical field). The research work undertaken gives a broad review of how deep learning techniques could be used in ECG signal arrhythmia detection in the human beings. Arrhythmia is the most common cardiac disorder that may lead to severe consequences in terms of health, so it is the interest of topic that is chosen for research & the outcome being presented here with results. In this context, the work finally closes by the development of a original hybrid deep learning model, called as, Automated Arrhythmia Classifications with Farmland based Fertility type of Algorithm based Hybridized Deep Learnings (AACFFAHDL) for real-time arrhythmia diagnosis in the Internet of Things (IoT) based framework and being introduced here. Simulation is carried out on a trained dataset-based validation on a benchmark ECG dataset which shows that the model achieves significantly better accuracy and F1-score compared to deep learning with hyperparameter tuning, as well as The Farmland Fertility Algorithm. The simulation results show the effectiveness of the methodology that is being presented in this research paper.