The COVID-19 emerged at the end of 2019 and has become a global pandemic. There are many methods for COVID-19 prediction using a single modality. However, none of them predicts with 100% accuracy, as each individual exhibits varied symptoms for the disease. To decrease the rate of misdiagnosis, multiple modalities can be used for prediction. Besides, there is also a need for a self-diagnosis system to narrow down the risk of virus spread in testing centres. Therefore, we propose a robust IoT and deep learning-based multi-modal data classification method for the accurate prediction of COVID-19. Generally, highly accurate models require deep architectures. In this work, we introduce two lightweight models, namely CovParaNet for audio (cough, speech, breathing) classification and CovTinyNet for image (X-rays, CT scans) classification. These two models were identified as the best unimodal models after comparative analysis with the existing benchmark models. Finally, the obtained results of the five independently trained unimodal models are integrated by a novel dynamic multimodal Random Forest classifier. The lightweight CovParaNet and CovTinyNet models attain a maximum accuracy of 97.45% and 99.19% respectively even with a small dataset. The proposed dynamic multimodal fusion model predicts the final result with 100% accuracy, precision, and recall, and the online retraining mechanism enables it to extend its support even in a noisy environment. Furthermore, the computational complexity of all the unimodal models is minimized tremendously and the system functions effectively with 100% reliability even in the absence of any one of the input modalities during testing.

Designing a multi-constrained QoS (Quality of service) communication protocol for mission-critical applications that seeks a path connecting source node and destination node that satisfies multiple QoS constrains such as energy cost, delay, and reliability imposes a great challenge in Wireless Sensor Networks (WSNs). In such challenging dynamic environment, traditional routing and layered infrastructure are inefficient and sometimes even infeasible. In recent research works, the opportunistic routing paradigm which delays the forwarding decision until reception of packets in forwarders by utilizing the broadcast nature of the wireless medium has been exploited to overcome the limitations of traditional routing. However, to guarantee the balance between the energy, delay and reliability requires the refinement of opportunistic routing through interaction between underlying layers known as cross-layer opportunistic routing. Indeed, these schemes fail to achieve optimal performance and hence require a new method to facilitate the adoption of the routing protocol to the dynamic challenging environments. In this paper, we propose a universal cross-layered opportunistic based communication protocol for WSNs for guaranteeing the user set constraints on multi-constrained QoS in low-duty-cycle WSN. Extensive simulation results show that the proposed work, Multi-Constrained QoS Opportunistic routing by optimal Power Tuning (MOR-PT) effectively achieves the feasible QoS trade-off constraints set by user by jointly considering the power control and selection diversity over established algorithms like DSF [1] and DTPC [2].

One of the most important challenges in the Wireless Sensor Networks is to improve the performance of the network by extending the lifetime of the sensor nodes. So the focus is on obtaining a trade-off between minimizing the delay involved and reducing the energy consumption of the sensor nodes which directly translate to an extended lifetime of the sensor nodes. An effective Sleep-wake scheduling mechanism can prolong the lifetime of the sensors by eliminating idle power listening, which could result in substantial delays. To counter this, an anycast forwarding scheme that could forward the packet opportunistically to the first awaken node may result in retransmissions as if the chosen node falls in resource constraints. The algorithm, namely Prim's-Dual is proposed to solve the said problem. The algorithm considers five crucial parameters, namely the residual energy of the nodes, transmission power, receiving power, packet loss rate, interference from which the next hop is determined to extend the lifetime of the sensor node. Since the proposed work is framed keeping critical event monitoring in mind, the sleep-wake scheduling is modified as low-power, high-power scheduling where all nodes are in low-power and the nodes needed for data transmission are respectively turned on to high-power mode. The integrated framework provides several opportunities for performance enhancement for conflict-free transmissions. The aim of our algorithm is to show reliable, energy efficient transfer without compromising on lifetime and delay. The further effectiveness of the protocol is verified. The results demonstrate that the proposed protocol can efficiently handle network scalability with acceptable latency and overhead.

No abstract

No abstract