Future wireless technologies play a crucial role in developing innovative real-time applications by enabling seamless communication [1,2]. They offer and facilitate real-time data sensing and exchange with low-latency connectivity channels for IoT-based solutions in innovative systems [3,4]. Emerging technologies that integrate with the Internet of Things (IoT) have revolutionized innovative development and communication for interaction with the physical world [5,6]. Fog computing enhances the capabilities of IoT networks by improving data processing closer to the source and improving system efficiency [7,8]. In healthcare IoT networks, such computing paradigms enhance real-time decision-making capabilities and reduce computational overhead under resource constraints [9,10]. Also, edge computing [11,12] is being utilized in many innovative applications to process the requested data at the edge, rather than by devices or local servers. However, many approaches have been proposed that combine edge computing with reduced computing power on constrained devices [13,14]. Nonetheless, it remains challenging to achieve effective load distribution in dynamic networks, particularly when addressing energy holes near the edges. Despite numerous advancements, the quality of service in IoT-based healthcare applications remains a pressing research issue, particularly when devices are heterogeneous [15,16]. On the other hand, trust computing ensures the reliability of e-health records by continuously evaluating and maintaining device trustworthiness [17,18]. In a critical healthcare system, timely identification of malicious behavior and maintenance of integrity are essential to protecting sensitive medical information. These systems require a significant contribution to establish trust in relaying services and offer the most robust multi-paths to enhance system performance in terms of resource management [19–21]. This research aims to introduce a fog-based intelligent framework using weigted trusted based aggregation modeling for efficient, and secured IoT-based healthcare communication. It enhances the security of health records and promptly transmits crucial data to processing servers, leveraging authentic fog-IoT-driven routing strategies. In real-time applications, the reliability of health data can be ensured through regular device monitoring using trust computing [22,23]. However, detecting faulty communication and malicious threats in healthcare applications in a timely and accurate manner remains a challenging research problem for IoT-driven systems.

Healthcare systems have become increasingly vulnerable due to inconsistent communication and potential attacks, thereby affecting the overall system. In addition, non-optimised decision-making schemes that do not account for network performance parameters degrade the resilience of critical applications to disruption and high-rate congestion. Despite several proposed approaches, most still lack fault-tolerant, robust trust-aware routing strategies and provide no guarantee of consistent device interactions in the face of malicious activity. Unreliable and compromised data is forwarded by network devices without establishing mutual trust or considering dynamic trustworthiness factors. It undermines the system’s confidence in timely decision-making and increases computational complexity when handling faulty traffic flow. Therefore, this study aims to address these issues by introducing an intelligent framework that integrates weighted-based trust aggregation with more reliable transmission channels, thereby ensuring consistent routing in the IoHT environment at the Fog-based communication layer. Based on the aforementioned discussion, the proposed framework addresses the following research questions.

RQ1: How can a healthcare system be scalable and provide reliable communication between devices in fog-based architecture against potential threats?

RQ2: How can a lightweight secured paradigm be attained while routing the healthcare records over an unpredictable environment?

RQ3: How can effective strategies be adopted to maintain the longevity of established routes and enhance the effectiveness of decision making system?

Ultimately, the main contributions of our research works are highlighted as follows. i.

A dynamic and scalable IoT-based healthcare architecture developed and integrated with fog computing to balance the network resources and enhance reliability against threats.

The research work is structured as follows. Related work is covered in Section 2. The proposed framework and its developed components are explained in Section 3. Section 4 provides an experimental discussion. Finally, the conclusion and future work are outlined in Section 5.

The IoT network enables seamless collection of sensor data by integrating wireless systems, enabling real-time monitoring in unpredictable environments [24,25]. The limitations of bounded networks pose distinct research challenges for processing, storage, and energy efficiency. Furthermore, the critical operations demand low-latency communication and a trust-aware monitoring system across distributed devices [26,27]. Maintaining network scalability and security in dynamic healthcare environments with dynamic conditions and environmental status presents significant challenges. Authors in [28] introduce a context-based adaptive trust model for innovative healthcare systems. It explores Bayesian approaches and similarity measures to address threats such as bad mouthing and ballot stuffing. Moreover, direct and indirect trust are computed using adaptive weights, entropy values to minimize bias, and context similarity to filter out malicious nodes. The proposed solution was simulated in Contiki-Cooja, and the results showed significant improvements in trust and reliability for healthcare systems. In [29], the authors introduce a novel protocol that integrates fog computing and software-defined networking (SDN) to enhance network scalability and support location-based services in VANETs. By exploring network lifetime, average distance, and signal-to-interference-plus-noise ratio, a new cluster-head (CH) selection algorithm is proposed. It reduces the overhead of long-distance transmission control and packet loss. Moreover, a dual-phase strategy is utilised so the AODV protocol acts as a fallback for handling large and complex packets. For the detection of network attacks and preseve data privacy with secure transmission, a study [30] proposed a novel method known as Cross-Layer and Cryptography-based Secure Routing (CLCSR). It is two phase framework: Firstly, a secure clustering mechanism is introduced using cross layer method and secondly a data is protected against threats based on lightweight cryptography principles. In addition, the use of cross-layer parameters and probability models, the proposed method provides the selection of most reliable and optimal cluster heads to improve the overall performance. Moreover, the lightweight Elliptic Curve Cryptography (ECC) is applied to impose the devices authentication and only grant the authorized access to network data. Authors of [31] present a two-layered hierarchical design for fog devices and explore cluster-based techniques for data aggregation and optimizing resources in healthcare applications. Three efficient algorithms are proposed for decreasing additional energy consumption and data latency. The proposed solutions are tested and validated using the iFogSim toolkit, and the performance results have shown significant improvements over existing approaches. Authors in [32] proposes a secure data forwarding mechanism for MQTT a widely used IoT communication protocol. It enhances reliability and data security by integrating fog-based computing resources with the MQTT ecosystem. Moreover, an informed data-transmission decision strategy is proposed that employs trust models for IoT and fog devices. To achieve real-time security and align with real-time network demands, the proposed solution utilizes key components, including TLS/SSL and cryptographic techniques. Authors in [33] introduce the DeW-IoMT framework for fog-cloud communication systems. The novel roof-computing layer reduces latency, processes tasks locally, and enhances security, thereby serving as an intermediary for communication. The proposed framework, simulated using iFogSim, minimizes computational overheads and ensures timely system responses for healthcare applications. Authors [34] introduce a fuzzy logic–based secure hierarchical routing scheme (FSRF) to enhance data privacy and routing effectiveness in healthcare systems. Moreover, a framework for fuzzy-based trust is explored to analyse device authenticity and ultimately increase the reliability of IoT communication against various attacks. In addition, the Firefly Algorithm is utilised for clustering and selecting optimal cluster heads, using a fitness function that evaluates multiple parameters.

This section provides a detailed discussion of the Fog-Assisted Anomaly-Aware Trust-Based Routing (FATR) framework, which aims to improve resource efficiency and enable trusted collaboration among health devices. Our proposed framework comprises three main phases: system design, anomaly detection using trusted computing, and communication performance improvement. Eq. (1) determines the device power consumption at the fog layer, Pfog, with the weighting sum of a particular power consumption of a device Pdevicei, and scaling factors αi reflecting each device’s operational state.

This section presents a performance evaluation of the FATR framework compared with existing studies. The simulation is conducted using iFogSim, which comprises sensors, fog, and infrastructure. The deployed network comprises various healthcare sensors, including wearable health-monitoring devices, interconnected with fog nodes. Fog devices are more stable and have greater computational and storage resources than IoT sensors. The transmission range is set to 5 m. The initial energy for each sensor is fixed at 5 J, and the simulation dimension is 3000 m × 3000 m. A varying health sensors are deployed to monitor vital parameters such as heart rate, temperature, and blood pressure. Attacker nodes are also introduced at rates of 10%–80% to simulate network threats and inject unauthenticated data. The experimental analysis is conducted across varying fog capacity and attacker intensity levels, focusing on energy consumption, network throughput, response time, and attack resilience. Table 1 declares the default parameters that are utilized in experiments. Moreover, approximately 50 simulations were conducted to generate synthetic data and a dataset for analysing the proposed FATR framework and related approaches. The dataset schema is defined in Table 2 to align with IoT-Fog-driven healthcare systems, including attributes such as device identity, timestamp, sensor readings, threat level, anomaly flag, and threat type.

Fig. 3a,b present the performance analysis of the FATR framework and related studies with respect to network throughput. The results demonstrated that the FATR framework enhances network throughput by an average of 36% and 41% over CLCSR and FSRF, respectively. This improvement is due to intelligent route prediction that accounts for network conditions and effective load balancing between IoT and fog systems. Additionally, integrating trust and secure routing paths enhances the FATR framework’s ability to identify anomalies and exclude faulty routes from communication channels. Using the Isolation Forest method, the FATR framework dynamically adjusts and optimizes the resources of healthcare applications, ultimately providing a more reliable ecosystem to handle a higher volume of health records for further processing.

Fig. 4a,b present the performance evaluation of the FATR framework for energy consumption compared with existing approaches. Across varying fog capacity and attacker intensity, the statistical results showed significant performance improvements for the FATR framework, averaging 39% and 46%, respectively. This improvement is due to adaptive routing schemes that integrate intelligence and leverage more stable forwarders within the trusted healthcare ecosystem. The prediction algorithm is developed for trust evaluation using weighted methods that explore multiple parameters, providing a consistent method for selecting the next-hop for transmitting health data to fog devices. Moreover, network conditions are also exploited to maintain efficient resource management.

Fig. 5a,b illustrate the performance of the FATR framework with CLCSR and FSRF in terms of attack resilience across varying fog capacity and attacker intensity. The results showed that the FATR framework outperformed existing studies by an average of 43% and 49%, respectively. This is due to the inclusion of dynamic trust updates and the exclusion of malicious devices from data transmission in healthcare applications. It not only provides a consistent path for long-term network connectivity but also enables distributed processing with intelligent threat-detection strategies to ensure data integrity and device authentication. Moreover, by excluding faulty links and devices from real-time health monitoring, the FATR framework reduces the additional delay caused by flooding vulnerable traffic, thereby improving its computational efficiency.

In Fig. 6a,b, the FATR framework and existing approaches are evaluated across varying fog capacity and attack devices in terms of system response. The performance results have demonstrated significant improvements in system response over CLCSR and FSRF, by an average of 47% and 51%, respectively. It is due to the consideration of multiple parameters to identify reliable connections between peer devices and to enhance the stability of IoT-based healthcare applications. In addition, the devices’ lightweight-ness and history are considered when assessing their trustworthiness and reducing the risk of bottlenecks, thereby improving data flow in constrained IoT devices.

Fog networks are rapidly utilised in many real-time systems for effective management of resources and timely response in a critical environment. The healthcare devices capture patient records and forward them to a processing platform for feedback from medical experts. The fog devices are placed closer to the source to reduce latency and improve response time when communicating with interacting devices. It not only balances resource consumption but also minimizes overheads in a constrained IoT-based environment. However, due to insecure channels and the presence of numerous malicious activities in an unpredictable network, many approaches fail to propose a trustworthiness solution and ultimately compromise sensitive information through disclosure. We introduce a fog-based framework to mitigate potential threats and optimize the healthcare network through efficient decision-making routing. The integration of a weighted-based trust aggregation model secures the IoT network against malicious threats and enhances the reliability of real-time communication. Moreover, lightweight methods minimize device energy consumption and help ensure a green network with an efficient anomaly-detection system. However, our proposed framework can protect network resources as malicious attacks increase in a distributed manner. In future work, we aim to explore deep learning techniques integrated with blockchain technology to reduce threat probabilities and enable seamless interactions and smarter disease identification.