Trust-Based Decision Making for Health IoT Systems
Abstract:
With the onset of the Internet of Things (IoT) era, the number of IoT devices and sensors is increasing tremendously. This paper is concerned with a health IoT system consisting of various IoT devices carried by members of an environmental health community. We propose a novel trust-based decision making protocol that uses trust-based information sharing a the health IoT devices, so that a collective knowledge base can be built to rate the environment at a particular location and time. This knowledge would enable an IoT device acting on behalf of its user to decide whether or not it should visit this place/environment for health reasons. Unlike existing trust management protocols, our trust-based health IoT protocol considers risk classification, reliability trust, and loss of health probability as three design dimensions for decision making, resulting in a protocol suitable for decision making in health IoT systems. Our protocol is resilient to noisy sensing data provided by IoT devices either unintentionally or intentionally. We present performance data of our trust-based health IoT protocol and conduct a comparative performance analysis of our protocol with two baseline protocols to demonstrate the feasibility.
Existing System:
There is a great potential for applying IoT technology across all sectors including both industrial and public to improve operation efficiency, reduce cost, and provide better service. Healthcare and public safety domains have a clear opportunity today to seize the benefits of IoT technology. Remote monitoring of medical parameters, smart hospital services, individual well-being, and emergency site and rescue are a few examples of applications that fall under these domains [3]. Environmental health IoT devices [4, 5] are available at a very affordable price, and when combined with a mobile application running on smart phones, can provide high quality readings for various environmental parameters, like CO levels, humidity, hydrocarbons, dust, noise, chemical fumes, fragrances, and so on. Since the measurement of the environment has a direct relation with healthcare of certain ailments and health in general, health IoT devices are expected to play a major role in providing excellent support in day-to-day healthcare. For example, an elderly person suffering with high blood pressure might not want to go to a place where noise levels are very high. Prior knowledge of the environment can safeguard decision making.
Proposed System:
Trust management of IoT systems is still in its infancy stage. Yan et al. [11] provided a survey of contemporary trust management techniques for IoT. However, no specific goals for health IoT were discussed. Our paper on the other hand has a specific goal. That is, we aim to help environment health conscious users carrying IoT devices become situation aware of surrounding environments. The novelty is in the use of trust management to effectively collect various geo-based health related data and to use this data for reliable decision making. Very recently [6, 7, 12, 13] discussed trust management for distributed IoT systems, where social relationships are established between things based on interactions. Both direct observations and indirect recommendations are factored into trust assessment of nodes. Unlike our work, their emphasis is how the social relationship of distributed IoT entities would affect the trust relationships and thus the service dispositions between IoT devices which provide services toward each other. While the trust management mechanisms proposed are valid for service composition and binding IoT applications, they cannot be applied to health IoT applications since the main characteristics of health IoT are not taken into consideration. Unlike [6, 7, 12, 13] cited above which consider only service providers’ trust scores for decision making, we specifically consider a patient’s risk classification and loss of health probability for trust-based decision making. Saied et al. [14] proposed a centralized IoT trust management system where a service requesting node is provided with the best assisting nodes to best answer the service request. This is achieved by computing “service context similarity” between reports stored centrally in the cloud and the target service where the weight of a report is based on the trustworthiness of the reporting node. Requesting nodes evaluate assisting nodes after the service is rendered by sending a report to the centralized trust management system in which it either rewards or punishes the assisting nodes. A recommender’s trust is based on the deviation between its reports with the majority of other reports with similar service context. Similar to [14], our work also recognizes the benefit of a centralized trust management system to offload the overhead from resource constrained devices and avoid communication overheads.
Conclusion:
In this paper we proposed and analyzed a trust-based decision making protocol for health IoT systems. We described the problem and thus the motivation to create a trust-based decision making protocol for a health IoT system. Our trust-based health IoT protocol considers risk classification, reliability trust, and loss of health probability as three design dimensions for decision making. We developed a trust computation protocol for a health IoT system to assess the reliability trust of individual IoT devices. We also developed a method to aggregate sensing data and derive the probability of health loss, should the user enter a given location at a given time. Based on the user’s vulnerability our system then assesses if the risk is low or high enough to support or refute the user’s request of entering the location specified in the query. Our simulation results demonstrated the feasibility of our approach with a high correct decision ratio (CDR) relative to the ground truth case with CDR=1 despite increasing malicious node population in a health IoT system. We also conducted a comparative performance analysis of our proposed trust-based health IoT protocol with two baseline protocols (NT and NMH) with convincing results.
Reference:
[1] P. Fraga-Lamas, T. Fernández-Caramés, M. Suárez-Albela, L. Castedo, and M. González-López, “A Review on Internet of Things for Defense and Public Safety,” Sensors, vol. 16, no. 10, p. 1644, 2016.
[2] S. Raza, P. Misra, Z. He, and T. Voigt, “Building the Internet of Things with bluetooth smart,” Ad Hoc Networks, 2016.
[3] E. Borgia, “The Internet of Things vision: Key features, applications and open issues,” Computer Communications, vol. 54, pp. 1-31, 2014.
[4] Adafruit Industries, New York City, NY, USA. Adafruit Industries products.2017.[Online].Available: https://learn.adafruit.com/category/adafruit-products [Accessed: 25- Apr- 2017].
[5] Sensorcon, Williamsville, NY, USA. Sensorcon Sensing Products by Molex. 2017. [Online]. Available: http://www.sensorcon.com [Accessed: 25- Apr- 2017].
[6] I. R. Chen, F. Bao, and J. Guo, “Trust-based service management for social internet of things systems,” IEEE Transactions on Dependable and Secure Computing, vol. 13, no. 6, pp. 684-696, 2016.
[7] I. R. Chen, J. Guo, and F. Bao, “Trust Management for SOA-Based IoT and Its Application to Service Composition,” IEEE Transactions on Services Computing, vol. 9, no. 3, pp. 482-495, 2016.
[8] K. Habib, A. Torjusen, and W. Leister, “Security analysis of a patient monitoring system for the Internet of Things in eHealth,” in Proceedings of the International Conference on eHealth, Telemedicine, and Social Medicine, 2015.
[9] S. C. Mukhopadhyay and N. Suryadevara, “Internet of Things: Challenges and Opportunities,” in Internet of Things: Springer, 2014, pp. 1-17.
[10] A. B. Pawar and S. Ghumbre, “A survey on IoT applications, security challenges and counter measures,” in IEEE International Conference on Computing, Analytics and Security Trends, 2016, pp. 294-299.