October 9, 2017
Lifting Healthcare through the Fog
By Arsalan Mosenia, Princeton University
Healthcare has been revolutionized through rapid advances in communication protocols and the miniaturization of transceivers in 1990s, along with the emergence of wearable medical sensors (WMSs) in the early 2000s. These sensor-based systems, which capture, analyze, and store various physiological signals for future use, are key to enabling proactive and holistic approaches to healthcare [1,2]. Fog computing and networking play a vital role in enabling these systems to work. Here’s how.
Limitations of cloud computing
Cloud computing, with its scalable storage and anytime/anywhere processing services, has significantly extended the application landscape of WMS-based systems (see  for a comprehensive survey). Researchers and organizations worldwide have been involved in the design, prototyping, development, and deployment of cloud-enabled WMS-based technologies and services. Despite its benefits, the cloud can’t be used in latency-sensitive WMS-based applications, such as real-time seizure detectors, due to two fundamental reasons:
- The delays caused by transferring data to the cloud, processing the data on the cloud, and sending the results back to the application aren’t acceptable for real-time healthcare applications.
- Even a short period of unavailability of the application, which may be caused by cloud failure or Internet connectivity loss, could be life-threatening.
Promise of fog computing
Fog computing enables on-time service delivery with high reliability while addressing several challenges associated with the use of cloud computing, such as network delay/failure or cost overheads of transmitting data to the cloud). Fog computing is a distributed horizontal architecture that exploits computational, storage, and networking resources along the cloud-to-thing continuum (see  for an extensive description of fog computing).
Driven by the rising market of personal smart devices that are powerful, ubiquitous, and can offer a variety of resources for fog computing, fog has emerged as an alternative to cloud for WMS-based systems. Fog offers three fundamental advantages over cloud-only approaches (see  or  for an in-depth discussion of these benefits):
- Low latency: Fog computing exploits close-to-the-user computational/storage resources. It also minimizes data transmission overheads since it offers local computation on the data. These enable the implementation of a variety of ultra low-latency/real-time applications using close-to-the-user resources with minimal reliance on the cloud.
- Privacy: Since raw data can be processed locally before being shared with third-party servers, inessential portions of data can be filtered [4, 5]. This significantly enhances the privacy of medical data.
- Resiliency against cloud/network failure: Exploring different resources along the cloud-to-thing continuum enables the safe recovery of healthcare applications in the presence of a cloud/network failure. A well-designed fog-based application can detect such failures and eaily address them by utilizing other available resources around the user. For example, cloud-dependent tasks can be divided between local resources.
|Fig. 1: Wearables are continuously collecting a huge amount of data. Within a fog architecture, the data can be processed on different components as it flows through several devices along the cloud-to-thing continuum.|
Research is paving the way for transforming patient outcomes through fog-based approaches. In one study, researchers  have proposed a fog-based system to detect, predict, and prevent falls by stroke patients, with technology that has lower energy consumption and faster response time over cloud-based alternatives. These are among the first steps toward enabling fog-based services in healthcare; the current trend shows that fog computing will become a very promising research direction in healthcare and will continue to grow in importance and applications as IoT conquers new grounds .
Enabling Fog in Healthcare
To further accelerate the research on fog computing, the OpenFog Consortium has brought together researchers and system architects from industry, academia, and non-profit organizations. We represent a commitment toward cooperative and open fog-based systems to further boost their development and deployment. As a technical member of the Security Working Group, I am part of a team that is exploring the security landscape of OpenFog architecture. We are defining and investigating various security/privacy-oriented domain-specific challenges. At Fog World Congress later this month, we will be discussing how fog computing can enhance security and privacy in healthcare and across all industries. Please look for us in Santa Clara, October 30-November 1, where we will be releasing our position paper on “OpenFog Security Requirements and Approaches” .
References A. Mosenia, S. Sur-Kolay, A. Raghunathan and N. K. Jha, “Wearable Medical Sensor-Based System Design: A Survey,” in IEEE Transactions on Multi-Scale Computing Systems, vol. 3, no. 2, pp. 124-138, 2017.  A. M. Nia, M. Mozaffari-Kermani, S. Sur-Kolay, A. Raghunathan, N. K. Jha, “Energy-efficient long-term continuous personal health monitoring,” IEEE Trans. Multi-Scale Computing Systems, vol. 1, no.2, pp. 85-98, 2015  OpenFog Consortium, “OpenFog Reference Architecture,” https://www.openfogconsortium.org/ra/, accessed: 09/01/2017  A. Mosenia, J. F. Bechara, P. Mittal, and M. Chiang, “ProCMotive: Bringing Programability and Connectivity into Isolated Vehicles,” https://arxiv.org/abs/1709.07450, accessed: 09/25/2017  B. A. Martin, F. Michaud, D. Banks, A. Mosenia, R. Zolfonoon, S. Irwan, S. Schrecker, and J. K. Zao, “OpenFog Security Requirements and Approaches,” in Proc. Fog World Congress, 2017  V. Stantchev, A. Barnawi, S. Ghulam, J. Schubert, and G. Tamm, “Smart items, Fog and Cloud computing as enablers of servitization in healthcare,” Sensors & Transducers, vol. 185, no. 2, p. 121, 2015  Y. Cao, S. Chen, P. Hou, and D. Brown, “FAST: A Fog computing assisted distributed analytics system to monitor fall for stroke mitigation,” in Proc. IEEE Int. Conf. Networking, Architecture and Storage, 2015, pp. 2–11.
About the author
Arsalan Mosenia received the B.Sc. degree in Computer Engineering from Sharif University of Technology in 2012, and the M.A. and Ph.D. in Electrical Engineering from Princeton University, in 2014 and 2016, respectively, under the supervision of Prof. Niraj K. Jha. Upon the completion of his Ph.D., he joined Profs. Mung Chiang’s (Purdue University) and Prateek Mittal’s (Princeton University) research groups as a postdoctoral research associate. He is also affiliated with Center of Information Technology and Policy (CITP). During his Ph.D., he investigated several security and privacy challenges of different IoT-enabled systems, namely implantable and wearable medical devices, smartphones, and industrial/home automation systems. As a postdoctoral research associate, he is currently exploring potential security threats against Internet-connected vehicles. Hi research lies at the intersection of Internet of Things (IoT), information security and privacy, and machine learning. His studies resulted in several academic papers, currently submitted to or published in top-tier conferences and IEEE Transactions, and five U.S./Provisional Patents. You can find further information about this author at http://princeton.edu/~arsalan