«Opportunistic IoT: Exploring the Harmonious Interaction between Human and the Internet of Things Bin Guo1 Daqing Zhang1,2 Zhu Wang1 Zhiwen Yu1 ...»
Opportunistic IoT: Exploring the Harmonious Interaction
between Human and the Internet of Things
Bin Guo1 Daqing Zhang1,2 Zhu Wang1 Zhiwen Yu1 Xingshe Zhou1
1. Northwestern Polytechnical University, China
2. Institut TELECOM SudParis, France
The traditional view of Internet of Things (IoT) attempts to connect all the physical
objects to build a global, infrastructure-based IoT. In this paper, however, we will
present opportunistic IoT, which is formed based on the ad hoc, opportunistic networking of devices (e.g., mobile phones, smart vehicles) using short-range radio techniques (e.g., Bluetooth, Wi-Fi). The opportunistic IoT demonstrates inherently the close relationship between human and opportunistic connection of smart things. It enables information forwarding and dissemination within and among the opportunistic communities formed based on the movement and opportunistic contact nature of human. We characterize the bi-directional effects between human and opportunistic IoT, discuss the technical challenges faced by this new research field, and propose a reference architecture for developing opportunistic IoT systems. Some of our ongoing practices, including opportunistic mobile social networking, opportunistic marketing, and community service provision are further presented to demonstrate the potential application areas and technical solutions of opportunistic IoT.
Keywords: opportunistic IoT, opportunistic mobile social networking, heterogeneous community orchestration, information dissemination, human-centric sensing
1. Introduction The Internet of Things (IoT) refers to the emerging trend of augmenting physical objects and devices with sensing, computing, and communication capabilities, connecting them to form a network and making use of the collective effect of the networked objects. Under the vision of IoT, the next-generation Internet will promote the harmonious interaction between human, societies, and smart things .
In the past few years, significant research efforts have been made on IoT, mainly from a thing-oriented perspective. A wide range of areas are covered, including object identification and tracking, object networking, sensing data visualization, privacy control, and so on . Nevertheless, the “harmonious” interaction between human and IoT, or in other words ― the social side of IoT, has yet little been explored.
In terms of its topology features, we can broadly category network connection into two types: infrastructure-enabled connection and ad hoc or opportunistic connection.
The prior type uses pre-existing infrastructure (e.g., base stations, routers, access points) and manages the data in a centralized manner. The latter one, however, is founded on the development of opportunistic networks , which uses infrastructure-free, shortrange radio techniques (Bluetooth, Wi-Fi, etc.) to build decentralized, ad hoc networks.
Opportunistic networks are human-centric because they inherently follow the way that people opportunistically get into contact. For instance, customer A can connect with other customers that opportunistically meet in a coffee shop to build an ad-hoc mobile phone network. Information sharing and communication can be further conducted among the members of this “opportunistic, physical-proximity-triggered” community.
When A leaves the coffee shop, the information she obtained from this opportunistic community (e.g., there will be an open-air concert in the Central Park tomorrow night) can be further disseminated to other newly-formed opportunistic communities (e.g., with his/her colleagues in the working place, with other passengers on the bus).
The traditional view of IoT attempts to connect all the physical objects to build a global, infrastructure-based IoT. In this paper, however, we will present opportunistic IoT, which addresses information dissemination and sharing within and among opportunistic communities (with pairs of devices) that are formed based on the movement and opportunistic contact nature of human. Various personal devices, such as mobile phones, wearable devices, vehicles, can form opportunistic IoT when they are equipped with the short-range communication and sensing modules. We illustrate the concept of opportunistic IoT through the following “opportunistic trading” use case.
Different from traditional market-based trading and online shopping, opportunistic trading is founded on the disseminating and matching of trading requests in opportunistic IoT environments. For example, Bob wants to buy a second-hand “Harry Potter” via the opportunistic trading agent (OTA) running on his mobile phone. While Bob moves each day, his trading request is shared by people in the vicinity (forming an opportunistic community using mobile phones). Since the moving range and mobility pattern of Bob is roughly fixed (the number of people he can encounter is thus limited), to increase the number of trading request receivers and speed up the request dissemination process, OTA will employ other mobile nodes as “brokers” to help store and forward Bob’s request. How to select brokers becomes a significant yet difficult problem, where we should consider the popularity of the node (in terms of its mobility patterns) and other social features (e.g., willingness to act as a broker). Two days later, Alice (the book seller), who lives in another district of the city, is found by OTA and the brokers.
The above scenario demonstrates the bi-directional relationship between human and opportunistic IoT. On one hand, opportunistic IoT becomes the primary media to sense and monitor human behaviours (e.g., mobility patterns can be learned from the GPS trajectories collected from user-carried mobile phones); on the other hand, the performance of IoT is also affected by human behaviours (e.g., social features are important for broker selection). In summary, opportunistic IoT presents a promising research domain to study the social side of the IoT. Further, according to the Oxford English Dictionary , collaboration is the act of working with another person or group of people to create or produce something. In technology, it encompasses a broad range of tools that enable groups of people to work together including social networking, instant messaging, web sharing, and so on. Wikipedia, Blogs, and Twitter are good examples of collaborative tools. By leveraging the opportunistic connection among people in the proximity, opportunistic IoT facilitates information dissemination and sharing, as well as spontaneous social networking (when the information exchanged is user profile ) among people in opportunistic communities, presenting a promising way to enhance instant human collaboration and data sharing.
In the following sections, we first describe the relations between our work and several closely-related research areas. The bi-directional effects between human and IoT will then be characterized. In Section 4, we discuss the research challenges on opportunistic IoT, followed by the description of a conceptual architecture in Section 5.
Our ongoing efforts to opportunistic IoT are presented in Section 6. Finally, we conclude the paper and present the future work.
2. Research Background and Related Work Research on opportunistic IoT can benefit from the ongoing and past research outcomes in pervasive computing, opportunistic networking, social computing, and mobile social networking.
In his seminal paper , Mark Weiser prophesied that pervasive computing can learn and adapt to human needs in an unobtrusive, ubiquitous manner. Over the last decade, mainframe studies on pervasive computing are about ubiquitous tracking/sensing , context-aware computing , personalization , mainly relying on the wireless infrastructure support (e.g., cellular networks, WLAN). Opportunistic IoT, however, addresses the limitation of wireless infrastructures, such as lacking network coverage, high cost, etc. In addition, the core of pervasive computing is context-awareness. Opportunistic IoT takes pervasive computing further, to explore the learned human behavior and social connection to enhance opportunistic data sharing.
Opportunistic networking is based on spontaneous connectivity between users with wireless devices , facilitating inter-device data routing and forwarding. Opportunistic IoT extends the opportunistic networking concept from two aspects: 1) it is rooted from the Internet of Things vision, which inherits the nature of smart things on ambient sensing. Local-sensed information (traffic dynamics, noise levels) can thus be opportunistically shared by others, i.e., supporting the so-called participatory sensing ; 2) it particularly explores the co-existing of opportunistic communities in the physical world and online communities in the virtual world, and study the interaction and collaboration between heterogeneous communities. There are also several studies that try to introduce the opportunistic element into IoT systems. For instance, Blackstock et al. have developed Magic Broker 2 , a lightweight middleware that supports spontaneous interaction between smart devices (public displays, mobile phones). Rohokale et al. have proposed a novel cooperative approach for the analysis of receiver sensitivity to enhance relay-based communication in wireless sensor networks . However, none of them explore human factors in IoT systems, especially the interaction and interplay between online and offline social communities.
Social computing refers to the computational facilitation of social studies and human interaction analysis as well as the design and use of technologies that consider social context . Similar to opportunistic IoT, social computing takes human factors and social behaviour analysis as key dimensions. However, social computing emphasis mainly on the analysis of human interaction using Web data, it does not target at the study of physical communities.
Mobile social networking (MSN) refers to social networking where individuals with similar interests connect with one another through their mobile devices . Similar to Web-based social networking, existing MSN services (e.g., Foursquare) occur in the virtual world, relying on full mobile access of the Internet. The opportunistic IoT, however, will drive a different form of MSN – the Opportunistic MSN , which aims to enhance spontaneous interaction/communication among people that opportunistically encounter in the physical world, without leveraging any infrastructure support.
In summary, opportunistic IoT shares many things in common with the aforementioned research areas, yet it goes beyond all those areas in terms of its focus and research challenges. Different from those areas that either focus on human behaviour/context analysis or opportunistic data sharing, the opportunistic IoT particularly addresses the interaction of the two research directions. Moreover, opportunistic IoT also studies the interlinking and collaboration between online communities and physical communities, as we present in the latter sections.
3. The Bi-directional Effects between Human and Opportunistic IoT By analyzing the tight-coupled relationship between human and opportunistic connection of smart things, we present the bi-directional effects between human/societies and opportunistic IoT, as shown in Fig.1.
Figure 1: The bi-directional effects between human, societies and IoT.
3.1. Human-Centric Sensing with Opportunistic IoT Various IoT devices (equipped with sensing and short-range communication capabilities) are weaved deeply into the fabric of everyday life. The diverse features of these devices present unprecedented opportunities to understand the aspects of interaction between humans and real-world entities. We characterize these humancentric interactions as human-object, human-environment, and human-human interactions. By analyzing the collected interaction data with advanced machine learning and data mining techniques, the opportunistic IoT is equipped with three sensing capabilities: user awareness, ambient awareness, and social awareness .
We characterize the attributes of them as follows.
User awareness refers to the ability to understand personal contexts and behavioral patterns. Examples include human activity, human popularity, preferences, etc.
Ambient awareness concerns status information on a particular space. Examples include space status and traffic dynamics (e.g., traffic jams).
Social awareness goes beyond personal contexts and extends to group and community levels. The objective is to reveal the patterns of social interaction (e.g., group detection, friendship prediction, situation reasoning), human mobility, etc.
3.2. The Impact of Human Behaviors on the Opportunistic IoT Data sharing is the major application area of opportunistic IoT, which exploits humans’ mobility and their gregarious nature to transmit information. Since the source node and destination nodes might be unaware of each other (e.g., in the opportunistic trading use case, Bob and Alice are unaware of each other) and may never meet in opportunistic networks, forwarding a message (e.g., selling a book called “Harry Potter”) from its sender to the nodes of interest (e.g., from Bob to Alice) becomes a big challenge. A trivial solution would be to flood the whole network with the message , but this would clearly saturate both network resources (in terms of available bandwidth) and device resources (e.g., in terms of energy, storage, and so on).
A better solution is to replicate the content to only selected nodes that have more chances to contact and influence others, and thus the broker-based solution is proposed (as demonstrated in the opportunistic trading use case). With this solution, each node (e.g., node Bob) carrying a message evaluates the suitability of any other node it makes contacts with as the broker (many social features are measured, as depicted later).
Messages are thus opportunistically disseminated by exploiting both the source node (e.g., node Bob) and the brokers selected, until they reach a node of interest (i.e., node Alice who lives in another district of the city wants to buy the book).
In opportunistic IoT, contacts between nodes are inherently tied with users’ social behaviors (e.g., two mobile phones contact when user A and B meet in a coffee shop).