All Available Positions

The Mobile Mesh Technology For Improved Connectivity in Canada will support 4 Postdoctoral Fellowships, 16 PhDs (or MSc students), and 10 undergraduate research assistants over the next 5 years. If you would like to be involved in this research program, please review the list of opportunities below to find the project that best matches your expertise and interests.

Research Abstract:

Connectivity in remote, rural, and Indigenous communities in Canada and around the world has led to a digital divide where millions lack access to online resources and the digital economy due to insufficient or non-existent infrastructure. To address this issue, Left has developed a novel approach known as RightMesh that capitalizes on the Wi-Fi, Wi-Fi Direct, and Bluetooth technologies that exist in most modern mobile devices. RightMesh turns individual user devices into “nodes” using mobile mesh networks, which can then be accessed by local users to connect to the internet. RightMesh enables data sharing and connectivity in the absence of, or working parallel with costly infrastructure. Moreover, the technology has been developed to seamlessly connect all nodes to the Internet so long as one of the nodes is connected to it. To ensure that Left continues to lead in this domain, it becomes necessary to optimize the network performance; develop best practices and methods that support remote, rural, and Indigenous software design strategies; implement engagement strategies to encourage participation in mesh-based apps; and develop incentives to foster connection with the RightMesh network. Each of these research areas will be explored through case studies in the rural and remote communities of northern Canada, with results used to inform the development of a simulation tool which will support expansion of the mesh networks to other locations around the world.


Jump directly to the various research project by clicking on one of the links below.

Click the following buttons to learn about the positions we are currently recruiting for.

GRADUATE POSITIONS

POST GRADUATE POSITIONS

For more information, see the following links:

Research Projects:

Please be sure to include the objective code [Obj#] in any correspondence related to this research program.

  • Obj1 Evaluation of the Mobile Mesh Network Performance
    The Left mobile mesh platform uses a unique and novel switching technology that has not been formally tested outside of the laboratory. As such, it is necessary to evaluate the network in real-world situations that might affect its performance. This includes understanding the network limitations and evaluating the performance of the network given different environmental conditions, available infrastructure, and under different traffic loads, as well as evaluating the performance of mobile mesh designed applications through the use of various case studies.

    • Obj1.1 Role Assignment Optimization [Postdoctoral Fellowship]
      Nodes in the network are assigned different roles (i.e. switching node, hotspot node, and normal client node). However, as nodes enter and exit the mesh (for example, when a mobile device moves in or out of a community), a node’s role could change. It is necessary to develop algorithms for neighbour discovery, understand how a network is affected by a change in a node’s role, and to develop algorithms to best assign network roles to optimize mesh coverage and capacity while minimizing delays. As such, the goal of this research question is to 1) develop and evaluate neighbourhood discovery algorithms, 2) understand the effect that a dynamic node structure will have on the performance of the mesh network, and 3) optimize the assignment of roles given different network scenarios. In addition to developing and evaluating these algorithms, the Postdoctoral Fellow will be responsible for working with and coordinating the various PhD interns that will research mesh network performance (i.e. Obj 1.2 through 1.9, where Postdoctoral Fellow and PhD intern timelines overlap). This work will build on existing prior research on topology control for mobile ad hoc networks which requires GPS positions. At the time the original work was published, GPS units were not common in mobile devices, but with low cost GPS available in almost all mobile smart phones, these types of techniques allow for building topologies which are energy efficient and perform well (Choi et. al, 2012).
    • Obj1.2 Device Density [PhD]
      It is expected that the utility and usability of the mesh network will depend on appropriate coverage within a community, and a user’s ability at connecting with the network. Presently, it is not known how dense the network needs to be (especially in communities where devices and people are transient), how many super peer devices are required to support networks of varying size, nor how the density might affect data transfer speeds from one device to another. As such, the goal of this research question is to explore the density dependence of the network to optimize performance given different densities, movement of nodes, and other factors. Signal strength measurements of neighbouring nodes (for each employed RAT) could be used to estimate density in a similar manner as suggested by Onur et. al. (2012) in order to inform this research.
    • Obj1.3 Switching Optimization [PhD]
      The mesh network relies on Bluetooth, Wi-Fi and Wi-Fi Direct technologies to connect devices. As a network expands, devices are assigned different roles to govern the movement of data from source node to its intended delivery node. It is necessary to evaluate the utility of Bluetooth, Wi-Fi and Wi-Fi Direct to act as a signaling agent, and further develop methods to optimally time the switching to achieve close to full two-way transmission of data (despite the switching). Further, it’s necessary to understand how network performance is affected as the number of switching nodes increases. Thus, the goal of this research question is 1) to develop and evaluate methods to optimize switching in a dynamic mesh, 2) to evaluate the effect of switching on network performance, and 3) to inform and update the mesh protocols. The research for this objective will be informed by existing work in software defined networks and virtual network interfaces such as work done by Bhatia et. al (Bhatia et. al, 2018).
    • Obj1.4 Routing Optimization [PhD]
      Depending on the density of the mobile mesh network, optimizing data packet routes becomes necessary to minimize wait times and to prevent or bypass data backlogs. To achieve this, it is necessary to identify the best metrics for route selection that consider current mesh density, current traffic load, and multiple device technologies (e.g. Bluetooth, Wi-Fi, Wi-Fi direct, etc.), and then to use this information to optimize routing within the network, possibly through the assignment or reassignment of role’s such that network coverage and capacity are maximized, while delays are minimized. The goal of this research question is to update the mesh protocol to achieve optimal routing given mesh density, traffic load, and the technologies present in the mesh.  Routing in wireless mesh networks is a well-studied field, however, due to the unique constraints of working in user space on mobile phones, new approaches must be adapted from existing work. In particular, this work will be driven by approaches such as multipath transmission control protocol (mTCP) (Sheu et. al, 2016), various metrics associated with Quality of Service (QoS) (Lemeshko et. al, 2013) and Quality of Experience (QoE) (Gomes et. al, 2010) in mobile ad hoc and mesh networks.
    • Obj1.5 Traffic Balance Optimization [PhD]
      With keeping network density, dynamic role assignment, and optimization routines for switching and routing data in mind, there is a need to develop a framework for selecting the most appropriate algorithm for a given situation. This will allow Left to manage traffic balance on the mesh. Specifically, it is essential to optimize traffic balance across multiple paths and considering the novel Left switching system. As such, the intern will work to identify different traffic balancing optimization algorithms in the literature, and evaluate them given different densities and dynamic role assignments. The results will be used to populate a framework for selecting the most appropriate optimization routine, and integrate it into the mesh network protocols. Many existing approaches in this area apply considering fixed infrastructure devices such as the approach by Takahashi and Asaka (2014), which would need to be adapted to work considering the mobility of the devices and the unpredictability of the routes.
    • Obj1.6 Develop Caching Protocols [PhD]
      Making use of caching nodes locally in the network is necessary for situations where the network is isolated spatially or temporally from other networks, or in situations where internet connectivity is broken or limited. This is particularly relevant in Rigolet and the circumpolar north, where the limited internet connection options are often overwhelmed by volume of users (especially during daytime hours). Using a local caching node will enable mesh-based apps to sync up with users outside of the local region via the Internet, without overwhelming the connection. Updates are stored in the caching device until a scheduled transmission when the local caching device syncs with the Internet node, which could store data centrally and pass along updates to other geographically separate meshes. Developing caching protocols will provide users with a sense of continuous performance, even with extremely limited “real” network capacity. As Ma and Jamalipour (2010) point out, “low cost wireless connectivities such as Bluetooth and IEEE 802.11 offer the mobile device an alternative way to communicate with each other … Unfortunately, end-to-end paths in MANETs may suffer intermittent connectivity due to node mobility, sporadic node density, short transmission range, and so on. This research aims to adopt approaches such as (Ma and Jamalipour, 2010) such that they operate under the constrained storage and resource availability present on mobile phones.
    • Obj1.7 Device Energy Conservation [PhD]
      Energy consumption and device battery life are important factors affecting usability, usefulness, and uptake of any new technology. As such, it is necessary to understand the impact of energy consumption on devices that participate as nodes in the mobile mesh. Specifically, it becomes necessary to understand factors affecting power consumption (e.g. mesh density, node roles, choice of signaling agent, etc.), how each affects the rate of power depletion, and to develop strategies and algorithms that prevent or slow power consumption. Further, in communities that have existing infrastructure (e.g. cellular towers, internet), it is necessary to explore the following questions: What is the battery discharge due to activity on the Mobile Mesh Network, and on what factors does this depend? How does the assignment and distribution of network roles affect battery consumption, and can this be minimized? How much energy can be saved by avoiding connectivity with the cellular towers and routing through the mesh? What is the relationship between network hops and the amount of energy saved, and can this be optimized given the factors that affect battery life? As such, the goal of this research project will be to answer these questions, and use this information to inform and update the mesh protocol to prolong device battery life. The annual carbon footprint for a mobile subscriber in 2007-2011 was 25kg, or roughly equivalent to running a 5W lamp for a year – including the supply chain, vendor, usage and end-of-life phases in the life cycle of the handset. This number has more recently been revised to upwards of 45kg, and is likely still growing with newer cellular access technology requiring even more energy. According to Huang et. al, LTE uses 23 times more energy than Wi-Fi and 3G uses 10.8 times more energy that Wi-Fi.
    • Obj1.8 Encryption [PhD]
      The Left library currently uses an adapted Open Whisper/Signal Protocol to enable end-to-end encryption. This library is used by Facebook, Whatsapp, and Skype, and has been extensively reviewed and found to be quite secure (see https://signal.org/blog/skype-partnership/, https://www.cyberscoop.com/signal-security-audit-encryption-facebook-messenger-whatsapp/). This security relies on Perfect Forward Secrecy, which periodically changes the encryption key. The hope is that if the key is changed quickly enough, even if it is broken, the amount of time to break the encryption will be slower than the switching time. Thus, the attacker will find it extremely difficult to penetrate the library. On mobile phones, keys cannot be switched on the fly because one phone may be temporarily unavailable due to poor connectivity or being in a low-power sleeping mode. The Open Whisper/Signal Protocol solution is to generate keys in batches of 100s and store them in a centralized server on the Internet. When the key is to be changed, the next key is requested from the server (Marlinspike, 2016). This is not possible in a mesh that may form without an Internet connection. We have adapted this approach for the offline case and have devices directly exchanging keys. However, this opens it up to man-in-the-middle attacks by forwarding nodes. The goal of this research question is to improve the exchange process in such a way that it is easy for end-users and developers, while maintain performance and security. To accomplish this, it is important to develop methods of encoding the encryption keys into a 2D barcode and exchange with others while they are nearby.
    • Obj1.9 Trust & Security [PhD]
      The goals of this objective are to develop a method to notify developers and end-users on the security level of data transmissions between peers (no encryption, somewhat insecure, mostly secure, and best security); develop a method to establish trust between devices within the mesh; and investigate how to establish trusted communications within the mesh and score the devices. Those devices which behave outside of expected operations would be scored poorly, and those that behave well and provide more resources would be scored high. A framework for scoring trust in distributed and delay-tolerant networks is described by Oleshchuk in 2017; an approach informed by these methods could be used to enhance the security level from the first part of the objective with measurable trust, so that developers and users know whether data transfers are occurring in a secure and trusted manner. Moreover, it is critical to develop a method to secure the low-level protocols, using encryption that does not rely on centralized communication, and which is suitable for disruption-tolerant ad-hoc networks. Trust models could be parametrized by data trust (accurate and honest data transmission), link trust (packet error, packet loss, etc.), node reputation (scoring by peers), similar to approaches used in untrusted sensor networks (Han, Jiang, et. al. 2015). Trust establishment and rating of nodes could be used to inform and improve upon Objective 1.1 as forwarding nodes must be trusted to deliver data with integrity. Finally, this PhD student will develop a tool for visualizing token transaction history, to help audit the flow of tokens in the mesh.
  • Obj2 Develop Best Practices for Mobile Mesh Implementation & App Design
    The continued use of, and engagement with the mobile mesh network will rely on the development of a diverse and robust ecosystem of mobile mesh network applications. To achieve this, it is necessary to develop a simple, straight forward, and easy to use mobile mesh network Software Development Kit (SDK) and Application Programming Interface (API). Software designers will also require guidance about how to develop tools that automatically integrate with the network, while providing a robust user experience. Given the potential of the mobile mesh networks in remote, rural, and/or Indigenous communities in Canada and around the world, which have varying levels of established infrastructure (e.g. internet, cell towers, satellite), it is necessary to explore software design methods that respect, honour, and value local and Indigenous culture and traditions. This also includes a need to understand best practices related to data sharing, data storage, and data mining as it pertains to mobile mesh apps that collect and share data between agencies and governments (such as those created to support Climate Change based community-led environment and health monitoring apps).

    • Obj2.1 Developing User Interface (UI) and User Experience (UX) Best Practices [PhD]
      We want the ecosystem for mobile mesh applications to grow in a manner that encourages user engagement and participation. To achieve this goal, it is necessary to develop a set of best practices for UI and UX design that consider the mobile mesh platform itself. This will include exploring and understanding how best to mitigate situations that might arise from slower than expected network performance, how best to maintain data integrity and user experience given the dynamic nature of the network (e.g. when a node leaves, or changes roles), how best (and if/when) to transition users from the mobile mesh to a Wi-Fi system, how best to visualize data exchange and cryptocurrency movement, and how best to incentivize the exchange of data through access to cryptocurrency sales and purchases. It also includes exploring and understanding how best to support developers through the debugging and testing process. While many of these challenges will need to be explored through testing, prototyping and iterative development, many of these questions will be explored through the application of modern User Interface Design frameworks. Based on popular frameworks, a optimally usable system must consider the effectiveness, responsiveness, platform-dependence, and intuitively of the system, as well as cultural factors of the primary users that might inform and influence design choices (Oppermann, 2002, Rincon et al., 2013). Moreover, cultural factors provoke three different consequences on UI and UX design: representation (whereby the layout of graphical elements, or interface tools is set according to social influences), prototype evaluation (which is affected by specific culturally-based decision making processes such as totalitarian, consensus-based, or democratic), and quality criteria (which involves the prioritization of UI design elements). To achieve this particular objective, it will become necessary to consider UI and UX from an Indigenous context to ensure that design decisions are informed by cultural and traditional factors, while also addressing issues such as slower than normal network performance (a common occurrence in, for example, Rigolet), and managing data integrity. Previous research in Columbia has explored this particular challenge. Specifically, researchers from the Université de Bretagne-Sud studied the socioeconomic and cultural influences on UI design while co-producing a software tool with the Nasa Columbian Native People. Through their work they were not only able to produce a successful and locally appropriate interface for their linguistic tool, but also produce a framework for cultural and socioeconomic considerations when designing User Interfaces (Rincon et. al., 2013). As such, we will use the Rincon et al. (2013) framework to select appropriate methods for establishing intuitive and useful designs that respect culture, while also informing users of activities specific to the mesh platform. In particular, open houses and focus groups facilitated by the community-research leads will be used to best identify UI and UX design choices. These will be integrated into new or updated to software tools for the community, and evaluated using standard usability heuristics.
    • Obj2.2 Decolonizing & Indigenizing the Software Design Process [PhD]
      Standard software development processes and design methods have been developed within the context of western science. While many of the development processes in use invoke user-centred design, iteration, and continuous improvement, they consider users as a group of individuals for which a system is being designed for. To ensure systems are developed to support Indigenous sovereignty and culture, it is necessary to explore and develop processes that are non-extractive, and co-creative. That is, the users must be involved at all development stages such that the system is something that is designed with them and not for them. Further, specific design methods may not respect the customs, cultures and traditions of many Indigenous communities which rely on consensus-based and decentralized decision making that is informed by empirical information, as well as culture and tradition (Kwiatkowski et al., 2009). For example, specific software design methods (such as A/B testing) are not necessarily as effective in communities where tradition dictates consensus be achieved after community-led discussions and debate. As such, it becomes necessary to explore these situations to better understand community traditions and develop a set of best practices that will support the design of mobile mesh applications that will be used in remote, rural, and Indigenous communities in Canada and abroad. As described in Section 2.3, participatory research and design methods have been successfully used to ensure that community perspectives are included during each stage of a research project (Schröter et al., 2005). Moreover, the participatory approach has been highlighted as an important tool to ensure marginalized and Indigenous communities are included in the research process – such that solutions are developed with the community, not for the community (Disalvo, et al., 2012, Hussain et al., 2012, Carroll and Rosson, 2007, Merkel et al., 2004, Holkup et al., 2004, Christopher, 2005, Jones, 2006). More recently, Cook (2018) described a case study in which participatory methods were used to develop a prototype health and environment monitoring app with an Inuit community in Nunatsiavut. With this in mind, we propose to explore and evaluate several case studies that use participatory methods to develop and test mesh-enabled software solutions for Indigenous community-based monitoring. As described by Cook (2018), standard participatory research and design tools such as surveys, interviews, focus groups, open houses, and workshops will be used to develop applications and to ensure that the community is involved at each stage of development. During each phase, we will also use participatory methods (in particular, open houses, workshops, and focus groups) to explore different design processes and tools standard to software design (e.g. requirements gathering, paper prototyping, etc.). That is, we will work with the community-research leads to present different methods of software development, then seek feedback from community in terms of their comfort with each method, noting in particular any potential pros and cons that have been identified. Workshops will be used specifically to develop software design skills with interested community members. Feedback obtained during the various events, as well as through follow-up semi-structured interviews and surveys will be used to evaluate the effectiveness of the methods explored. Community-research leads will play a vital role in this objective, helping to organize and facilitate open houses and workshops, and in the identification of community practices and traditions that might inform the selection of software design processes and tools. Moreover, through active community engagement throughout the design process, the community-research leads will help facilitate skills development related to software design to the broader community.
    • Obj2.3 Data Mining Best Practices [PhD] Many of the mesh applications that will be developed for rural and remote communities in northern Canada will support Indigenous-led community based monitoring programs with the goal of collecting environment and health data in response to climate change impacts. In addition to data available through remote sensors and weather stations, vast amounts of both quantitative and qualitative health and environment data will be collected through mesh based apps. Using data mining methods, these data will be used to support public health, wellness, and environmental stewardship decision making. However, it remains to be determined how best to process the qualitative and quantitative data sets in a manner that supports Indigenous self-determination and sovereignty, while providing the most accurate information for decision makers pertaining to the status of the environment and health. As such, there is a need to explore and develop a set of best practices for data mining within this context. The goal of this research question will be to develop a set of best practices for Indigenous data mining. As outlined in Section 2.3, the research team will follow the principles and guidelines of data collection and management as described by the Tri-council TCPS-2, and in the National Inuit Strategy on Research (ITK, 2018), and the FNIGC OCAP®. Data to be collected will be co-determined by the community and the research team, stored locally (where feasible) and stored in a manner that allows for access, ownership, and possession as determined by the community. Moreover, data collection will follow the principle of free prior and informed consent at all levels. Extreme care will be taken when analyzing these data, so that learned information is used solely with the full consent of community members (UNDESA Division for Inclusive Social Development, 2017). To reduce the chance for further exploitation by researchers and data collection, Indigenous communities are increasingly requiring the use of participatory and decolonizing research processes (Simonds & Christopher, 2013). The decolonization of the research and data collection processes is still being fully explored, however successful examples are becoming more frequent in the literature. One such example is that of the participatory and decolonizing approaches used by researchers with the community of Apsáalooke Reservation, Montana. In particular, the research team sought to evaluate a public health issue (i.e. cervical cancer) through the use of simple data mining techniques (i.e. clustering and classification). To accomplish this, the research team worked with the community to identify research goals, to identify the boundaries of data collection, and to outline the potential outcomes of data analyses. Moreover, they clearly identified how data could be used, as well as what could be shared and how. After the completion of 83 extensive interviews with community members, data were co-analyzed using classification methods. Findings of the study were used to inform health intervention and education protocols for the community (Christopher et. al., 2008). Other examples include Holkup et al. (2004), Ford et al. (2016), Kouril et al. (2016), and Cook, (2018). As such, we will work with community-research leads to identify research questions that support community goals specific to their community-based monitoring programs. Participatory workshops will be established to provide community with a set of data mining methods (e.g. association, classification, clustering, predictions and decision trees) to explore the research question(s), with additional focus groups created to identify and evaluate those methods most aligned with community cultures, traditions, and needs. Focus groups will also be used to explore how data might be used, which results might be shared, and how results might be shared. Finally, co-analysis will require that community-research leads (or other interested parties) participate in the implementation, analysis, and reporting of analyzed data to ensure that the process and outcomes meet the explicit priorities and aims of the indigenous community.
    • Obj2.4 Data Sharing & Management for Use in Multisectoral Decision Making [Postdoctoral Fellowship] It is expected that other communities will wish to develop RightMesh enabled mobile applications that will support the collection of data in communities with low to no connectivity. While Rigolet, Nunatsiavut has already identified their partnership with this research program, other communities have been identified as potential collaborators (e.g. Little Salmon Carmacks First Nation, Yukon, Pangnirtung, Nunavut). While the data will be community-owned, local, regional, territorial, and federal decision makers will need access to the data in some regionally aggregated form. To accomplish this, it becomes necessary to explore data sharing and data management schemes in other domains (e.g. public health, government, industry), particularly those that have Indigenous connections, to inform the development of best practices for data sharing and management in an Indigenous context. The goal of this research question will be to establish a set of best practices for sharing data and managing them given local, regional, and federal needs/requirements. The Postdoctoral Fellow will also be responsible for co-ordinating and managing the related PhD intern research questions. Further, the Postdoctoral Fellow will necessarily need to work with Indigenous community members to co-ordinate data and information sharing needs/requirements. This includes working directly with our community partners, but also establishing and fostering relationships with other communities who have already developed data management plans, or are looking to get involved in this particular space. Similar to Objective 2.3, the research team will follow the principles and guidelines of data collection and management as described by the Tri-council TCPS-2, and in the National Inuit Strategy on Research (ITK, 2018), and the FNIGC OCAP®. Data to be collected will be co-determined by the community and the research team, stored locally (where feasible) and stored in a manner that allows for access, ownership, and possession as determined by the community. To achieve this objective, it will be necessary to review data sharing and data management frameworks, and more specifically, those data sharing and management frameworks that focus on Indigenous data. Two potentially guiding examples include 1) the Canadian Polar Data Network (Scassa and Taylor, 2017), and 2) the First Nation Data Centre (https://fnigc.ca/fndc). These will be used to guide focus groups and workshops related to data sharing and management, to better understand which data are shared, and how they are shared.
  • Obj3 Evaluation of Methods for User Engagement
    To help grow the number of users who connect and participate in a mobile mesh network, it is necessary to develop a robust and varied ecosystem of engaging mesh-based applications that encourage consistent interactions over a period of time. As such, it becomes necessary to identify and evaluate various engagement methods that are appropriate for remote, rural, and Indigenous communities which can inform the development of a framework to support software design choices.

    • Obj3.1 Engagement Using Competitive Gamification Tools [PhD]
      Gamification, or the act of introducing elements of gaming to non-gaming environments, has been shown to improve engagement in several domains. This includes, for example, the domains of education, health and wellness, and in at least one case, Indigenous language training for children (Hew et al., 2016, Mekler et al., 2013, Johnson, 2016, Ramirez, 2018, Barata, 2013). However, it is unknown if these tools can be used to support the development of mesh-based applications, especially those that are used to support community-led monitoring programs in remote, rural, and Indigenous communities (e.g. Rigolet, Nunatsiavut, Pangnirtung, Nunavut, Little Salmon Carmacks First Nation, Yukon). As such, mesh apps will be developed that explore the utility of self-competition and social-competition will be developed to increase user engagement. Ultimately, the findings of this research question will be the development of best practices for designing mesh-based software and user engagement. To achieve this particular objective, gamification frameworks such as those proposed by Werbach and Hunter (2012), Marache-Francisco and Brangier (2013), or Chou (2013) will be used to guide gamification techniques selected for use in mesh-based apps designed for Indigenous communities. For example, Chou’s Octalysis framework could be used with the community to identify different types of gamification tools that best reflect Indigenous culture and tradition. Specifically, through the use of workshops, open houses, and other participatory research and design methods, and with an understanding of the specific goals of their community-based monitoring program, community members will be able to select standard competition-based gamification tools (such as leaderboards, badges, etc.) to achieve any of eight core drivers for motivating humans[1] that might lead to improved engagement and enhanced data collection for community-based monitoring programs. With this in mind, it will be necessary to first review the available literature to identify and evaluate gamification frameworks for use in Indigenous communities. Evaluation, in particular, will require the participation and expertise of our community-research leads to identify a framework that best reflects community cultures and traditions. Following this, focus groups will be used to specifically identify the self- and social-competition gamification tools most likely to improve engagement. Tools will then be developed using the participatory approaches previously described, and evaluated to determine the effect on user engagement. Engagement will be measured to quantify long-term engagement (i.e. sustained engagement and re-engagement), and to quantify high quality interactions (i.e. engagement that leads to improved data collection), as outlined in Looyestyn et al. (2017).
    • Obj3.2 Engagement Through Choice [PhD]
      There has been some success increasing user engagement using choice –providing users with multiple paths to complete a task. While this could relate to a self or social competition, the goal here is to use choice to encourage consistent and repeated engagement with a mesh-based app. This is particular relevant in remote communities such as Rigolet, where a desire to preserve and share local Inuit Knowledge with youth has been identified as a priority by the community. This research question will explore how youth might be engaged through the development of, for example, skills trees that connect students and elders as the students learn and build on traditional skills on their way to becoming experts in local Inuit tradition and cultural practice. The research question will provide a set of best practices for engaging users through choice, both in terms of developing the set of choices available, and the design process for implementing them. Like self- and social-competition, allowing user-defined paths (choices) through gamification has been shown to promote autonomy, a factor in improving user engagement (see for example, Kumar and Raghavendran, 2015, Yang et al., 2017, and Zainuddin, 2018). Using a similar process as has been outlined in Obj3.1, we will develop tools that provide users with choices as they complete tasks or gain skills. While traditional skills (e.g. hunting, fishing, harvesting) have already been identified by the community of Rigolet, we will begin the process by engaging community through open-houses and focus groups to identify the skills or tasks that best reflect community interests while supporting their community-based monitoring program. Once identified, mesh-based software will be developed (or updated) to provide users with choice. Engagement measures (as described by Looyestyn et al., 2017) will be collected and compared with engagement measures collected prior to the introduction of choice, or by establishing groups of users with and without the choice feature. The most appropriate option will be determined by consultation with community-research leads to ensure that local cultures and traditions are respected.
    • Obj3.3 The Utility of User Models [PhD] User models can be used in software design to improve user experience, and thereby encourage engagement (or at least reduce user fatigue with an app). Importantly, these models can be used to support engagement with mesh-based applications by 1) informing structural changes to the design and layout of various screens that simplify the user interface or improving the user experience through personalized customizations, or 2) providing the user with content or recommendations based on their existing use patterns. To evaluate choice, community-led mesh-based apps will be developed that support rural, remote, and Indigenous community environments, resources, and health monitoring programs. The goal of this research question will be to explore the utility of user models to improve engagement through custom interfaces and curated content. Typically, the goal of a user model is to allow the personalization of content and/or structure of software to reduce on-screen clutter, stream line processes for the user, or provide content that limits information overload (Bozdag and Timmermans, 2001, Brusilovsky, 1996, Brusilovsky, 2007, Razmerita et al., 2012, Kolias et al., 2008, Germanakos et al., 2009, Steichen et al., 2012). User models that allow for the personalization of content do not affect the navigation or methods of interacting with a system. Instead they provide content to a user that might differ from another (Brusilovsky, 1996). Personalization of structure is used to, for example, more easily guide a user to information they might need. This could mean a change in the navigational structure, or a modification to the on-screen elements presented to the user (Brusilovsky, 1996, Tsandilas et al., 2003, Brusilovsky, 2007). To achieve personalization with user models, typically two methods are used: domain-based filtering, and collaborative-based filtering. In the former case, information is recommended to users by developing a user model that assigns relevance to a resource, then recommending other resources that have similar features (e.g. recommending documents that have similar content to one that the user might have selected) (Steichen et al., 2012, Balabanovic and Shoham, 1997). Collaborative-based filtering provides recommendations to a user based on the actions and interests of other users who have similar past interests. For example, a user who has viewed several documents pertaining to a particular subject might be recommended related topics based on document requests of users with similar past interests (Goldberg, 2001, Ricci et al., 2011). To achieve this research objective, community-research leads will help to identify the most appropriate personalization goals and methods considering both the goals of their community-based monitoring program, and the cultures and traditions of the community. Methods will also be explored through participatory focus groups and workshops. Once identified, mesh-based software tools will be developed (or updated) to incorporate user models. Engagement will be measured using similar metrics outlined for Obj3.1.
    • Obj3.4 Participatory Design Methods for User Engagement [Postdoctoral Fellowship]
      Participatory design methods have been successfully used to co-create community-led monitoring programs. The methods have also been used to co-create and pilot an environment and health monitoring application in Rigolet, Nunatsiavut, Labrador to support a community-led health and environment monitoring program, to better understand and adapt to the effects of Climate Change. However, while it is expected that co-creation will lead to improved engagement with the final product (because it has been designed with the intended user, and not just for the intended user), it is unknown what effect (if any) participatory design will have on engagement. As such, it is necessary to evaluate the impact on engagement by exploring mesh-based software design in a participatory fashion. In addition to achieving the goals of this research question, the Postdoctoral Fellow will be responsible to co-ordinate and manage the PhD interns working to understand user engagement related to RightMesh enabled mobile applications.
  • Obj4 Evaluation of Cryptocurrencies for Incentivization
    The mesh network relies on the use of blockchain and cryptocurrency technologies to incentivize users to participate in the mobile mesh network. However, given the variety and growth of the blockchain domain, this project will identify and evaluate different blockchain protocols, and develop tools evaluate new blockchain technologies. There is also a need to understand how these apps might affect the performance of the network, or users’ perceptions of it. As such, there is a need to develop evaluation criteria to select the most appropriate technology and incentivization route to ensure continued user engagement, growth, and satisfaction in the network.

    • Obj4.1 Evaluation of the Blockchain Transaction Rate on Network Performance [Postdoctoral Fellowship]
      To ensure the mesh network is performing optimally despite blockchain exchanges, it becomes necessary to understand the state of blockchain technologies. Further, it becomes necessary to develop evaluation criteria relevant to the needs of Left to determine the blockchain technologies that will perform well enough to support incentivization of the data transmission in the mobile mesh network. Using these criteria as guide, a set of scenarios and implementations will be developed that can screen different blockchain technologies, to determine the strengths and weaknesses of each. The results of these simulations will be used to develop a framework for the selection of blockchain technologies in general. They will also be used to select the most appropriate technology for Left, or provide incentive to improve or develop a new blockchain specific to the Left’s mesh networks.
    • Obj4.2 Evaluation of Micropayment Channels Performance Limits, and the Effect on Network Performance Evaluation of Micropayment Channels Performance Limits, and the Effect on Network Performance [PhD]
      Because it is well known that many of the existing blockchain technologies do not scale (in terms of transaction number) while maintaining transaction consistency and security, one of the solutions proposed is micropayment channels (Scherer, 2017). A micropayment channel allows Left to bypass the expensive and time-consuming process to log transactions directly on the blockchain. Instead, micropayment channels use the smart contract capability of Ethereum so that parties exchanging payment do so “off-chain” by exchanging signed transactions of each small payment. At some later point, each party can settle the full balance of the transactions on the blockchain, rather than with each transaction. However, micropayment channels still require on-chain actions such as opening and closing the channel. The goal of this research questions is to evaluate the impact of these actions on the system, identify any limitations or bottlenecks, and determine how quickly remote transactions can be signed using commodity phones. We can obtain answers to these questions by applying the same scenarios and experiments that we apply to blockchain technologies to micropayment channels. We will also develop new scenarios to demonstrate any possible scaling problems related to creating or closing payment channels. There are also questions to be answered about the Left superpeers – one of four node types in the RightMesh network that facilitates connectivity through the management of micropayment channels. For example, how does the number of tokens held by the superpeer affect how many channels may be open at a given time? And what should the strategy be for the superpeer when it comes to opening and closing channels? Because it costs money to open a channel, all parties are motivated to keep the channels open as long as possible. However, closing a channel is the only way to recover the funds in it. Therefore, we must determine whether there are ways to circulate the funds without closing the channel.
    • Obj 4.3 Evaluation of Incentivization Methods, Economics of Token Model [PhD]
      Left has proposed various methods to incentivize the mesh. The simple solution is to incentivize the raw data, storage and processing capabilities of the phones. There are also more complex methods, such as feeding into the system externally by showing ads and distributing apps. More complex still is rewarding mobility and collecting sensor data, or distributing content. In each of these cases, there are questions about whether the users will accept these methods or how the methods might affect the overall token economy. Moreover, we do not know if tokens will continue to move within the network, move too much, or what the right balance is between token utility, inflation and scarcity. Moreover, we need to determine if the balance is affected by the type of incentivization.

Note: given the nature of the research, the projects listed are subject to change.