Research Topics/Projects

Wireless Sensor Networks Powered by Ambient Energy Harvesting (WSN-HEAP)

contact: Prof Winston Seah .

Bulk of the research on wireless sensor networks assume that sensors rely on a limited power source like batteries, and aim to maximize the lifetime of the network through efficient energy usage. When the batteries run out, a wireless sensor network is essentially useless. The ideal wireless sensor network is one in which the sensors do not have to rely on any onboard energy source. The sensors harvest the energy from the environment to power up, initialize, take readings and transmit the data. Each sensor has only sufficient harvested energy to transmit the data once (if not more), and they must coordinate among themselves to relay the data to the sink. While this approach has alleviated the energy problem, it exacerbates the synchronization problem among sensors. Synchronizing packet transmissions is very important to minimize collisions among sensors which, in our case, can be critical as there is very limited energy available. Another is knowing the neighbourhood (or locality) so that a next-hop node can be found to forward a packet to. This project focuses on the following closely related issues:
  1. node synchronization (for medium access control)
  2. locality awareness (for data forwarding)
  3. energy-efficient data forwarding
  4. cognitive methods for sensor data processing, such as, fusion, aggregation, and optimization
  5. energy-efficient modulation and coding scheme
  6. reliability mechanisms

Wireless Protocol Design for Body Area Sensor Networks in Healthcare

contact: Prof Winston Seah .

In healthcare, a number of tiny bio-sensor devices are deployed on the human body to monitor vital signs of human such as blood pressure, ECG, spO2, and movement activities. These sensors create a wireless body area network (WBAN) for transmission of sensed vital signs from the sensors to a dedicated sink. Primary batteries can have excessive weight and size that limit the lifespan and autonomy of electronic devices because of the need of replacement. This makes them unsuitable in systems with limited accessibility and the cost becomes prohibitive in wireless microsensor networks with a large quantity of powered devices. Long lifespan and small dimensions of the power source are particularly important and advantageous for wearable electronics and systems. These have stimulated worldwide research in the field of energy-harvesting devices. However, the field of energy-harvesting technologies in wearable computing is still a new area due to its specific challenges such as human safety, form factor limitations and wearability. Recent investigations have been carried out with the goal to design and develop miniaturized wearable power generators. For example, a micro thermoelectric generator can produce small, though usable power from temperature differences as low as 5 K, introducing the capability to power devices from body heat. Energy can also be captured by muscular activity so that the human body can be used as steady energy source.

When applying energy-harvesting technologies to body area networks, the energy capturing rates varying according to the energy-harvesting technologies. Furthermore, the energy of sensor nodes will be unevenly distributed throughout the body depending on the physical locations of the sensors. As wireless communications consume most of the energy of a sensor node as compared to other energy consumers such as sensing data acquisition and computing, the transmission capability of a sensor node depends on how much energy it is able to capture during its operation. In this research, two areas will be investigated: (1) the existing energy-harvesting technologies and the off-the-shelf products that may be suitable to be applied to body area networks. (2) the impact of the above technologies on the wireless protocol design, possibly through the study of energy states and distributions. Students working on this project will also have the opportunity to do an internship at the Institute for Infocomm Research, Singapore, for a period of between 3 and 6 months.

Robust end-to-end Wireless Multihop Protocols for Harsh Environments

contact: Prof Winston Seah .

In wireless multihop networks, it is common for nodes to experience constantly fluctuating link quality due to node movement, interference from the physical environment, power limitation that require communications to be minimized in order to save power, etc. When nodes move or the physical environment around the nodes changes, links may break due to obstructions, fading, and other forms of interference. In order to conserve power, it is often better to withhold the data or shutdown the link until conditions recover before resuming communications again. This results in intermittent connectivity, and even network partitioning when there is no path between a pair of communicating nodes. Some examples of such intermittently connected networks include: a) sensor networks that operate with stringent physical and environmental constraints and scheduled to be turned off periodically to conserve resources; b) military ad hoc networks where nodes (e.g. troops, vehicles, aircraft) may move randomly, may be destroyed or need to stop transmitting in order to avoid being detected; c) underwater wireless networks deployed for offshore engineering applications where the underwater acoustic communications channel is subjected to spatial and temporal blackouts; and d) vehicular networks where nodes move at high velocities. Protocols for such networks must be able to establish stable routes and also adapt quickly to changes in the environment. Failing which, the applications must first be able to tolerate delays beyond conventional IP forwarding delays, giving rise to what is commonly referred to as delay tolerant networks (DTN). This project encompasses the following aspects:
  1. Study the fundamentals of routing for wireless networks and design algorithms that select the most stable (but not necessarily shortest) routes, preferably without affecting the existing routing fabric. (See HICSS-45 paper for details on idea and preliminary results.)
  2. Study the performance of known routing protocols in delay tolerant networks, and design schemes for DTNs to meet the performance requirements of realistic application scenarios.

Green Cognitive Wireless Networks - Algorithms, Protocols & Systems

contact: Prof Winston Seah .

Wireless communications is highly dependent on the availability of reliable energy sources to operate. In recent years, the immense amount of energy consumed by the wireless communications infrastructure and devices has become a critical issue. Despite the emergence of more energy efficient technologies, we remain tethered to energy sources. Portable energy sources like batteries need to be replaced and/or recharged, and also pose environmental risks. The promise of untethered freedom by wireless communications remains hampered by the energy source needed to drive it. With the increasing concerns about global warming and the urgent need to reduce energy usage, it is therefore inevitable that we need to find alternative and more sustainable ways to deploy and operate wireless communication systems and networks. Furthermore, with the ever changing environment, networks require cognitive capabilities to adapt and evolve.

This research aims to develop new wireless communications technologies that can operate from minimal amounts of energy harvested from the environment and/or for long durations without the need to replace the portable sources (batteries) thus minimizing environmental damage from human intervention during the process of energy supply or replenishment. Ideally, such wireless communications networks will operate without the need for any human intervention and remain sustainable for years to come. Some of the areas to be studied include:
  1. More efficient ways of harnessing and utilizing the ambient energy that is present in the environment to power wireless communications. Research on the development of energy harvesting technology is not studied here, but rather, the focus is on the design of algorithms and protocols that fully utilize the scarce amounts of harvested energy.
  2. New wireless networking technologies that reduce energy consumption and improve energy efficiency by minimizing, e.g. unnecessary periodic transmissions like beacons, persistent channel access techniques, data redundancy, etc.
  3. Applying environmentally friendly metrics for route discovery and selection in wireless communication networks.
  4. Cognitive methods that learn and evolve to adapt to dynamic environmental conditions and network traffic loads.

Communications Architecture for Network of Unmanned Autonomous Swarms (CANUAS)

contact: Prof Winston Seah .

The use of many small low-cost unmanned air/ground/underwater vehicles (UAV/UGV/UUV or collectively as UxV) working as a collective group is a viable approach for surveillance and detection of potential threats/obstacles in harsh environments, e.g. underground, underwater, disaster zones, and battlefields. Effective communication mechanisms in such groups of unmanned vehicles (UxVs) are a key requirement for their meaningful deployment. A fully distributed communications architecture is necessary for reliability and robustness, especially since the communications system needs to support both the transport of data among the UxVs, as well as, commands that influence the motion of the UxVs; the latter is very critical for ensuring that the group of UxVs maintain formation while performing their tasks, and the commands must reach their desired destinations (nodes) within the given time constraints, so that the targeted nodes have sufficient time to act on the command. An example scenario can be a UxV of the group straying away and nearest UxVs transmitting commands to direct it back towards the group. In this project, we focus on designing protocols that ensures high-priority messages like motion commands are given precedence to transmit over the wireless medium. Given the stringent time constraints, a time-scheduled mechanism appears to be a more viable approach. The aim of this project is to study the state-of-the-art in networking schemes for such harsh environments, and design a suitable scheme for swarm control and communications.

Theoretical Approaches in Wireless Communications Networks and Systems Design

contact: Prof Winston Seah .

Traditional networks are built to cooperate based on a mandatory network communication semantic to achieve desirable qualities such as efficiency and scalability. With technological maturity and widespread technical know-how, a different set of network problems emerged - clever users try to alter network behavior in a way to benefit themselves at the expense of others. The problem is more pronounced in mobile ad hoc networks (MANET) where network ownership can be largely public. At one extreme is the malicious node that aims to disrupt the operation of the network. We focus instead on selfish, rational user misbehavior while keeping this danger in mind. In this project, pricing and promiscuous listening are avoided because of known technical feasibility in actual deployment scenarios. While adopting the punishment approach, we avoid network-wide punishment as it is an easy avenue for denial-of-service (DoS) attacks. Some approaches for study include:
  1. Model based on the work of Masaki Aoyagi's imperfect private monitoring for the dynamic Bertrand oligopoly, and fit it to wireless multi-hop networks. The model further relates the relevance of routing broadcasts and packet acknowledgments to network cooperation.
  2. Apply game theory to study the performance of bandwidth sharing in communal networks where each member of the network is expected to its fair share of network resources in order to gain access to other members' network resources, and most importantly, discourage unfair and selfish behaviours.
  3. Model the performance of wireless multihop networks for network and civil resource optimization and pricing. This can be applied to the realistic scenarios of networking resources that support intelligent transportation systems, emergency services, logistics, etc.

Exploiting Radio Irregularity in Wireless Sensor Networks

contact: Prof Winston Seah .

Wireless communications, which is an integral part of IoT, suffers from radio irregularity -- a phenomenon referring to radio waves being selectively absorbed, reflected or scattered by objects in their paths, e.g., human bodies that comprises liquid, bone and flesh. Radio irregularity is often regarded as a problem in wireless communications but, with the envisioned pervasiveness of IoT, we aim to exploit radio irregularity as a means to detect the presence of people, animals and other objects. Applications for this technology, include motion/instrusion detection for security and surveillance, automated people counting, wildlife detection and tracking for determining absolute abundance or pest control, vehicular traffic monitoring, etc.