Search

Introduction Mild cognitive impairment (MCI) is characterized by subtle deficits that functional assessment via informant‐report measures may not detect. Sensors can potentially detect deficits in everyday functioning in MCI. This study aims to establish feasibility and acceptability of using sensors in a smart home for performance‐based assessments of two instrumental activities of daily living (IADLs). Methods Thirty‐five older adults (>65 years) performed two IADL tasks in a smart home laboratory equipped with sensors and a web camera. Participants’ cognitive states were determined using published criteria including measures of global cognition and comprehensive neuropsychological test batteries. Selected subtasks of the IADL assessment were autonomously captured by the sensors. Total time taken for each task and subtask were computed. A point scoring system captured accuracy and number of attempts. Acceptability of the smart home setup was assessed. Results Participants with MCI (n = 21) took longer to complete both tasks than participants with healthy cognition (HC; n = 14), with significant time differences observed only in "Cost calculation." Completion time for IADL tasks and scores correlated in the expected direction with global cognition. Over 95% of the participants found the smart home assessment acceptable and a positive experience. Discussion We demonstrated the feasibility and acceptability of the use of unobtrusive commercially available sensors in a smart home for facilitating parts of the objective assessment of IADL in older adults. Future studies need to identify more IADLs that are suitable for semi‐automated or automated assessments through the use of simple, low‐cost sensors.

Singapore faces a major challenge in providing care and support for senior citizens due to its rapidly ageing population and declining old-age support ratio. The concept of Ageing-in-Place was introduced by the Singapore government [1] to allow older people to live independently in their own homes and communities so that the need for institutionalised care will only be utilised when necessary. We have three fundamental questions that this project will answer: 1. How to make community care services effective through innovations in care delivery? How to lower the cost of service delivery and improve 2. productivity of caregivers, by leveraging information and communications technology (ICT)? 3. Can we quantify such productivity gains?

The selection of a path in a network from one node to another is performed by a routing protocol, often using the shortest path approach. However, it has been shown that in wireless networks, the shortest path approach tends to select paths with wireless links that are only intermittently available and often unstable, resulting in extensive loss of packets in the data delivery process. Instead of taking the well-trodden approach of designing a new protocol or new routing metric, we propose a method that influences the perception of the network topology for different network nodes with the aim of presenting to the routing protocol a network topology that comprises more stable wireless links. We refer to this method as the {$\delta$}-scheme (``delta'' scheme). We implement a proof-of-concept prototype of the {$\delta$}-scheme using commercial-off-the-shelf routers and demonstrate through experimentation the viability of this scheme in a realistic scenario.

In wireless networks, it is important to determine the outcome of packet transmissions for networking protocols. In this paper, we design a transmission outcome classifier for IEEE 802.15.4 wireless networks based on received signal strength indicator and link quality indicator values. Our classifier performs loss differentiation by analyzing statistical differences between weak signal and collision losses. We implement our proposed classifier using the CC2500 RF transceiver and evaluate it experimentally. The results show that our classifier can accurately detect packet transmissions as well as distinguish wireless losses due to weak signals and multiple access collisions, with a maximum error rate of 15%. We apply the classifier to probabilistic polling, which is a MAC protocol designed for energy harvesting wireless sensor networks, and show experimentally that it is able to achieve close to or even exceed the theoretical throughput due to packet capture effect.

This paper proposes XOR-BiDO, a data delivery scheme (DDS) suited for multihop underwater acoustic networks characterized by regular bidirectional traffic streams. DDSs for multihop underwater networks usually assume a single-sink configuration, where most of the transmissions flow in a single direction. Throughput performance decreases and fairness suffers when such DDSs are subjected to heavy bidirectional traffic streams due to increase of queueing delays at relay nodes. This paper proposes the integration of an opportunistic XOR network coding layer into BiDO, a scheme previously designed for single-sink underwater networks. This coding layer actively and opportunistically mixes data frames whenever a coding opportunity arises. By effectively transmitting more data in fewer transmissions at each relay node, the result is a scheme that (1) maximizes usage of the underwater acoustic channel's scarce bandwidth, (2) provides better fairness for bidirectional traffic and (3) improves energy-efficiency of submerged devices, while retaining the opportunistic-routing and network-wide overhearing features of BiDO. Using simulations, we demonstrate the network performance improvement of XOR-BiDO in terms of packet latency, fairness and energy-efficiency for bidirectional network traffic. While maintaining comparable reliability over a 10-hop network, XOR-BiDO achieves a 12% improvement in latency and consumes 24% less energy per delivered packet against BiDO.

Energy harvesting wireless sensor networks (EH-WSNs) are gaining importance in smart homes, environmental monitoring, health care and transportation systems, since they enable much longer operation time as energy can be replenished through energy harvesting. This is unlike sensor nodes that use non-rechargeable batteries which need to be replaced once energy is depleted. However, the sporadic availability of ambient energy makes the design of networking protocols and predicting network performance very challenging. In this paper, we perform an empirical energy characterization of a time-slotted solar energy harvesting node with different system and environmental parameters. We use six different statistical models (uniform distribution, geometric distribution, transformed geometric distribution, Poisson distribution, transformed Poisson distribution and a Markovian model) to fit the empirical datasets. Our results show that there is no single statistical model that can fit all the datasets, thus justifying the need to use empirical data to validate the theoretical analysis of time-slotted MAC protocols for EH-WSNs.

In an environmentally-powered wireless sensor network, where nodes are powered solely by harvesting energy from ambient sources, node operation highly depends on the energy availability and harvesting rate. For duty-cycling schemes to support existing wireless sensor network applications, they need to adapt the nodes' sleep-wake schedules according to energy harvesting and consumption rates. In this paper, we propose a harvesting-aware duty-cycling scheme, perform an empirical study of this scheme over a solar harvesting powered wireless sensor network in a source-relay-sink configuration, and provide some experimental results to illustrate its throughput performance under various light conditions.

Motion and intrusion detection are often cited among various Wireless Sensor Network (WSN) applications. A typical configuration comprises clusters of wireless nodes equipped with motion sensors to detect human motion. Currently, the performance of WSN is subject to several constraints, mainly the phenomenon of radio irregularity and finite onboard computation/energy resources. In Radio Frequency (RF) propagation, radio irregularity rises to a higher level in the presence of human activity due to the absorption effect of the human body. In this paper, the feasibility of monitoring RF transmission for the purpose of intrusion detection is investigated. With empirical data obtained from the Crossbow TelosB platform in several different environments, the impact of human activity on the signal strength of RF signals in a WSN is evaluated. This paper offers a novel approach to intrusion detection by turning a constraint in WSN, namely radio irregularity, into an advantage for the purpose of intrusion detection. Unlike most related work, the “intruders” neither transmit nor receive any RF signals. By enabling existing wireless infrastructures to serve as intrusion detectors instead of deploying numerous costly sensors, this approach shows great promise for providing novel solutions.

A wireless multihop network is a communications network composed of nodes equipped with wireless interfaces and organized in an ad hoc manner. It has become an attractive and practical solution for providing flexible and extended wireless coverage over large areas. Measurement of physical layer quantities can provide useful information in the design of networking protocols and cross-layer optimization. In this paper, we present a cost-effective approach of performing physical-layer and link-layer measurements by exploiting the capabilities of inexpensive commercial-off-the-shelf (COTS) wireless routers and open-source software tools. Our experimental results show that the measurements made using inexpensive wireless routers are consistent with theoretical models and hence reliable. It is therefore possible to utilize the measurements to devise models or algorithms that will improve the performance of networking protocols.

The use of wireless sensor networks (WSNs) for structural health monitoring is gaining popularity since it allows for a low-cost, rapid and robust assessment of structural integrity. Meanwhile, recent advances in ambient energy harvesting technology have made it a viable alternative source of energy for powering WSNs. WSNs powered by ambient energy harvesting (WSN-HEAP) are potentially moreuseful and economical in the long term than traditionalbattery-powered WSNs as they can operate for very long periods of time without the need for human involvement, thus paving the way towards alleviating energy constraints that continue to challenge WSNs. In this paper, we evaluate the impact of transmit power control on the usefulness of a multi-sink WSN-HEAP, deployed in uniform string topology for railway track monitoring. Based on current achievable energy harvesting rates from track deflections, and commercially available sensor mote parameters, our analysis reveals that availability can be maximised while maintaining good data delivery ratio and throughput-fairness by appropriate setting of the transmit power over a wide range of deployment density, Signal-to-Noise Ratio requirements and energy harvesting characteristics.

Underwater acoustic networking is an emerging technology platform for oceanographic data collection, pollution monitoring, offshore exploration and tactical surveillance applications. Design of reliable and efficient communications protocols is challenging due to the unique characteristics of underwater acoustic channels. In this paper, we present a modular and lightweight implementation of an opportunistic multihop automatic repeat request (ARQ) scheme in a real system. We evaluate the performance of the opportunistic ARQ using inexpensive underwater acoustic modems in a shallow underwater environment.

Motion and intrusion detection are often cited as wireless sensor network (WSN) applications with typical configurations comprising clusters of wireless nodes equipped with motion sensors to detect human motion. Currently, WSN performance is subjected to several constraints, namely radio irregularity and finite on-board computation/energy resources. Radio irregularity in radio frequency (RF) propagation rises to a higher level in the presence of human activity due to the absorption effect of the human body. In this paper, we investigate the feasibility of monitoring RF transmission for the purpose of intrusion detection through experimentation. With empirical data obtained from the Crossbow TelosB platform in several different environments, the impact of human activity on the signal strength of RF signals in a WSN is evaluated. We then propose a novel approach to intrusion detection by turning a constraint in WSN, namely radio irregularity, into an advantage for the purpose of intrusion detection, using signal fluctuations to detect the presence of human activity within the WSN. Unlike RF fingerprinting, the 'intruders' here neither transmit nor receive any RF signals. By enabling existing wireless infrastructures to serve as intrusion detectors instead of deploying numerous costly sensors, this approach shows great promise for providing novel solutions.