Neighbor Discovery (ND) is the procedure employed by wireless devices to establish a first contact. All ND protocols involve devices sending beacons, and also listening for them. Protocols differ in terms of how the beacon transmissions and reception windows are scheduled, and the device sleeps in between consecutive transmissions and reception windows in order to save energy. A successful discovery constitutes a sending device’s beacon coinciding with a receiving device’s reception window. The goal of all ND protocols is to minimize the discovery latency. In spite of the ubiquity of ND protocols and active research on this topic for over two decades, the basic question “Given a power budget, what is the minimum guaranteed ND latency?”, however, has still remained unanswered. This paper is on the best-achievable ND latency for a given power budget between a pair of devices. In order to compute this lower bound, we introduce a concept called coverage maps, that allows us to analyze the ND procedure in a protocol-independent manner. Using it, we derive discovery latencies for different scenarios, e.g., when both devices have the same or different power budgets. We also show that some existing protocols can be parametrized such that they perform optimally. Our results are restricted to the case when a few devices discover each other at a time, as is the case in most real-life scenarios, while scenarios with large numbers of devices need further study.

Using WiFi's Channel State Information for human activity recognition—referred to as WiFi sensing—has attracted considerable attention. But despite this interest and many publications over a decade, WiFi sensing has not yet found its way into practice because of a lack of robustness of the inference results. In this paper, we quantitatively show that even “microchanges” in the environment can significantly impact WiFi signals, and potentially alter the ML inference results. We therefore argue that new training and inference techniques might be necessary for mainstream adoption of WiFi sensing.

In the context of smart buildings and smart cities, the design of low-cost and privacy-aware solutions for recognizing the presence of humans and their activities is becoming of great interest. Existing solutions exploiting wearables and video-based systems have several drawbacks, such as high cost, low usability, poor portability, and privacy-related issues. Consequently, more ubiquitous and accessible solutions, such as WiFi sensing, became the focus of attention. However, at the current state-of-the-art, WiFi sensing is subject to low accuracy and poor generalization, primarily affected by environmental factors, such as humidity and temperature variations, and furniture position changes. Such is-sues are partially solved at the cost of complex data preprocessing pipelines. In this paper, we present a highly accurate, resource-efficient deep learning-based occupancy detection solution, which is resilient to variations in humidity and temperature. The approach is tested on an extensive benchmark, where people are free to move and the furniture layout does change. In addition, based on a consolidated algorithm of explainable AI, we quantify the importance of the WiFi signal w.r.t. humidity and temperature for the proposed approach. Notably, humidity and temperature can indeed be predicted based on WiFi signals; this promotes the expressivity of the WiFi signal and at the same time the need for a non-linear model to properly deal with it.

Identity recognition is increasingly used to control access to sensitive data, restricted areas in industrial, healthcare, and defense settings, as well as in consumer electronics. To this end, existing approaches are typically based on collecting and analyzing biometric data and imply severe privacy con-cerns. Particularly when cameras are involved, users might even reject or dismiss an identity recognition system. Furthermore, iris or fingerprint scanners, cameras, microphones, etc., imply installation and maintenance costs and require the user's active participation in the recognition procedure. This paper proposes a non-intrusive identity recognition system based on analyzing WiFi's Channel State Information (CSI). We show that CSI data attenuated by a person's body and typical movements allows for a reliable identification - even in a sitting posture. We further propose a lightweight deep learning algorithm trained using CSI data, which we implemented and evaluated on an embedded platform (i.e., a Raspberry Pi 4B). Our results obtained using real-world experiments suggest a high accuracy in recognizing people's identity, with a specificity of 98% and a sensitivity of 99%, while requiring a low training effort and negligible cost.

Neighbor discovery (ND) protocols are used for establishing a first contact between multiple wireless devices. The energy consumption and discovery latency of this procedure are determined by the parametrization of the protocol. In most existing protocols, reception and transmission are temporally coupled. Such schemes are referred to asslotted, for which the problem of finding optimized parametrizations has been studied thoroughly in the literature. However, slotted approaches are not efficient in applications in which new devices join the network gradually and only the joining devices and a master node need to run the ND protocol simultaneously. For example, this is typically the case in IoT scenarios or bluetooth low energy (BLE) piconets. Here,slotlessprotocols that decouple reception and transmission can achieve significantly lower worst-case latencies than slotted ones. In this paper, we study slotless, BLE-like protocols, which schedule receptions and transmissions independently using periodic intervals (PI). For this class of protocols, optimal parameter values remain unknown. To address this, we propose an optimization framework for PI-based protocols, which translates any specified duty-cycle (and therefore energy budget) into a set of optimized parameter values. We show that the parametrizations resulting from one variant of our proposed scheme are optimal when one receiver discovers one transmitter, and no other parametrization or ND protocol – neither slotted nor slotless – can guarantee lower discovery latencies for a given duty-cycle in this scenario. Since the resulting protocol utilizes the channel more aggressively than other ND protocols, beacons will collide more frequently. Hence, due to collisions, the rate of successful discoveries gracefully decreases for larger numbers of devices discovering each other simultaneously. We also propose a scheme for configuring the BLE protocol (and not just BLE-likeprotocols). Though it is not clear whether the resulting parametrizations minimize the latencies of BLE, reasonably low worst-case latencies can be guaranteed.

Hearing screening devices emit an acoustic signal in the outer ear, which invokes a specific response from a healthy inner ear. However, the high cost of such devices prevents widely deploying them in schools or private homes, especially in developing countries. In this paper, we for the first time show that such tests are also feasible with a device that consists of only one speaker for emitting the signal and using the same speaker - now as a microphone - for also recording the response. Existing devices rely on a speaker and microphone pair, which makes them significantly more complex and costly. We further outline the embedded systems and signal processing challenges that such a setup entails. If successful, it has the potential to make hearing screening available to a much wider population in developing countries.

Power management is quintessential to the successful deployment of edge devices, such as smartphones, in power-, thermal-, and energy-constrained environments. Governors and schedulers operate system sub-routines for power management at the edge. There exist several tools for debugging power issues in Android applications. However, there exists no tool to identify and classify inevitable misdecisions by power managers, given their often inefficient underlying heuristics. In this work, we introduce the first tool - BrezeFlow - designed for unified (scheduling and frequency scaling) power debugging of CPU power managers on Android edge devices. BrezeFlow enables kernel developers to evaluate designs of their power managers retrospectively with closed-source applications in real-world scenarios based on any user-defined strategy and thereby gain insights for better future governor designs. BrezeFlow detected an average of 815 misdecisions per second for the commonly deployed duo, ondemand governor and Completely Fair Scheduler, on mobile edge devices running popular applications.

In many wireless control networks, sensor data and controller data are exchanged periodically, which requires periodic packet transmissions between the physical plant and the controller. As an alternative, event-triggered control paradigms imply that data is only exchanged when there are significant changes in the state of the plant, e.g., because of disturbances. This is the nature of many IoT scenarios and requires that a receiving device has to listen to the channel for incoming packets during all times. However, especially in mobile networks, in which all devices are battery-powered, continuous scanning would drain the battery quickly and hence, reception needs to be duty-cycled. When optimizing such duty-cycled operation, significant energy savings are possible using intelligent software-enabled communication scheduling. In this paper, we propose a wireless transmission scheme that supports loosely time-triggered control. When optimizing the scheduling of transmissions and reception windows in the communication protocol, our proposed scheme allows for energy-efficient communication without requiring strict clock-synchronization between the devices. We show that such a scheme is practical and can greatly reduce the energy consumption in event-triggered control applications.

Bluetooth Low Energy (BLE) is a wireless protocol optimized for low-power communication. To design energy-efficient devices, the protocol provides a number of parameters that need to be optimized within an energy, latency, and throughput design space. Therefore, an energy model that can predict the energy consumption of a BLE-based wireless device for different parameter value settings is needed. As BLE differs from the well-known Bluetooth Basic Rate (BR) significantly, models for Bluetooth BR cannot be easily applied to the BLE protocol. In past years, there have been a couple of proposals on energy models for BLE. However, none of them can model all the operating modes of the protocol. This article presents an energy model of the BLE protocol, which allows the computation of a device’s power consumption in all possible operating modes. To the best of our knowledge, our proposed model is not only one of the most accurate ones known so far (because it accounts for all protocol parameters), but it is also the only one that models all the operating modes of BLE. Based on this model, guidelines for system designers are presented that help choose the right parameters for optimizing the energy consumption. The model is publicly available as a software library for download.

Wireless devices are becoming increasingly pervasive in our everyday life and hence, robust wireless connectivity is a crucial requirement for a host of applications. In particular, with the IoT becoming a reality, an increasing number of wireless devices are frequently brought into their range of reception simultaneously, which need to interact with each other in an ad-hoc fashion. While effective solutions for reliable connected communication exist, establishing a first contact between devices remains the Achilles' heel of mobile wireless devices. For a reliable operation, every device needs to discover all devices in its range within bounded time. Thus, wireless protocols that bound the time within which devices are discovered need to be designed. However, even in such protocols, it is inevitable that a certain fraction of packets sent by multiple devices discovering each other simultaneously collide. As a result, some discoveries fail and a dependable operation can only be realized if the fraction of failed discoveries is kept negligibly low, even in worst-case scenarios. In addition, most devices are battery-powered and rely on a low-power operation for achieving reasonably long battery lifetimes. In this paper, we discuss how wireless protocols can be designed to realize a highly reliable and dependable connection setup, while at the same time meeting the energy-constraints of mobile IoT devices. Towards this, we analyze the challenges in protocol design and discuss which properties of the wireless hardware impact the robustness of the connection setup procedure of IoT devices.

Mobile devices apply neighbor discovery (ND) protocols to wirelessly initiate a first contact within the shortest possible amount of time and with minimal energy consumption. For this purpose, over the last decade, a vast number of ND protocols have been proposed, which have progressively reduced the relation between the time within which discovery is guaranteed and the energy consumption. In spite of the simplicity of the problem statement, even after more than 10 years of research on this specific topic, new solutions are still proposed even today. Despite the large number of known ND protocols, given an energy budget, what is the best achievable latency still remains unclear. This paper addresses this question and for the first time presents safe and tight, duty-cycle-dependent bounds on the worst-case discovery latency that no ND protocol can beat. Surprisingly, several existing protocols are indeed optimal, which has not been known until now. We conclude that there is no further potential to improve the relation between latency and duty-cycle, but future ND protocols can improve their robustness against beacon collisions.

Neighbor discovery is the procedure in which two wireless devices initiate a first contact. In low power ad-hoc networks, radios are duty-cycled and the latency until a packet meets a reception phase of another device is determined by a random process. Most research considers slotted protocols, in which the points in time for reception are temporally coupled to beacon transmissions. In contrast, many recent protocols, such as ANT/ANT+ and Bluetooth Low Energy (BLE) use a slotless, periodic-interval based scheme for neighbor discovery. Here, one device periodically broadcasts packets, whereas the other device periodically listens to the channel. Both periods are independent from each other and drawn over continuous time. Such protocols provide 3 degrees of freedom (viz., the intervals for advertising and scanning and the duration of each scan phase). Though billions of existing BLE devices rely on these protocols, neither their expected latencies nor beneficial configurations with good latency-duty-cycle relations are known. Parametrizations for the participating devices are usually determined based on a “good guess”. In this paper, we, for the first time, present a mathematical theory which can compute the neighbor discovery latencies for all possible parametrizations. Further, our theory shows that upper bounds on the latency can be guaranteed for all parametrizations, except for a finite number of singularities. Therefore, slotless, periodic interval-based protocols can be used in applications with deterministic latency demands, which have been reserved for slotted protocols until now. Our proposed theory can be used for analyzing the neighbor discovery latencies, for tweaking protocol parameters and for developing new protocols.

The process of two wireless devices meeting over-the-air for the first time is referred to as neighbor discovery. In mobile ad-hoc networks, battery powered devices duty-cycle their radios during neighbor discovery. As a result, they transmit and receive for very short durations of time and sleep at other times. Energy-efficient protocols, which guarantee short, bounded latencies while achieving low energy-consumptions are highly important for long battery lifetimes. In the past, neighbor discovery has been carried out mostly using slotted protocols, which subdivide time into multiple, equal length periods, called slots. An alternative are slotless protocols, which decouple beaconing from listening and can potentially achieve lower latency-duty-cycle-relations. As in slotted protocols, they also guarantee bounded latencies. However, understanding the mechanisms that ensure these deterministic bounds is more complex than for slotted protocols, since they rely on less intuitive concepts. In this demo, we propose a setup that visualizes the operation of two radios with slotless protocols in real-time, thereby providing insights that help in understanding slotless neighbor discovery. This demo is supposed to accompany the paper entitled "Griassdi: Mutually Assisted Slotless Neighbor Discovery Protocols", which appeared at IPSN 2017 as a regular paper.

Recent results show that slotless, purely-interval based neighbor discovery protocols, in which time is assumed to be continuous, achieve significantly lower worst-case discovery latencies than time-slotted protocols. In slotted protocols, the discovery of device A by B and vice-versa occurs within the same slot, and hence the latencies for one-way and two-way discovery are identical. However, in purely interval-based protocols, these latencies are independent from each other, leading to longer mean latencies for two-way discovery. In this paper, we propose a cooperative approach to reduce this two-way discovery latency. In particular, each side broadcasts information on the time-period until its next reception phase takes place. The remote device adjusts its beacon schedule accordingly once a first packet is received. Compared to non-cooperative slotless protocols, this technique can reduce the two-way discovery latency by up to 43 %. We propose a theory to model such protocols and show that with an optimized schedule, our proposed protocol achieves considerably shorter mean latencies than all known protocols, while still guaranteeing worst-case latencies that are similar to the best known solutions. For example, compared to Searchlight-Striped, our proposed protocol achieves by up to 89 % lower mean latencies and by up to 86 % lower worst-case latencies.

With the Internet of Things (IoT) becoming a reality, power-efficient techniques are crucial to achieve sufficient battery lifetimes. Whereas current medical IoT devices typically acquire data with a constant quality, we propose an architecture that dynamically adjusts the data quality adaptively based on the current medical condition of the subject being monitored. Since transmission and processing make up a large fraction of the energy consumption, the reduction of the link traffic and processing effort caused by such an adjustment results in a decreased energy consumption of the devices. For example, if anomalies in the monitored data are detected, the monitoring is performed with an increased granularity and more exhaustive processing. Further, not all data generated by the medical sensors needs to be transmitted during all times. Only if certain events are detected, the transmission of the complete data needs to be activated. In this paper, we present a novel approach for body-worn medical IoT devices. In particular, a generic, distributed architecture for the power-management of the whole system, which is based on dynamically switching services, is presented. We show that such an architecture can reduce the energy-consumption of medical sensors by up to 80 % in real-world measurements.

Reducing the energy consumption to the minimum is a crucial design requirement for all body area sensor networks. Sensors deployed on the human body, especially at the limbs often move along different positions. Usually, the transmit power is set to a sufficiently high value to achieve reliable transmission for the constellation with highest attenuation. For periodic movements, data transmission can be carried out at the position of the lowest path loss between the sender and the receiver, provided this position can be reliably identified. We propose a novel framework that predicts this position using acceleration data and the received signal strength. By learning a correlation between these signals, accurate predictions can be performed and up to 24.7% of the power spent by a Bluetooth Low Energy module for the transmission of a packet can be saved while still achieving the same packet error rate as with sending using the higher transmit power.

Bluetooth Low Energy is a time-slotted wireless protocol aimed towards low power communication for battery-driven devices. As a power-management capability, whenever there is less data to send, the slave is allowed to remain in a low power mode during a given number of time-slots in a row. However, since the master does not know the exact sleep behavior of the slave, it has to wake-up at every time-slot and repeat its packets until the slave is awake. As a result, applications with variable throughput lead to many energy-consuming idle-slots at the master. In such applications, usually the connection parameters are chosen considering the worst case at design time and remain constant during operation. In this paper, we propose a novel power-management framework for BLE. Rather than skipping slots at the slave side, the proposed system updates the interval between two consecutive time-slots during runtime by applying online algorithms. To avoid data-loss or high delays, the framework guarantees that constraints on latency are met and buffers never overflow. Energy measurements of three different test-cases show that up to 42 percent of the energy consumption of a BLE master can be saved with our power management system.

In this paper, we present Smart2, an advanced smartphone charger that mitigates battery's capacity fading, which until now has usually been ignored. Smart2 exploits the fact that many users charge their phones over night. Since the overnight charging duration is unnecessarily long, the battery is subjected to a high average state of charge (SOC), which accelerates battery aging. Therefore, we delay the charging adaptively to be done shortly before the phone is unplugged. With this scheme, clearly when averaged over the duration of the night, the average SOC is lower and hence aging is reduced. Indicators are a set alarm clock and/or statistics of previous usage. Similarly, we lower the maximum target SOC. To enable this, the main challenges are firstly to find a solution that does not negatively influence the usability and secondly to quantify the achieved savings in terms of aging mitigation. Towards this, we propose a novel charging scheme which can be implemented in the smartphone's firmware. Furthermore, we propose a modified battery charging device that can be used with almost all existing smart phone models. Using our proposed techniques, the average battery cycle life can be nearly doubled from 3.7 to 6.6 years.

With the rapid progress in semiconductor technology and the shrinking of device geometries, the resulting processors are increasingly becoming prone to effects like aging and soft errors. As a processor ages, its electrical characteristics degrade, i.e., the switching times of its transistors increase. Hence, the processor cannot continue error-free operation at the same clock frequency and/or voltage for which it was originally designed. In order to mitigate such effects, recent research proposes to equip processors with special circuitry that automatically adapts its clock frequency in response to changes in its circuit-level timing properties (arising from changes in its electrical characteristics). From the point of view of tasks running on these processors, such autonomic frequency scaling(AFS) processors become slower as they gradually age. This leads to additional execution delay for tasks, which needs to be analyzed carefully, particularly in the context of hard real time or safety-critical systems. Hence, for real-time systems based on AFS processors, the associated schedulability analysis should be aging-aware which is a relatively unexplored topic so far. In this paper we propose a schedulability analysis framework that accounts such aging-induced degradation and changes in timing properties of the processor, when designing hard real-time systems. In particular, we address the schedulability and task mapping problem by taking a lifetime constraint of the system into account. In other words, the system should be designed to be fully operational (i.e., meet all deadlines) till a given minimum period of time (i.e., its lifetime). The proposed framework is based on an aging model of the processor which we discuss in detail. In addition to studying the effects of aging on the schedulability of real-time tasks, we also discuss its impact on task mapping and resource dimensioning.