Trajectory similarity search is a basic operation in spatial data analysis for nonterrestrial networks applications, such as asset tracking, maritime transportation, and urban air mobility. In many real-world applications (e.g., Flickr, Bikely, and Foursquare), users are allowed to share comments or experiences along with their trajectories, called semantic trajectories. In this article, we study a new type of similarity search on semantic trajectories, which takes into account spatial and textual information together. Given a query trajectory$t_{q}$and a similarity threshold$\theta $, semantic trajectory similarity search retrieves all trajectories whose spatial-textual similarity to$t_{q}$is no greater than$\theta $. To address the problem efficiently, we first develop a novel bitmap-based index, called SET-tree, to organize the semantic trajectory hierarchically. It contains two types of bitmaps, global gram bitmap and positional gram bitmap, that enable us to efficiently prune the search space by spatial and textual similarity simultaneously. Besides, we develop some novel upper and lower bounds on textual and spatial similarity for pruning. We provide the spatial-side enhancement and the textual-side enhancement for further improvement. Finally, we report on extensive experiments using real and synthetic trajectory data sets that offer insight into the performance of our algorithm, and show that our algorithm is capable of achieving superior efficiency and good scalability.

Maximizing Influence (Max-Inf) query is a fundamental operation in spatial data management. Given a set of weighted objects, this query aims to find an optimal location from a candidate set to maximize itsinfluence, which is the total weight of its reverse nearest neighbors. Existing work commonly assumes that every object is in a fixed location. In real life, however, there are a wide variety of drive-in services (e.g., food joints, pharmacies, ATMs, etc.) that are widely accessed by mobile users (i.e., trajectories) instead of the fixed ones. In this paper, we first define the Maximizing Influence query over Trajectories, namely, MIT query, which aims to find an optimal location to maximize the total weight of influenced trajectories. We propose a novel index, QB-tree to hierarchically group trajectories with similar activity regions together for subsequent unified processing, and classify trajectories inside the same node into multiple buckets according to their motion patterns. For each bucket, we construct a rectilinear polygon using the trajectories in it to exclude some irrelevant areas in the minimum boundary rectangle. Moreover, we develop a branch-and-bound approach called BBM to efficiently solve the MIT query. The algorithm adaptively partitions the candidates into disjoint regions and prunes the regions without containing optimal results. Then, by exploiting the QB-tree, the upper and lower bounds are efficiently computed with three-level pruning technique. Practically, we also study a variant of the MIT query, called MDT query. We propose novel pruning bounds in cooperation with QB-tree to answer MDT queries efficiently. Finally, extensive experiments on real and synthetic datasets demonstrate that our index and algorithms have high performance in terms of efficiency, scalability, and genericity.

Maximizing Influence (Max-Inf) query is a fundamental operation in spatial data management. Given a set of weighted objects, this query aims to find an optimal location from a candidate set to maximize its influence, which is the total weight of its reverse nearest neighbors. Existing work commonly assumes that every object is in a fixed location. In real life, however, there are a wide variety of drive-in services (e.g., food joints, fuel stations, ATMs, etc.) that are widely accessed by mobile users (i.e., trajectories) instead of the fixed ones. It is urgent and challenging to solve the Max-Inf query in trajectory data (MIT). In this paper, we first define the MIT query which aims to find the optimal location to maximize the total weight of influenced trajectories. We propose a novel index structure, QB-tree to hierarchically group trajectories with similar activity regions together for subsequent unified processing, and classify trajectories inside the same node into multiple buckets according to their motion patterns. For each bucket, we construct a rectilinear polygon using the trajectories in it to exclude some irrelevant areas in the minimum boundary rectangle. Moreover, we develop a branch-and-bound approach called BBM to efficiently solve the MIT query. The algorithm adaptively partitions the candidates into disjoint regions and prunes the regions without containing optimal results. Then, by exploiting the QB-tree, the upper and lower bounds are efficiently computed with three-level pruning technique. Finally, we conduct extensive experiments on real and synthetic datasets to evaluate our index and algorithms, and the experimental results demonstrate that our algorithm has high performance in terms of efficiency, scalability, and genericity.

A fundamental problem of every intermittently-powered sensing system is that signals acquired by these systems over a longer period in time are also intermittent. As a consequence, these systems fail to capture parts of a longer-duration event that spans over multiple charge-discharge cycles of the capacitor that stores the harvested energy. From an application's perspective, this is viewed as sporadic bursts of missing values in the input data - which may not be recoverable using statistical interpolation or imputation methods. In this paper, we study this problem in the light of an intermittent audio classification system and design an end-to-end system - SoundSieve - that is capable of accurately classifying audio events that span multiple on-off cycles of the intermittent system. SoundSieve employs an offline audio analyzer that learns to identify and predict important segments of an audio clip that must be sampled to ensure accurate classification of the audio. At runtime, SoundSieve employs a lightweight, energy- and content-aware audio sampler that decides when the system should wake up to capture the next chunk of audio; and a lightweight, intermittence-aware audio classifier that performs imputation and on-device inference. Through extensive evaluations using popular audio datasets as well as real systems, we demonstrate that SoundSieve yields 5%--30% more accurate inference results than the state-of-the-art.

Resource-optimized deep neural networks (DNNs) nowadays run on microcontrollers to perform a wide variety of audio, image and sensor data classification tasks. Despite comprehensive support for deep learning tools for 32-bit microcontrollers, performing deep learning inferences on 16-bit microcontrollers still remains a chal-lenge. Although there are some tools for implementing neural net-works on 16-bit systems, generally, there is a large gap in efficiency between the development tools for 16-bit microcontrollers and 32-bit (or higher) systems. There is also a steep learning curve that discourages beginners inexperienced with microcontrollers and programming in C to develop efficient and effective deep learning models for 16-bit microcontrollers. To fill this gap, we have created a neural network model generator that (1) automatically transfers parameters of a pre-trained DNN or CNN model from commonly used frameworks to a 16-bit microcontroller, and (2) automatically implements the model on the microcontroller to perform on-device inference. The optimization of data transfer saves time and mini-mizes chances of error, and the automatic implementation reduces the complexity to implement DNNs and CNNs on ultra-low-power microcontrollers.

We propose SmartON, a batteryless system that learns to wake up proactively at the right moment in order to detect events of interest. It does so by adapting the duty cycle to match the distribution of event arrival times under the constraints of harvested energy. While existing energy harvesting systems either wake up periodically at a fixed rate to sense and process the data, or wake up only in accordance with the availability of the energy source, SmartON employs a three-phase learning framework to learn the energy harvesting pattern as well as the pattern of events at run-time, and uses that knowledge to wake itself up when events are most likely to occur. The three-phase learning framework enables rapid adaptation to environmental changes in both short and long terms. Being able to remain asleep more often than a CTID (charging-then-immediate-discharging) wake-up system and adapt to the event pattern, SmartON is able to reduce energy waste, increase energy efficiency, and capture more events. To realize Smar-tON we have developed a dedicated hardware platform whose power management module activates capacitors on-the-fly to dynamically increase its storage capacitance. We conduct both simulation-driven and real-system experiments to demonstrate that SmartON captures 1X–7X more events and is 8X–17X more energy-efficient than a CTID system. All source code and hardware design files are open-sourced1.

In this paper, we argue that the fusion of machine learning (ML) and batteryless computing systems enables true lifelong learning in mobile devices. The lack of learning from experience in current batteryless systems makes them ignorant of changes in their operating environment. Due to high communication cost, latency, privacy, and dependency issues of offloading computation to an edge device, on-device training is a solution for batteryless systems to learn and adapt in dynamically changing environments. Combining batteryless systems and ML is however a challenging task. Sporadic energy supply and limited resources in a batteryless system cause execution-discontinuity and data-constraints in ML processes. To understand these challenges, we identify suitable ML tasks for such systems and study the energy producers, i.e., harvesters, and consumers, i.e., intermittently executable tasks in a ML pipeline. Using a trace-driven simulation, we demonstrate the feasibility of on-device training of a batteryless learner.

We propose a novel text steganography method using RNN Encoder-Decoder structure to generate quatrains, one genre of Chinese poetry. Compared to other text-generation based steganography methods which have either very low embedding rate or flaws in the naturalness of generated texts, our method has higher embedding rate and better text quality. In this paper, we use the LSTM Encoder-Decoder model to generate the first line of a quatrain with a keyword and then generate the following lines one by one. RNN has proved effective in generating poetry, but when applied to steganograpy, poetry quality decreases sharply, because of the redundancy we create to hide information. To overcome this problem, we propose a template-constrained generation method and develop a word-choosing approach using inner-word mutual information. Through a series of experiments, it is proven that our approach outperforms other poetry steganography methods in both embedding rate and poetry quality.