The use of Dynamic Random Access Memory (DRAM) for storing Machine Learning (ML) models plays a critical role in accelerating ML inference tasks in the next generation of communication systems. However, periodic refreshment of DRAM results in wasteful energy consumption during standby periods, which is significant for resource-constrained Internet of Things (IoT) devices. To solve this problem, this work advocates two novel approaches: 1) wireless memory activation and 2) wireless memory approximation. These enable the wireless devices to efficiently manage the available memory by considering the timing aspects and relevance of ML model usage; hence, reducing the overall energy consumption. Numerical results show that our proposed scheme can realize smaller energy consumption than the always-on approach while satisfying the retrieval accuracy constraint.

This paper investigates how to achieve both low-power operations of sensor nodes and accurate state estimation using Kalman filter for internet of things (IoT) monitoring employing wireless sensor networks under radio resource constraint. We consider two policies used by the base station to collect observations from the sensor nodes: (i) an oblivious policy, based on statistics of the observations, and (ii) a decentralized policy, based on autonomous decision of each sensor based on its instantaneous observation. This work introduces a wake-up receiver and wake-up signaling to both policies to improve the energy efficiency of the sensor nodes. The decentralized policy designed with random access prioritizes transmissions of instantaneous observations that are highly likely to contribute to the improvement of state estimation. Our numerical results show that the decentralized policy improves the accuracy of the estimation in comparison to the oblivious policy under the constraint on the radio resource and consumed energy when the correlation between the processes observed by the sensor nodes is low. We also clarify the degree of correlation in which the superiority of two policies changes.

Energy efficiency and information freshness are key requirements for sensor nodes serving Industrial Internet of Things (IIoT) applications, where a sink node collects informative and fresh data before a deadline, e.g., to control an external actuator. Content-based wake-up (CoWu) activates a subset of nodes that hold data relevant for the sink’s goal, thereby offering an energy-efficient way to attain objectives related to information freshness. This paper focuses on a scenario where the sink collects fresh information on top-kvalues, defined as data from the nodes observing thekhighest readings at the deadline. We introduce a new metric called top-kQuery Age of Information (k-QAoI), which allows us to characterize the performance of CoWu by considering the characteristics of the physical process. Further, we show how to select the CoWu parameters, such as its timing and threshold, to attain both information freshness and energy efficiency. The numerical results reveal the effectiveness of the CoWu approach, which is able to collect top-kdata with higher energy efficiency while reducingk-QAoI when compared to round-robin scheduling, especially when the number of nodes is large and the required size ofkis small.

This paper considers energy-efficient connectivity for Internet of Things (IoT) devices in a coexistence scenario between two distinctive communication models: pull- and push-based communication models. In pull-based communication, the base station (BS) decides when to retrieve a specific type of data from the IoT devices equipped with wake-up receivers, while in push-based communication, the IoT device decides when and which data to transmit. To efficiently manage both types of traffic, this paper applies content-based wake-up (CoWu) and designs a medium access control (MAC) frame. This enables the BS to activate a subset of pull-based nodes and collect the relevant data to fulfill its tasks, while receiving data from the push-based communication nodes. This paper analyzes the basic trade-off through the MAC layer operations: allocating longer duration for collecting data from pull-based nodes can lead to high retrieval accuracy while decreasing the probability of data transmission success for push-based nodes, and vice versa. Numerical results show that CoWu can manage communication requirements for both pull-based and push-based nodes while realizing high energy efficiency (up to 38%) of IoT devices, compared to the baseline.

This paper introduces EcoPull, a sustainable Internet of Things (IoT) framework powered by Tiny Machine Learning (TinyML) models for efficient image retrieval from multiple devices. The devices are equipped with two types of TinyML models: i) a behavior model and ii) an image compressor model. The behavior model filters out irrelevant images based on the current task, minimizing unnecessary data transmission and reducing communication resource competition among devices. The image compressor model enables devices to communicate with the edge server (ES) using latent representations of images, thereby reducing communication bandwidth usage. While integrating TinyML models into IoT devices does increase energy consumption due to the inference process, this cost is carefully accounted for in our design. Numerical results show that the proposed framework can achieve over 77% and 43% energy savings compared to the simple offloading and a state-of-the-art baseline while still maintaining the quality of the retrieved images at the ES.

This paper focuses on top–k query in cluster–based wireless sensor networks (WSNs) employing wake–up receivers, which aims to grasp the information on top–k data efficiently in terms of energy consumed by sensor nodes. As a wake–up control for top–k query, countdown content–based wake–up (CDCoWu) has been proposed. CDCoWU selectively activates sensor nodes storing data belonging to top–k dataset, and its superiority to a conventional identity–based wake–up (IDWu) has been confirmed for a single–hop network accommodating a large number of nodes. In this paper, we extend these wake–up control to a cluster–based WSN, a common network structure of WSN. The cluster–based WSN contains nodes called cluster heads that relay wake–up signals and data for their cluster members. The proposed cluster–based wake–up control conducts activations and data collections of different clusters sequentially while the results of data collections at a cluster are exploited for reducing unnecessary data transmissions at the other clusters. We also introduce a hybrid mechanism of wake–up control, where a wake–up control employed at each cluster is selected between CDCoWu and IDWu according to its number of cluster members. Our simulation results show that the proposed hybrid wake–up control achieves smaller energy consumption and data collection delay than the control solely employing CDCoWu or IDWu.

This paper aims to improve the efficiency of wake-up control for top-$k$query in wireless sensor networks (WSNs) by introducing a function to estimate the data distribution of sensors. In order to minimize the wasteful wake-up of nodes outside the top-$k$set, the sink gradually enlarges the search region of data from the highest value by employing countdown content-based wake-up (CDCoWu). In this case, the step size to enlarge the search region plays an important role: a larger step causes many nodes to simultaneously wake up, which leads to severe congestion, while a smaller step leads to the failed wake-up trial with no replies. In this paper, in order for the sink to appropriately select the step size, we introduce kernel density estimation into wake-up control, by which the sink first estimates the probability density function (PDF) of data owned by sensor nodes. The efficiency of CDCoWu can be improved with more accurate estimation on PDF, which, however, requires larger estimation cost. In this paper, we analyze the trade off between the accuracy of estimation and efficiency of CDCoWu, and investigate whether the overall efficiency in terms of energy consumption and delay can be improved by the proposed wake-up control with density estimation. Our numerical results show that the proposed wake-up control achieves better energy effi-ciency and data collection delay than the conventional CDCoWu employing a fixed step size.

This paper considers the periodical top-k query for wireless sensors, where a sink periodically seeks top-k values and corresponding nodes-IDs in a sensing field. We advocate applying wake-up receivers in this scenario in order to reduce the wasteful energy consumption of sensor nodes. A wake-up control is proposed for the sink to collect top-k data with temporal correlation with small amount of energy. Specifically, the proposed scheme employs different types of wake-up control, which aims to wake up only the subset of nodes whose owning data contribute to the identification of current top-k set by exploiting temporal correlation. The numerical results confirm the effectiveness of our proposed scheme compared with the conventional wake-up control for top-k query in terms of average energy consumption, especially when the degree of correlation is high.

This paper proposes content-based control of wakeup receivers for data collection in wireless sensor networks. The wake-up procedure is designed with a goal of waking up only the subset of the sensor nodes which have the relevant data observations. This prevents the sensors with less relevant data from waking up and wasting energy, which is inevitable when employing conventional ID-based wake-up control. We apply the proposed content-based wake-up scheme to top- k query, where the sink attempts to collect information on the set of nodes that own top- k observations from the sensing field. Assuming medium access based on p-persistent CSMA, we design a content-based wake-up control scheme suited for the data collection of top- k query. We analyze the scheme theoretically in terms of data collection delay and energy-efficiency and compare it to the ID-based wake-up. The numerical results confirm the effectiveness of the proposed content-based wake-up control, especially when the number of sensor nodes is large.