Tactical operations like search and rescue or surveillance necessitate the rapid synthesis of physically dispersed assets and mobile compute nodes into a network capable of efficient and reliable information gathering, dissemination, and processing. We formalize this network synthesis problem as selecting one among a set of potentially deployable networks which optimally supports the distributed execution of complex applications. We present the NSDC (network synthesis for dispersed computing) framework; a general framework for studying this type of problem and use it to provide a solution for one well-motivated variant. We discuss how the framework can be extended to support other objectives, parameters, and constraints as well as more scalable solution approaches.
The commercial industry has been working to develop ever-larger and more capable machine learning (ML) models (such as recent models from OpenAI, Microsoft and Google with more than ten billion parameters) for everything from general language processing to computer vision that have larger and larger computational, memory and other resource requirements. These powerful models generally need hardware accelerators to accommodate their workloads. GPU systems have been a popular choice among users and deep learning model designers either for their ability to run the inherently parallel deep learning workloads efficiently or for their large memory resource that would fit large deep learning models. We profile and analyze different deep learning workloads using various GPUs and configurations, emphasizing how deep learning architectures have diverse compute requirements. We analyze three popular deep learning workloads: Ultralytics’s You-Only-Look-Once model (Yolov5) on COCO dataset, Bidirectional Encoder Representations from Transformers (BERT), and Deep Learning Recommendation Model (DLRM) on the Criteo Kaggle Display Advertising Challenge Dataset. This work aims to shed light on the performance bottlenecks when using GPU systems as accelerators for training recent deep learning models.
We develop a network synthesis scenario, which is built around a concrete perimeter surveillance application, yet we believe captures a number of the challenges and requirements that are common to other tactical communication and computational network applications. The proposed scenario addresses the problem of binary population identification within a perimeter: our goal is to synthesize a sensing and computing network that classifies people moving within a given perimeter into one of two categories (e.g., friend or foe). We discuss several open challenges that we organize across the following clusters: sensor placement, communication network provisioning and optimization, computational task placement, dynamic re-synthesis and resilience under adversarial settings. We also briefly discuss approaches that attempt to address such challenges.
Energy saving is one of the major concerns for low rate personal area networks. This paper models energy consumption for beacon-enabled time-slotted media accessing control cooperated with sleeping scheduling in a star network formulation for IEEE 802.15.4 standard. We investigate two different upstream (data transfer from devices to a network coordinator) strategies: a) tracking strategy: the devices wake up and check status (track the beacon) in each time slot; b) non-tracking strategy: nodes only wake-up upon data arriving and
stay awake till data transmitted to the coordinator. We consider the tradeoff between energy cost and average data transmission delay for both strategies. Both scenarios are formulated as optimization problems and the optimal solutions are discussed. Our results show that different data arrival rate and system parameters (such
as contention access period interval, upstream speed etc.) result in different strategies in terms of energy optimization with maximum delay constraints. Hence, according to different applications and system settings, different strategies might be chosen by each node to achieve energy optimization for both self-interested view
and system view. We give the relation among the tunable parameters by formulas and plots to illustrate which strategy is better under corresponding parameters. There are two main points emphasized in our results with delay constraints: on one hand, when the system setting is fixed by coordinator, nodes in the network can intelligently change their strategies according to corresponding application data arrival rate; on the other hand, when the nodes' applications are known by the coordinator, the coordinator can tune the system parameters to achieve optimal system energy consumption.
Due to the broadcast and error prone nature of wireless medium, novel routing mechanisms based on receiver contention
have been proposed recently. The intuition of this strategy is, transmitters make routing decisions based on contentions
of nodes that have successful reception. A remarkable advantage of receiver contention is the long average advancement
of transmissions. To the best our knowledge, existing works utilizing receiver contention schemes are all based on a
common assumption. That is, feedback packets sent by contending nodes are all destined to the transmitters. However,
probability of reception is a function of distance. The longer the distance is, the lower the reception probability will be5.
According to this relation, we argue that transmitters may not be the best nodes to taking care of contention packets. In
this paper, we consider uniformly distributed sensor networks, and propose the optimal locations, in terms of
maximizing the expected advancement of each transmission, to place nodes which are responsible for handling feedback
packets. We call these nodes feedback handlers. Based on the simulation results, placing the feedback handlers on the
optimal locations can raise expected advancement up to about 30 percent, comparing to existing works.
KEYWORDS: Sensors, Diffusion, Sensor networks, Data fusion, Head, Process modeling, Data modeling, Filtering (signal processing), Fusion energy, Signal to noise ratio
The monitoring of a diffuse process, such as the propagation of a toxic gas in an area, using the partial differential equation (PED) model via autonomous wireless sensor networks is studied in this research. Sensor nodes update the base station with their estimates of PDE model parameters rather than raw sensor measurements. Then, the base station can reconstruct the phenomenon through model parameters and initial and boundary conditions. In-network processing techniques to estimate the PDE coefficients are presented. A scheme is presented to provide a hybrid combination of decision and data fusion to find a proper tradeoff between estimate accuracy and energy efficiency. Besides, several open issues in this research context, such as identifiability of parameters, monitoring of time varying boundary conditions and unknown sources, are discussed.
KEYWORDS: Sensors, Data communications, Reliability, Failure analysis, Data modeling, Data processing, Sensor networks, Mathematics, Environmental sensing, Analytical research
The impact of the non-uniform individual sensor node lifetime on the
connectivity of a data gathering tree over time is studied in this
research. The lifetime of sensor devices depends on the device failure rate and/or battery energy depletion, and surviving nodes may not preserve the uniform node density across the network as nodes age. We first examine the general node aging problem by considering the energy consumption rate and the node failure rate. The energy consumption rate in a data gathering tree is presented with or without data aggregation. The nodes in each hop level show a different energy depletion rates even with data aggregation, which is studied by mathematical analysis as well as simulation results. Then, the resulting non-uniform connectivity over time in a data gathering tree is examined with a node's survivor function. It is shown by mathematical analysis and simulation results that the node aging process has a significant impact on the connectivity
as the hop distance increases.
Energy-efficient tracking of a target using a sensor network has received significant attention in recent research. Our earlier study on energy-quality tradeoffs in target tracking with binary sensors showed that optimal selective activation of sensor nodes based on prediction of the target's trajectory could achieve orders of magnitude savings in the energy expenditure over naive and random activation, while achieving almost the same tracking quality. In this paper, we consider a more realistic sensor model and extend the analysis of activation strategies to account for the presence of noise in sensor measurements. Our results confirm that the best quality of tracking that can be obtained with selective activation depends on the noise level in sensor measurements and that the optimal radius of activation depends on the noise level and the density of deployment. We also show how duty cycling with selective activation can be used to obtain flexible tradeoffs between the energy expenditure and quality of tracking.
We describe the application of image processing techniques for data refinement in sensor networks, by mapping network nodes to pixels in an image. Due to their localized, distributed nature, these techniques are inherently scalable and therefore desirable for use in large sensor networks. We examine two specific problems: cleaning of uncorrelated sensor noise, and the decentralized detection of edges (such as the perimeter of a chemical leak). Our simulation results show that the performance of these processing techniques depends critically upon both sensor density and radio range.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.