Search

Health monitoring of civil infrastructures is a key application of Internet of things (IoT), while edge computing is an important component of IoT. In this context, swarms of autonomous in...

We address energy-efficient placement of data and analytics components of composite analytics services on a wireless network to minimize execution-time energy consumption (computation and communication) subject to compute, storage and network resource constraints.We introduce an expressive analytics service hypergraph model for representing k-ary composability relationships (k ≥ 2) between various analytics and data components and leverage binary quadratic programming (BQP) to minimize the total energy consumption of a given placement of the analytics hypergraph nodes on the network subject to resource availability constraints. Then, after defining a potential energy functional Φ(•) to model the affinities of analytics components and network resources using analogs of attractive and repulsive forces in physics, we propose a decentralized Metropolis Monte Carlo (MMC) sampling method which seeks to minimize Φ by moving analytics and data on the network. Although Φ is non-convex, using a potential game formulation, we identify conditions under which the algorithm provably converges to a local minimum energy equilibrium placement configuration.Trace-based simulations of the placement of a deep-neural-network analytics service on a realistic wireless network show that for smaller problem instances our MMC algorithm yields placements with total energy within a small factor of BQP and more balanced workload distributions; for larger problems, it yields low-energy configurations while the BQP approach fails.

Capturing and managing the data needed to build effective machine learning models for custom IoT environments requires a great deal of effort. The amount of data generated from IoT devices is abundant, but tools to find datasets appropriate for the desired models are lacking. This paper presents a data capture system and data management catalog with solutions addressing the challenges of curating IoT data applied to purpose-built machine learning deployments.

The reporting on nocturnal sounds like cough and snore is not only relevant to follow the progress of respiratory diseases of patients but also to assess the quality of sleep of subjects. In this study, we discuss an audio analysis approach to count individual cough events and the duration of snore sounds in presence of air-conditioner noise through recordings of a smartphone and computationally efficient classifiers. A new audio data set of cough and snore sounds was acquired from 26 subjects. Energy threshold-based segmentation was applied to identify cough or snore events in the original low noise dataset. A k-nearest neighbor classifier was trained to merge cough phases belonging to the same cough event, to derive the proper ground-truth labeling. The original audio signal was augmented by the superposition of air-conditioner noise, with a signal-to-noise ratio of -40dB to 40dB, to enrich the training set of the binary classifier. Nine out of 40 mel-frequency cepstral coefficients in combination with the logarithm of energy from an entire cough or snore event were computed. Various classifiers, such as k-nearest neighbor (k-NN), rule-based classifier, decision tree, random forest, naive Bayes, and support vector machine were benchmarked against each other. The k-NN classifier with k=1 resulted in the highest F_ 1 scores of.85 and.88 in the binary classification task using generalized and personalized models, respectively, considering noise augmented samples. These results underline the potential of smartphones to objectively report on patient symptoms through audio recordings at night.

Supervised machine learning (ML) algorithms are aimed at maximizing classification performance under available energy and storage constraints. They try to map the training data to the corresponding labels while ensuring generalizability to unseen data. However, they do not integrate meaning-based relationships among labels in the decision process. On the other hand, natural language processing (NLP) algorithms emphasize the importance of semantic information. In this paper, we synthesize the complementary advantages of supervised ML and NLP algorithms into one method that we refer to as SECRET (Semantically Enhanced Classification of REal-world Tasks). SECRET performs classifications by fusing the semantic information of the labels with the available data: it combines the feature space of the supervised algorithms with the semantic space of the NLP algorithms and predicts labels based on this joint space. Experimental results indicate that, compared to traditional supervised learning, SECRET achieves up to 14.0% accuracy and 13.1% F1 score improvements. Moreover, compared to ensemble methods, SECRET achieves up to 12.7% accuracy and 13.3% F1 score improvements. This points to a new research direction for supervised classification based on incorporation of semantic information., Comment: 16 pages, 20 figures, 2 tables - IEEE Transactions on Computers

Deep neural networks (DNNs) are emerging as important drivers for GPU (Graphical Processing Unit) usage. Routinely, now, cloud offerings include GPU-capable VMs, and GPUs are used for training and testing DNNs. A popular way to run inference (or testing) tasks with DNNs is to use middleware called a serving system. Tensorflow-Serving (TF-Serving) is an example of a DNN serving system. In this paper, we consider the problem of carefully scheduling multiple concurrent DNNs in a serving system on a single GPU to achieve fairness or service differentiation objectives, a capability crucial to cloud-based TF-Serving offerings. In scheduling DNNs, we face two challenges: how to schedule, and switch between, different DNN jobs at low overhead; and, how to account for their usage. Our system, Olympian, extends TF-Serving to enable fair sharing of a GPU across multiple concurrent large DNNs at low overhead, a capability TF-Serving by itself is not able to achieve. Specifically, Olympian can run concurrent instances of several large DNN models such as Inception, ResNet, GoogLeNet, AlexNet and VGG, provide each with an equal share of the GPU, while interleaving them at timescales of 1-2 ms, and incurring an overhead of less than 2%. It achieves this by leveraging the predictability of GPU computations to profile GPU resource usage models offline, then using these to achieve low overhead switching between DNNs.

With the rise of small, networked sensors, the volume of data generated increasingly require curation by AI to analyze which events are of sufficient importance to report to human operators. We consider the ultimate limit of edge computing, when it is impractical to employ external resources for the curation, but individual devices have insufficient computing resources to perform the analytics themselves. In a previous paper we introduced a decenralized method that distributes the analytics over the network of devices, employing simulated annealing, based on physics-inspired Metropolis Monte Carlo. If the present paper we discuss the capability of this method to balance the energy consumption of the placement on a network of heterogeneous resources. We introduce the balanced utilization index (BUI), an adaptation of Jain’s Fairness Index, to measure this balance.

A wide array of military and commercial applications rely on the collection and processing of audio data. One approach to perform analytics and machine learning on such data is to upload and process them at a central server (e.g., cloud) which offers abundant processing resources and the ability to run sophisticated machine learning models and analytics on the audio data. This approach can be inefficient due to the low bandwidth and energy limitations of mobile devices as well as intermittent connectivity to a central collection point such as the cloud. It is also problematic as audio data are often highly sensitive and subject to privacy constraints. An alternative approach is to perform audio analytics at edge of the network where data is generated. The challenge in this approach is the requirement to perform analytics subject to resource constraints which limit performance and accuracy of predictive analytics. In this paper, we present a system for performing predictive analytics on audio data, where the training is executed on the cloud and the classification can be executed at the edge. We present the design principles and architecture of the system, and quantify the performance tradeoff of executing analytics at contemporary edge devices versus the cloud.

In the current Named Data Networking implementation, forwarding a data request requires finding an exact match between the prefix of the name carried in the request and a forwarding table entry. However, consumers may not always know the exact naming, or an exact prefix, of their desired data. The current approach to this problem---establishing naming conventions and performing name lookup---can be infeasible in highly ad hoc, heterogeneous, and dynamic environments: the same data can be named using different terms or even languages, naming conventions may be minimal if they exist at all, and name lookups can be costly. In this paper, we present a fuzzy Interest forwarding approach that exploits semantic similarities between the names carried in Interest packets and the names of potentially matching data in CS and entries in FIB. We describe the fuzzy Interest forwarding approach and outline the semantic understanding function that determines the name matching.

The fundamental aim of this paper is to position the opportunities and challenges for adopting Named Data Networking (NDN) in the specific context of military coalition operations and tactical networks. The characteristic properties of tactical networks include high dynamics in multiple dimensions: bandwidth, network congestion, frequent topological changes, geographical mobility of assets, as well as dynamic changes in information access policies. Furthermore, coalition networks must provide secure and efficient communication across coalition boundaries and mitigate the impact of adversarial entities attempting to obstruct the mission. In this paper, we elaborate on the basic NDN architecture characteristics, including robust data discovery and retrieval over ad hoc and intermittent connectivity, inherent security, efficient content distribution, and automatic in-network caching; we also articulate how the above properties can all be utilized to enable resilient and secure data collection, improve the analytics capacity of the network, and to speed up and improve the quality of distributed decision making in challenging coalition environments.

We study the problem of flexibly, dynamically, and adaptively moving, positioning, and instantiating computing tasks and data in federated, distributed edge systems. We call this process “agile code and agile data” (ACAD). We explore the adaptation of physics-inspired models, used for atomistic simulations, to the ACAD problem, treating the code and data as particles on a graph, interacting through different potential energy models. We discuss the mapping between the different elements of ACAD problem and our particles-on-a-graph model, considering different frameworks for data analytics. We explore gravitational, elastic and Coulombic models, both with global and local energy minimization, finding that the Coulombic model obtains the most efficient solution.

One of the challenges in a military wireless sensor network is the determination of an information collection infrastructure which minimizes battery power consumption. The problem of determining the right information collection infrastructure can be viewed as a variation of the network design problem, with the additional constraints related to battery power minimization and redundancy. The problem in its generality is NP-hard and various heuristics have been developed over time to address various issues associated with it. In this paper, we propose a heuristic based on the mammalian circulatory system, which results in a better solution to the design problem than the state of the art alternatives.

The dynamic environment of hybrid mobile and fixed wireless networks used in military operations poses significant challenges in the efficient provisioning of software functionality to application clients. With their transient topology, the software services hosted on mobile nodes may become temporarily unavailable or the cost of transferring data across the network may become too high. To address this problem we have designed a placement technique that allows the dynamic repositioning of services within the network as it evolves. The technique repositions services in reaction to changes in network topology as well as in various system properties, such as service dependencies or the workload generated by application clients. In our approach we use a multi-layer model to represent a service-based software system embedded in a network topology. We apply constraint programming and its linear programming relaxation to solve the model to find where best to place or reposition the services. We evaluate the technique in terms of its effectiveness and cost under various simulated operational conditions.

In machine-to-machine (M2M) communication, smart spectrum management is vital for the system performance since there are a large number of wireless devices sharing the same limited spectrum and the spectrum resource is scarce. For the spectrum sharing, it's not only encountered in ISM band, but also in the networks with dedicated spectrums when the network is under evolution or upgrading, such as the system updating from narrow-band to broadband in SmartGrid communications. In this paper, we summarize the typical methods in wireless cellular network to improve the spectrum utilization and compare their advantages and limitations as applied to M2M communications. Based on the analysis, we propose a novel mechanism for dynamic spectrum allocation in a cognitive radio environment for SmartGrid applications using OFDMA technology and give some evaluations based on field test data. The spectrum allocation we proposed can be divided into two stages: network entry stage based on initial sensing and dynamic interference avoidance stage based on periodical sensing. At the network entry stage, we focus the discussion on the design of optimal spectrum auto planning algorithm and provide a new scheme based on backup list generation and internal/external interference differentiation. After the OFDMA based system enters into normal communication stage, interference avoidance within the allocated spectrum band is the main task of the M2M system. The in-band spectrum will be segmented into several sub-bands and be scheduled by the system based on periodical spectrum sensing results to avoid interferences coming from external devices(SCADA or unknown system) which share the same spectrum with the current system. The spectrum allocation mechanism we proposed has been used in IBM Wireless Internet-of-Thing (IoT) platform and achieved good performance during the field test.

The quality of computing certain aggregation functions based on incomplete measurements for the purpose of distributed network monitoring is considered. Network monitoring plays a fundamental role in network management systems by providing timely information on the network status, which is crucial for administration purposes. To reduce network overhead and for easier assimilation, this information is usually presented by calculating a few key aggregate metrics. The aggregates are periodically computed from a large number of detailed events collected continuously during the course of the network operations. Under errors induced by network delays, the accuracy of typical aggregation functions used in network management systems is evaluated both analytically and by simulations. The results provide a quantifiable trade-off between accuracy and timeliness of the information acquired, which can then be used to design and optimize network management systems.

Discovering the topology of a network and detecting link state changes (e.g.: link failures) is an essential element for various network management and monitoring tasks. In this paper, we investigate scalable mechanisms to monitor the topology and link states of networks based on information available in network nodes' routing tables. We first present an algorithm that infers the network topology based on the full or partial information about network distances between nodes, based on which we obtain a scalable network topology discovery solution via a novel use of random walk in graphs. We then present scalable algorithms to detect the state changes of remote links by monitoring the routing tables of a small fraction of the routers, where the routers to be monitored are selected by a greedy approach to an NP-complete Tree Cover problem. We show the efficacy and scalability of our topology monitoring algorithms through experimental evaluation performed both on synthetic topologies and on a large topology data-set from a real enterprise network.

To increase the utilization of the available frequency channel space in 802.11-based wireless mesh networks, recent work has explored solutions based on multi-radio stations. This paper reports on our design and experimental study of a distributed, self-stabilizing mechanism that assigns channels to multi-radio nodes in wireless mesh networks. We take a modular approach by decoupling the channel selection decision from the data forwarding mechanism, which makes our solution readily applicable to real-world operation when used with emerging multi-radio routing solutions. We demonstrate the efficacy of our protocol on a real-world, 14-node testbed comprised of nodes, each equipped with an 802.11a card and an 802.11g card. We show via extensive measurements on our testbed that our channel assignment algorithm improves the network capacity by 50% in comparison to a homogeneous channel assignment and by 20% in comparison to a random assignment.

Motivated by the need to enable easier data sharing and curb rising storage management costs, storage systems are becoming increasingly consolidated and thereby shared by a large number of users and applications. In such environments, service differentiation becomes increasingly important. Since caching is a fundamental and pervasive technique employed to improve the performance of storage systems, providing differentiated services from a storage cache is a crucial component of the entire end-to-end QoS solution. In this paper we discuss a QoS architecture for a shared storage proxy cache which can provide long-term hit rate assurances to competing classes. The proposed architecture consists of three components: (a) per-class feedback controllers that track the performance of each class, (b) a fairness controller that allocates excess resources fairly in the case when all goals are met, and (c) a contention resolver that decides cache allocation in the case when at least one class does not meet its target hit rate. We compare the performance of various feedback per-class controllers, and provide guidelines for designing QoS mechanisms for such a dynamic environment.

Emerging large scale distributed networking systems, such as P2P file sharing systems, sensor networks, and ad hoc wireless networks, require replication of content, functionality, or configuration to enact or optimize communication tasks. The placement of these replicated resources can significantly impact performance. We present a novel self-stabilizing, fully distributed, asynchronous, scalable protocol that can be used to place replicated resources such that each node is "close" to some copy of any object. We describe our protocol in the context of a graph with colored nodes, where a node's color indicates the replica/task that it is assigned. Our combination of theoretical results and simulation prove stabilization of the protocol, and evaluate its performance in the context of convergence time, message transmissions, and color distance. Our results show that the protocol generates colorings that are close to the optimal under a set of metrics, making such a protocol ideal for emerging networking systems.

Quality of service for high-bandwidth or delay-sensitive applications in the Internet, such as streaming media and online games, can be significantly improved by replicating server content. We present a decentralized algorithm that allocates server resources to replicated servers in large-scale client-server networks to reduce network distance between each client and the nearby replicated server hosting the resources of interest to that client. Preliminary simulation results show that our algorithm converges quickly to an allocation that reduces the expected client-server distance by almost half compared to the distance when the assignment of replicated servers is done at random.

Large scale interactive online games aim to support a very large number of game players simultaneously. To support hundreds of thousands of concurrent players, game providers have so far focused on developing highly scalable game server architectures and extensible network infrastructures. Recently, distributed online games are beginning to incorporate more interactive features and action sequences; thus, it becomes increasingly important to provision server resources in an efficient manner to support real-time interaction between the users. In this paper, we present a novel distributed algorithm to select game servers for a group of clients participating in a large scale interactive online game session. The goal of server selection is to minimize the server resource usage while satisfying a real-time delay constraint. We develop a synchronization delay model for interactive games and formulate the server selection problem, and prove that the considered problem is NP-hard. The proposed algorithm, called zoom-in-zoom-out, is adaptive to session dynamics (e.g., clients join) and lets the clients select appropriate servers in a distributed manner such that the server resource is efficiently utilized. Using simulation, we study the performance of the proposed algorithm and show that it is simple, yet effective in achieving its design goal. In particular, we show that the performance of our algorithm is comparable to, or sometimes even better than, that of centralized greedy algorithms, which require global information and extensive computations.

Connectivity within ad-hoc and peer-to-peer networks undergoes constant change. One approach to reducing the cost of finding information within these networks is to replicate the information among multiple points within the network. A desirable replication approach should cache copies of all pieces of information as close to each node as possible without exceeding the storage resources of the nodes within the network. In addition, the approach should require minimum communication overhead among participating nodes and should adjust the locations of stored content as connectivity within the network changes. Here, we formulate this caching problem as a graph coloring problem, where the color of the node determines the content that the node should store. We present a distributed algorithm where each node chooses its color in a greedy manner, minimizing its own distance to the color furthest from it. We demonstrate convergence of this algorithm and evaluate its performance in the context of its ability to place information near all nodes in the network.© (2002) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.