DDPQN: An Efficient DNN Offloading Strategy in Local-Edge-Cloud Collaborative Environments

With the rapid development of the Internet of Things (IoT) and communication technology, deep neural network (DNN) applications, such as medical imaging, speech transcription, handwritten text recognition, have been widely used in IoT devices. However, due to resource constraints on these devices, e.g., limited memory capacity, weak computing capacity, and low battery capacity, IoT devices cannot support complicated DNN operation effectively and, thus, fail to fulfill the requirements of Quality of Service of mobile users. One promising approach is to migrate the DNN model to a remote cloud server to reduce the computing burden on IoT devices. Unfortunately, it still suffers from high delay and low bandwidth when communicating with cloud servers. Although the transmission delay of the edge server is low, its computing capacity lacks scalability and elasticity.To make matters worse, in the real world, the wireless connection between IoT devices and the cloud is intermittent, which can cause offloading failures during large-scale DNN data transmission. In this article, we describe a DNN model migration framework to overcome the abovementioned challenges, which consists of three parts: DNN model preprocessing, partition-offloading plan, and partition-uploading plan. Accordingly, we introduce the operation of the DNN migration and the available methods for each part. In addition, we improve the DNN partition-uploading plan in a multiuser edge-cloud collaborative computing environment. Finally, we highlight the important challenges of achieving more efficient DNN migration and point out the unresolved issues of DNN migration, which may shed light on future research directions.

In the Internet of Things (IoT) paradigm, data is gathered for treatment from locals, machines, vehicles, etc. Cloud computing is providing suitable hardware and software for data processing. Thus, the integration of IoT with cloud capabilities offers several benefits for many applications. However, challenges persist for some use-cases like delay-sensitive services due to the huge amount of information collected by IoT devices and to be processed by cloud servers. Fog computing is expected to overcome several limits and challenges in cloud computing concerning the Quality of Service (QoS) requirements like latency, bandwidth and location awareness. Nevertheless, researchers still have to deal with several issues namely the architectural level and the energetic aspect. In this paper, we investigate fog system architectures and energy consumption in literature, while considering QoS requirements in the synthesis. A system model is then introduced with a potential solution for QoS management for fog computing environment.

It is predicted that the number of Internet of Things (IoT) devices will be more than 75 billion by 2025, where a large portion of IoT devices will be long-range (LoRa) powered by batteries. The battery lifetime limitations and the spectrum shortage have been the main problems in realizing LoRa wide area network (LoRAWAN) for the devices in hard-toreach areas. The dynamic spectrum access technique has gained tremendous research interest as a promising paradigm due to its outstanding performance in improving spectrum efficiency. How to realize intelligent resource allocation (RA) to avoid collisions among IoT devices with low energy consumption is an important problem in LoRaWAN. However, either synchronization and prior information estimation, such as channel state information (CSI), are required, or the energy consumption of LoRa devices is not considered in related work, which may decrease the energy efficiency of the LoRa devices. In addition, the necessary prior information may be challenging to obtain in future networks. To address these issues, we propose a deep Q learning-based RA (DQLRA) method for LoRaWAN. In our proposed method, the gateway (GW) trains the deep neural network (DNN) only based on the transmission state, i.e., transmission failure or success, and the corresponding device number of each LoRa device. Then, each LoRa device can make decisions based on its device number and ACK or NACK information using the trained DNN. Synchronization and prior information estimation are not required in our proposed method, which may improve the energy efficiency of IoT devices. Simulation results show that the proposed method can achieve the optimal frame success rate (FSR) in most scenarios.

Online dynamic stability analysis plays a key role in modern power systems and is becoming even more vital with the penetration of Distributed Energy Resources (DERs). Identification of an accurate lower-order model for the ‘external system” makes it possible to carry out the stability analysis of a study system faster than the analysis for the original detailed system. However, the identified lower-order model becomes invalid when the operating state changes in the power system. ‘Black box’ representation of the ‘external system’ with the application of Deep Neural Networks (DNN) is widely discussed in the literature. In such equivalent models, the DNN response changes based on two aspects i.e. the operating point of the power system and the location/type of the fault. In this paper, a clustering algorithm is proposed to identify the clusters of the ‘external system’ models for different states of the power system, and a DNN is trained for each cluster. The clustering algorithm is based on the pre-fault steady-state power flow from the ‘external system’ to the ‘study system’, which is a reflection of changes in the system state. Mean Absolute Error (MAE) between responses of the actual fault and the predicted fault is used to evaluate the accuracy of the identified system. The proposed technique is easy to implement and requires a lesser number of DNN models compared to identifying an individual DNN model for each operating point to represent the ‘external system’.

The Observing Resources extension of the Constrained Application Protocol (CoAP) is expected to play a major role in future Internet of Things (IoT) applications, since it enables asynchronous monitoring of IoT resources in an efficient manner. As the number of IoT applications will grow, solutions are needed to manage multiple observe relationships for the same resource, especially if hosted by a constrained IoT device. The problem is further exacerbated since multiple observers may have heterogeneous Quality of Service (QoS) requirements on the minimum and maximum frequency with which to receive notifications, while the resource-constrained server may only be able to provide notifications with one fixed period for all. In this work, we propose an efficient solution to support multiple CoAP observers with QoS requirements. In particular, we formulate the problem of finding the notification period which minimizes the amount of messages sent by the server, so as to reduce energy consumption of IoT devices and congestion in constrained networks, while ensuring all observers' requirements are met. We then propose an algorithm, named Observation Period Selection (OPS), which is proved to find the optimal solution to this problem. OPS is also evaluated experimentally to analyze its effectiveness in some simple scenarios.

Deep learning is being incorporated in many modern software systems. Deep learning approaches train a deep neural network (DNN) model using training examples, and then use the DNN model for prediction. While the structure of a DNN model as layers is observable, the model is treated in its entirety as a monolithic component. To change the logic implemented by the model, e.g. to add/remove logic that recognizes inputs belonging to a certain class, or to replace the logic with an alternative, the training examples need to be changed and the DNN needs to be retrained using the new set of examples. We argue that decomposing a DNN into DNN modules— akin to decomposing a monolithic software code into modules—can bring the benefits of modularity to deep learning. In this work, we develop a methodology for decomposing DNNs for multi-class problems into DNN modules. For four canonical problems, namely MNIST, EMNIST, FMNIST, and KMNIST, we demonstrate that such decomposition enables reuse of DNN modules to create different DNNs, enables replacement of one DNN module in a DNN with another without needing to retrain. The DNN models formed by composing DNN modules are at least as good as traditional monolithic DNNs in terms of test accuracy for our problems.

In the wireless Internet of Things (IoT) networks with resource-constrained devices, fog computing has been introduced to deal with the computation-intensive applications at the edges of the networks. While fog computing decreases the computation delay and fronthaul traffic data, it also brings the severe challenge on complex resource allocation of the available computation and communication resources under the stringent quality of service (QoS) requirements. In this paper, we investigate the problem of tasks scheduling and heterogeneous resource allocation for multiple devices in the wireless IoT networks. The IoT devices that collect a massive amount of data need to make proper offloading decision to transfer the data to the fog computing nodes (FNs). Moreover, to support a massive number of device connections and transfer a huge amount of data with low latency and limited resource, we consider the deployment of non-orthogonal multiple access (NOMA) in IoT networks, which enables multiple IoT devices to simultaneously transmit data to the same FN at the same time, frequency, and code domain. We jointly optimize the allocation of resource blocks and transmit power of multiple IoT devices, subject to the respective QoS requirements. Furthermore, the optimization problem is formulated as a mixed-integer nonlinear programming problem to minimize the system energy consumption. Since it is an NP-hard problem, we introduce an improved genetic algorithm (IGA) to solve it. The simulation results show that the proposed scheme achieves good performance in throughput, delay, outage probability, and energy consumption.

The Internet of Things (IoT) devices is able to communicate with another IoT devices in heterogeneous environments. They are also able to create a network by their own using wireless capabilities without a centralized approach. This network is called wireless mobile ad hoc network (MANET). The MANET facilitate IoT devices to communicate with each other in an infrastructure less network of the Internet of Things environment. The IoT devices can dynamically join, create connection and exchange information among them. The role of cloud in this framework is to store and process the data by IoT devices. The communication Security has been raised as one of the most critical issues of cloud computing where resolving such an issue would result in a constant growth in the use and popularity of cloud. Our purpose of this study is to create a framework for providing the communication in an integrated framework of Cloud and MANET among IoT devices. Our main contribution links a new study that integrates cloud computing to the MANET and provides communication to the internet of things. This study can be used in the framework of the Internet of Things in future.

The Internet of Things (IoT) has had a considerable influence on our daily lives by enabling enhanced connection of devices, systems, and services that extends beyond machine-to-machine interactions and encompasses a wide range of protocols, domains, and applications. However, despite privacy concerns shown by IoT users, little has been done to reduce and protect individual information exposure. It is extremely difficult to mitigate IoT devices from reidentification threats which is why it is still a major challenge for IoT users to securely protect their information. The trust controls how we regulate privacy in our IoT platforms in the same way that it governs personal relationships. As IoT devices become increasingly linked, more data is shared across individuals, businesses, governments, and ecosystems. Technologies, sensors, machines, data, and cloud connections all rely largely on trust relationships that have been formed. With the rapid growth of additional types of IoT devices that are being introduced, it, therefore, expands privacy concerns and is difficult to develop trust with an IoT system or device without the option to regulate information privacy settings. Privacy has always been a barrier for many devices as they race for the early adoption of IoT technologies. Several Internet of Things devices or systems will continue to pose privacy threats. As a result, the main objective of this study was to examine the individual understanding of privacy and to promote information privacy protection awareness not only to IoT users but also to organizations that use IoT devices or platforms to run their day-to-day business operations. Furthermore, the objective extends to compare user knowledge and concerns about IoT privacy, as well as to identify any common attitudes and variances. However, in terms of enhancing individuals’ knowledge, an artifact was developed to educate and enhance information privacy awareness among IoT users. A pre- and postquestionnaire was generated to test and validate user knowledge regarding information privacy protection in IoT. The study was conducted using a quantitative research method. Findings indicate that IoT users’ awareness of information privacy protection turned out to be average, suggesting a need for education and awareness. Several participants stated that information privacy protection awareness is required within the community to educate, raise awareness, eliminate human error, and enable individuals to be conscious of their privacy when surfing the Internet.

As applications using deep neural networks (DNNs) are increasingly deployed on mobile devices, researchers are exploring various methods to achieve low energy consumption and high performance. Recently, advances in STT-MRAM have shown promise in offering non-volatility, high performance, and low energy consumption when used to replace DRAM in main memory. Most of the memory space used in DNN applications is occupied by weight and activation data. Typically, the contents of weight data remain unchanged during inference, whereas activation data are modified to store intermediate results across the layers of the DNN. As large DNN applications consume significant energy and require high performance, STT-MRAM is an ideal candidate to fully or partially replace DRAM in main memory. However, the long write latency of STT-MRAM compared to DRAM presents a performance bottleneck. In this work, we propose a reduced retention-time MRAM-based main memory to address this issue. We divide the MRAM into multiple partitions, with each partition implemented with different retention times, tailored for DNN applications. In this approach, the DNN weights can be assigned to the long retention-time partition, while the activation data can be allocated to the short retention-time partition, optimizing for the varying characteristics of data reuse. To achieve high performance, we propose two mapping schemes: intra-segment and inter-segment circular buffers, which dynamically map DNN activation data (i.e., virtual pages of streaming data) to physical pages in a circular fashion to exploit reuse patterns of DNN data. These circular buffers are mapped to the short retention-time MRAM partition as much as possible. The intra-circular buffer mapping scheme achieves an average improvement of 14.4% in bandwidth and 12.6% in inference latency compared to DRAM for on-device DNN applications. Furthermore, the inter-circular buffer mapping scheme offers 11.1% bandwidth and 11.2% latency improvements on average using only 16 MBytes of the short retention-time partition.

Deep neural network (DNN) models typically have many hyperparameters that can be configured to achieve optimal performance on a particular dataset. Practitioners usually tune the hyperparameters of their DNN models by training a number of trial models with different configurations of the hyperparameters, to find the optimal hyperparameter configuration that maximizes the training accuracy or minimizes the training loss. As such hyperparameter tuning usually focuses on the model accuracy or the loss function, it is not clear and remains under-explored how the process impacts other performance properties of DNN models, such as inference latency and model size. On the other hand, standard DNN models are often large in size and computing-intensive, prohibiting them from being directly deployed in resource-bounded environments such as mobile devices and Internet of Things (IoT) devices. To tackle this problem, various model optimization techniques (e.g., pruning or quantization) are proposed to make DNN models smaller and less computing-intensive so that they are better suited for resource-bounded environments. However, it is neither clear how the model optimization techniques impact other performance properties of DNN models such as inference latency and battery consumption, nor how the model optimization techniques impact the effect of hyperparameter tuning (i.e., the compounding effect). Therefore, in this paper, we perform a comprehensive study on four representative and widely-adopted DNN models, i.e., CNN image classification , Resnet-50 , CNN text classification , and LSTM sentiment classification , to investigate how different DNN model hyperparameters affect the standard DNN models, as well as how the hyperparameter tuning combined with model optimization affect the optimized DNN models, in terms of various performance properties (e.g., inference latency or battery consumption). Our empirical results indicate that tuning specific hyperparameters has heterogeneous impact on the performance of DNN models across different models and different performance properties. In particular, although the top tuned DNN models usually have very similar accuracy, they may have significantly different performance in terms of other aspects (e.g., inference latency). We also observe that model optimization has a confounding effect on the impact of hyperparameters on DNN model performance. For example, two sets of hyperparameters may result in standard models with similar performance but their performance may become significantly different after they are optimized and deployed on the mobile device. Our findings highlight that practitioners can benefit from paying attention to a variety of performance properties and the confounding effect of model optimization when tuning and optimizing their DNN models.

Internet-of-Things (IoT) devices are becoming both intelligent and green. On the one hand, Deep Neural Network (DNN) compression techniques make it possible to run deep learning applications on devices equipped with low-end microcontrollers (MCUs). By performing deep learning on IoT devices, in-situ decision-making can be made, which can improve the responsiveness of such devices to the environment and reduce data uploading to edge servers or clouds to save valuable network bandwidth. On the other hand, many IoT devices in the future will be powered by energy harvesters instead of batteries to reduce environmental pollution and achieve permanent service free of battery maintenance. As the energy output of energy harvesters is tiny and unstable, energy harvesting IoT (EH-IoT) devices will experience frequent power failures during their execution, making the software task hard to progress. The deep learning tasks running on such devices must face this challenge and, at the same time, ensure satisfactory execution efficiency. We believe deploying deep learning on EH-IoT devices that execute intermittently will be a challenging yet promising research direction. To motivate research in this direction, this paper summarizes existing solutions and provides an in-depth discussion of future challenges that deserve further investigation. With IoT devices becoming more intelligent and green, DNN inference on EH-IoT devices will generate a much more significant impact in the future in academia and industry.

Seeing is not always believing: Insights on IoT manufacturing from firmware composition analysis and vendor survey

With the advancement in science and technology, the demand for electronic Internet of Things (IoT) devices has increased to store and process digital data. With the advancement in IoT technology, we are more reliant on IoT devices to perform our regular tasks. IoT Forensics deals with the investigation of IoT devices such as smartphones, tablets, smart watches, or any device which is connected to the Internet. The data on these devices can still be retrieved via IoT forensics even if they are destroyed, resold, or misplaced. Before discarding any IoT device, we must consider how to secure the data stored on them. We are then going to discuss the several types of personal data that could be retrieved from discarded IoT devices. Finally, we are going to discuss the protection measures for the safe disposal of IoT devices.

Bayesian optimization has been used for the global optimization of communication parameters of Vehicular Ad Hoc Networks (VANETs) for safety applications with stringent quality of service (QoS) requirements. However, the effectiveness of the methodology relies on an accurate analytic model for querying a distribution over functions, which is not practical. Furthermore, incorporating QoS requirements as constraints into the search process is cumbersome, timeconsuming, and even unreliable. In this paper, we present a new approach to the constrained Bayesian optimization of IEEE 802.11 based VANETs for safety messaging with the help of deep learning neural networks (DLNNs). First, we design and train a DLNN using data collected from the analytic models or channel measurements to approximate a mapping from the search parameter space to the QoS metrics. The QoS constraints are naturally incorporated into the DLNN through preprocessing data pairs that cannot meet the QoS requirements. Then, the Bayesian optimization is conducted to find the optimal communication parameters for the best channel usage on the condition that all QoS requirements are met. Accordingly, experiments are carried out on the Google Colab platform where the impact of DLNN structure, data sampling rate, and other optimization parameters are investigated. In comparison to other optimization approaches, utilizing a DLNN in the Bayesian optimization process is more time efficient and flexible.