A two-level architecture for deep learning applications in mobile edge computing

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.

With the rapid development of computing power and artificial intelligence, IoT devices equipped with ubiquitous sensors are gradually installed with intelligence. People can enjoy many conveniences with intelligent devices, such as face recognition, video understanding, and motion estimation. Currently, deep neural networks are the mainstream technology in intelligent mobile applications. Inspired by DNN model partition schemes, the paradigm of edge computing could be utilized collaboratively to improve the effectiveness of intelligent task execution in IoT devices. However, due to the dynamics of the wireless network environment and the increasing number of IoT devices, a DNN partition policy without adequate consideration would pose a significant challenge to the efficiency of task inference. Moreover, the shortage and high rental cost of edge computing resources make the optimization of DNN-based task execution more difficult. To cope with those situations, we propose a joint method by a self-adaptive DNN partition with cost-effective resource allocation to facilitate collaborative computation between IoT devices and edge servers. Our proposed online algorithm can be proved to ensure the overall rental cost within an upper bound above the optimal solution while guaranteeing the latency for DNN-based task inference. To evaluate the performance of our strategy, we conduct extensive trace-driven illustrative studies and show that the proposed method can achieve sub-optimal results and outperforms other alternative methods.

Mobile edge computing (MEC) is an emerging computing and communication model that brings extremely resourceful servers at a network's edge. This reduces the latency associated with cloud computing as applications executing on mobile devices do not always need to communicate with a cloud server because most of the functionalities provided by the cloud server are available on an edge server. Moreover, with the emergence of 5G really high amount of bandwidth is available to a mobile device, hence communication latency between the edge server and a mobile device is low. The combination of 5G and edge computing has a potential to better enable many interesting time-critical and/or computationally expensive computing and communication use cases, for example, connected cars. The stated combination enables development of computationally expensive applications for mobile devices and sensors nodes as these devices can offload their computational tasks to the edge server. Task offloading is a promising feature of edge computing, therefore here priority-class-based methods are presented to allocated CPU cycles to an offloaded task at the edge server. In one of the presented methods, a mobile device or a senor node assign their task to one of the available priority classes, and then offload the task to the edge server. Afterwards, the edge server assigns CPU cycles to such a task depending upon its priority class. Similarly, another method is presented that assigns priority to each device based on the type of tasks the device is offloading. While assigning CPU cycles to different offloaded tasks, the edge server uses their device priority. The simulation results demonstrate that the presented methods are capable of providing service differentiation.

Mobile edge computing (MEC) plays a pivotal role in the Internet of Vehicles and the Internet of Things. Edge server deployment is the initial step in establishing edge computing systems, which impact the overall system performance significantly. Besides, the performance of an edge computing system is also contingent upon the type of service deployed on servers, in the case of the same server deployment, different deployment of services will bring different profits. Most current studies concentrate solely on the former aspect, neglecting the optimization of service deployment in MEC system. In this article, we proposed a two-step method KPSOP for edge server and edge service deployment, aiming to reduce time delay, balance load, and improve the profit of MEC system, and KPSOP includes clustering algorithm and heuristic algorithm. We considered the location distribution of base stations, the task requests of vehicle users, the resource limitations of edge servers, etc. First, the edge server deployment was completed with the goal of minimizing time delay and load balancing. Second, the service deployment was completed with the goal of maximizing edge server profit. The experiments were based on real world base station information. The simulation results validate that our algorithm is more stable and converges faster. In addition, compared to other algorithms, it performs better in load balance and increasing profit.

The real-time communication requirement of the Internet of Things (IoT) applications promotes the convergence of IoT and Mobile Edge Computing (MEC). The MEC paradigm greatly shortens the IoT service delay by leveraging cloudlets (edge servers) of MEC in the proximity of IoT devices. Considering limited computing and storage resources in an MEC network, it is challenging to provide efficient IoT-enabled service provisioning in such a network. In this article, we study the service home identification problem of service provisioning for multi-source IoT applications in an MEC network, by identifying a service home (cloudlet) of each multi-source IoT application for its data processing, querying and storage. Each multi-source IoT application consists of multiple sources located at different geographical locations and each source uploads its data stream via a gateway (its nearby access point) to the MEC network and the uploaded data then is aggregated with the stream data of the other sources of the IoT application at the service home. We here focus on two novel service home identification problems: the service operational cost minimization problem with the aim to minimize the total service operational cost by admitting as many multi-source IoT applications as possible, and the online throughput maximization problem with the aim to maximize the number of multi-source IoT application requests admitted. We first show that both the problems are NP-hard. We then formulate an Integer Linear Programming (ILP) solution to the service operational cost minimization problem, and propose a randomized algorithm with high probability and a deterministic approximation algorithm respectively, at moderate resource capacity violations. We third develop an efficient heuristic algorithm for the problem without any resource violation. Furthermore, we deal with the online throughput maximization problem under an assumption that multi-source IoT application requests arrive one by one without the knowledge of future arrivals, for which we formulate an Integer Linear Programming (ILP) solution to its offline version, followed by devising an online algorithm with competitive ratio. We finally evaluate the performance of the proposed algorithms through experimental simulations. Simulation results demonstrate that the proposed algorithms are promising, and outperform their comparison counterparts.

Traditionally Deep Neural Networks (DNNs) services are deployed in the cloud, due to the computation-intensive DNN models. In recent years, as emerging edge computing provides new possibilities for DNN applications, we have opportunities to process DNN models in the cloud and on the device collaboratively, i.e., edge-cloud computing. Since cloud and edge devices demonstrate significant diversity on inference latency, network transmission overhead, memory capacity and power consumption, it is a big challenge to determine the DNN model deployment in the cloud and on edge devices. In this paper, we characterize the behaviours of three types of DNN models, i.e., CNN, LSTM and MLP, on four types of platforms, i.e., server-class CPU, server-class GPU, embedded device with GPU, and smart-phones. Our experimental results demonstrate that we can carefully tune a deployment strategy for DNN models in the cloud, and on big and (or) little cores of the edge device, to balance performance and power consumption.

For several reasons, the cloud computing paradigm, e.g., mobile edge computing (MEC), is suffering from the problem of privacy issues. MEC servers provide personalization services to mobile users for better QoE qualities, but the ongoing migrated data from the source edge server to the destination edge server cause users to have privacy concerns and unwillingness of self‐disclosure, which further leads to a sparsity problem. As a result, personalization services ignore valuable user profiles across edges where users have accounts in and tend to predict users’ potential purchases with insufficient sources, thereby limiting further improvement of QoE through personalization of the contents. This paper proposes a novel model, called CEPTM, which (1) collects mobile user data across multiple MEC edge servers, (2) improves the users’ experience in personalization services by loading collected diverse data, and (3) lowers their privacy concern with the improved personalization. This model also reveals that famous topics in one edge server can migrate into several other edge servers with users’ favorite content tags and that the diverse types of items could increase the possibility of users accepting the personalization service. In the experiment section, we use exploratory factor analysis to mathematically evaluate the correlations among those factors that influence users’ information disclosure in the MEC network, and the results indicate that CEPTM (1) achieves a high rate of personalization acceptance due to the availability of more data as input and highly diverse personalization as output and (2) gains the users’ trust because it collects user data while respecting individual privacy concerns and providing better personalization. It outperforms a traditional personalization service that runs on a single‐edge server. This paper provides new insights into MEC diverse personalization services and privacy problems, and researchers and personalization providers can apply this model to merge popular users’ like trends throughout the MEC edge servers and generate better data management strategies.

A Novel Lyapunov based Dynamic Resource Allocation for UAVs-assisted Edge Computing

Mobile edge computing (MEC) provides a promising approach to significantly reduce network operational cost and improve quality of service (QoS) of mobile users by pushing computation resources to the network edges, and enables a scalable Internet of Things (IoT) architecture for time-sensitive applications (e-healthcare, real-time monitoring, and so on.). However, the mobility of mobile users and the limited coverage of edge servers can result in significant network performance degradation, dramatic drop in QoS, and even interruption of ongoing edge services; therefore, it is difficult to ensure service continuity. Service migration has great potential to address the issues, which decides when or where these services are migrated following user mobility and the changes of demand. In this paper, two conceptions similar to service migration, i.e., live migration for data centers and handover in cellular networks, are first discussed. Next, the cutting-edge research efforts on service migration in MEC are reviewed, and a devisal of taxonomy based on various research directions for efficient service migration is presented. Subsequently, a summary of three technologies for hosting services on edge servers, i.e., virtual machine, container, and agent, is provided. At last, open research challenges in service migration are identified and discussed.

Mobile Edge Computing has already become a new paradigm to reduce the latency in data transmission for resource-limited mobile devices by offloading computation tasks onto edge servers. However, for mobility-aware computation-intensive services, existing offloading strategies cannot handle the offloading procedure properly because of the lack of collaboration among edge servers. A data stream application is partitionable if it can be presented by a directed acyclic dataflow graph, which makes cross-edge collaboration possible. In this paper, we propose a cross-edge computation offloading (CCO) framework for partitionable applications. The transmission, execution and coordination cost, as well as the penalty for task failure, are considered. An online algorithm based on Lyapunov optimization is proposed to jointly determine edge site-selection and energy harvesting without priori knowledge. By stabilizing the battery energy level of each mobile device around a positive constant, the proposed algorithm can obtain asymptotic optimality. The-oretical analysis about the complexity and the effectiveness of the proposed framework is provided. Experimental results based on a real-life dataset corroborate that CCO can achieve superior performance compared with benchmarks where crossedge collaboration is not allowed.

In this paper, we investigate a new mobile blockchain-enabled edge computing (MBEC) network, where mobile users can join the empowered process of public blockchains and meanwhile offload computation-intensive mining tasks to the mobile edge computing (MEC) server. However, the trustiness of the MEC server and the fairness of computation resources allocated by the MEC server for each user become key challenges. To tackle these challenges, we consider an untrusted MEC server and propose a nonce hash computing ordering (HCO) mechanism in MBEC networks. Then we formulate nonce hash computing demands of an individual user as a non-cooperative game that maximizes the personal revenue. Moreover, we also analyze the existence of Nash equilibrium of the non-cooperative game and design an alternating optimization algorithm to achieve the optimal nonce selection strategies for all users. With the proposed HCO mechanism, the MEC server can provide much fairer computation resources for all users, and we can achieve the optimal nonce strategies of hash computing demands by using the proposed alternating optimization algorithm. Numerical results demonstrate that the proposed HCO mechanism can provide fairer computation resource allocation than the traditional weighted round-robin mechanism, and further verify the effectiveness of this alternating optimization algorithm.

This article aims to improve the accuracy of real-time image recognition in the context of the Internet of Things (IoT), reduce the core network pressure of the IoT and the proportion of IoT broadband, and meet people’s demand for internet image transmission. An intelligent image fusion system based on mobile edge computing (MEC) and deep learning is proposed, which can extract the features of images and optimize the sum of intra-class distance and inter-class distance relying on the hierarchical mode of deep learning, and realize distributed computing with the edge server and base station. Through comparison with other algorithms and strategies on the text and character data sets, the effectiveness of the constructed system is verified in the performance of the algorithm and the IoT. The results reveal that the application of the unsupervised learning hierarchical discriminant analysis (HDA) has better accuracy and recall in various databases compared with conventional image recognition algorithms. When the sum intra-class and inter-class distance K is 2, the accuracy of the algorithm can be as high as 98%. The combination of MEC and layered algorithms effectively reduces the pressures of core network and shortens the response time, greatly reduces the broadband occupancy ratio. The performance of IoT is increased by 37.03% compared with the general extraction and common cloud computing. Image recognition based on the MEC architecture can reduce the amount of network transmission and reduce the response time under the premise of ensuring the recognition rate, which can provide a theoretical basis for the research and application of user image recognition under the IoT.

Mobile edge computing enables the execution of compute-intensive applications, e.g. deep learning applications, on the end devices with limited computation resources. However, the deep learning applications bring the performance bottleneck in mobile edge computing, due to the movements of a large amount of data incurred by the large number of layers and millions of weights. In this paper, the computing model for parallel deep learning applications in mobile edge computing is proposed, by considering the occupancy allocation of processors, cost of context switch, and multi-processors in edge server and remote cloud. The problem of minimizing the completion time for deep learning applications is formulated, and the NP-hardness of the problem is proved. To solve the problem, an integrated algorithm by merging and scheduling is proposed. Moreover, a real-world distributed platform is developed for evaluating the proposed algorithm. Experimental results show that, the completion time of deep learning application for the proposed algorithm is decreased by 63% and 75%, respectively, without extra control costs, compared with the existing algorithms.

Autonomous vehicles and mobile edge computing’s confluence have raised an innovative model for immediate decision-making and improved computational abilities. But, enhancing vehicle management systems to guarantee effective enactment remains an important challenge. Present approaches regularly depend on intricate algorithms and multiple sensors, that result in improved computational overhead and potential latency. The current study resolves the present gap by offering a new hybrid framework, which synergistically mixes optimization algorithms and deep neural networks through the advantages of mobile edge computing. Precisely, this research presents a hybrid model for autonomous vehicle management by integrating a refined version of the RPO or Pelican Optimizer with deep neural networks attuned to mobile edge computing environments. The chief contributions of the present study have been threefold: (1) the improvement of a particular autonomous driving method optimized for mobile edge computing platforms; (2) the arrangement of an optimized MobileNet method employing the RPO algorithm that uses LiDAR sensor data for effective object recognition and path design; and (3) the construction of an indoor vehicle prototype by mean of a microcontroller and LiDAR sensors, after a comprehensive performance evaluation of inference models, and analyzing the trade-offs between input size and computational effectiveness. Experimental outcomes show the efficiency and reliability of the suggested hybrid model, through improving autonomous vehicle management and decision-making abilities within the mobile edge computing paradigm. The current study contributes to the enhancement of autonomous vehicle research and provides an innovative solution for effective and precise vehicle control within edge computing environments.

CICIDS2017 is a well-known benchmark dataset for intrusion detection system (IDS) model creation. However, several studies revealed that the CICIDS2017 dataset still has several weaknesses. This research aims to improve R-CICIDS2017 by removing unrealistic features. After removing unrealistic features, we use a deep neural network (DNN) to predict the R-CICIDS-2017 dataset. We focus our predictions on attacks labeled denial of service (DoS) and distributed DoS (DDoS) in the dataset, as these attacks are the most frequent in the Internet of Things (IoT) world. Because implementation in IoT must be lightweight, we only use 0.1 of the entire dataset. We develop quantization for several data types to determine which model is best suited for running on IoT edge devices. The test results show that the Shannon entropy of the sampled and original datasets for each label is identical. This answer indicates that the DNN model for IDS on IoT with R-CICIDS2017 uses a low-complexity algorithm. The DNN test on CICIDS2017 and R-CICIDS2017 reveals that DNNs using unrealistic datasets exhibit better performance; however, this is not feasible for implementation. DNN with R-CICIDS2017 is optimal because its performance surpasses that of GNB and uses features that can be implemented in the IoT environment. Finally, the test shows that dynamic quantization has better accuracy loss than static int 8 quantization, and its model size is better than static float 16 quantization. We obtain a DNN model that achieves optimum efficiency for IDS deployment on edge devices within IoT networks, utilizing R-CICIDS2017.