Tactile Internet (TI) requires ultra-low latency and high reliability to ensure stability and transparency in touch-enabled teleoperation. However, variable delays and packet loss present significant challenges to maintaining immersive haptic communication. To address this, we propose a predictive framework that integrates Discrete Mode Decomposition (DMD) with Shapley Mode Value (SMV) for accurate and timely haptic signal prediction. DMD decomposes haptic signals into interpretable intrinsic modes, while SMV evaluates each mode's contribution to prediction accuracy, which is well-aligned with the goal-oriented semantic communication. Integrating SMV with DMD further accelerates inference, enabling efficient communication and smooth teleoperation even under adverse network conditions.Extensive experiments show that DMD+SMV, combined with a Transformer architecture, outperforms baseline methods significantly. It achieves 98.9% accuracy for 1-sample prediction and 92.5% for 100-sample prediction, as well as extremely low inference latency: 0.056 ms and 2 ms, respectively. These results demonstrate that the proposed framework has strong potential to ease the stringent latency and reliability requirements of TI without compromising performance, highlighting its feasibility for real-world deployment in TI systems.

As touch-enabled teleoperation applications envisioned in the Tactile Internet are highly sensitive to latency, minimizing Medium Access Delay is essential. Two commonly employed resource scheduling schemes for this purpose are Semi-Persistent Scheduling (SPS) and Grant-Free Access (GFA) with M-repetition transmission. While SPS is well-suited for periodic traffic and GFA is for sporadic transmissions, the bursty and dynamic nature of haptic traffic in the Tactile Internet necessitates a hybrid scheduling policy that adapts to varying traffic states. To address this challenge, we propose a Diffusion-based Offline Reinforcement Learning framework that efficiently learns effective scheduling policies from static datasets, rather than online interaction. Using synthetic traffic datasets from the human experiment and the 5G emulator, the simulation results show a general improvement over other offline RL benchmarks.

With the rise of mega-satellite constellations, the integration of hierarchical non-terrestrial and terrestrial networks has become a cornerstone of 6G coverage enhancements. In these hierarchical satellite networks, controllers manage satellite switches within their assigned domains. However, the high mobility of LEO satellites and field-of-view (FOV) constraints pose fundamental challenges to efficient domain partitioning. Centralized control approaches face scalability bottlenecks, while distributed architectures with onboard controllers often disregard FOV limitations, leading to excessive signaling overhead. LEO satellites outside a controller’s FOV require an average of five additional hops, resulting in a 10.6-fold increase in response time. To address these challenges, we propose Eunomia, a three-step domain-partitioning framework that leverages movement-aware FOV segmentation within a hybrid control plane combining ground stations and MEO satellites. Eunomia reduces control plane latency by constraining domains to FOV-aware regions and ensures single-hop signaling. It further balances traffic load through spectral clustering on a Control Overhead Relationship Graph and optimizes controller assignment via the Kuhn-Munkres algorithm. We implement Eunomia on the Plotinus emulation platform with realistic constellation parameters. Experimental results demonstrate that Eunomia reduces request loss by up to 58.3%, control overhead by up to 50.3%, and algorithm execution time by 77.7%, significantly outperforming current state-of-the-art solutions.

Haptic teleoperation typically demands sub-millisecond latency and ultra-high reliability (99.999%) in Tactile Internet. At a 1 kHz haptic signal sampling rate, this translates into an extremely high packet transmission rate, posing significant challenges for timely delivery and introducing substantial complexity and overhead in radio resource allocation. To address this critical challenge, we introduce a novel DL model that estimates force feedback using multi-modal input, i.e. both force measurements from the remote side and local operator motion signals. The DL model can capture complex temporal features of haptic time-series with the use of CNN and LSTM layers, followed by a transformer encoder, and autoregressively produce a highly accurate estimation of the next force values for different teleoperation activities. By ensuring that the estimation error is within a predefined threshold, the teleoperation system can safely relax its strict delay requirements. This enables the batching and transmission of multiple haptic packets within a single resource block, improving resource efficiency and facilitating scheduling in resource allocation. Through extensive simulations, we evaluated network performance in terms of reliability and capacity. Results show that, for both dynamic and rigid object interactions, the proposed method increases the number of reliably served users by up to 66%.

Enabling video-haptic radio resource slicing in the Tactile Internet requires a sophisticated strategy to meet the distinct requirements of video and haptic data, ensure their synchronized transmission, and address the stringent latency demands of haptic feedback. This paper introduces a Deep Reinforcement Learning-based radio resource slicing framework that addresses video-haptic teleoperation challenges by dynamically balancing radio resources between the video and haptic modalities. The proposed framework employs a refined reward function that considers latency, packet loss, data rate, and the synchronization requirements of both modalities to optimize resource allocation. By catering to the specific service requirements of video-haptic teleoperation, the proposed framework achieves up to a 25 % increase in user satisfaction over existing methods, while maintaining effective resource slicing with execution intervals up to 50 ms.

Local differential privacy techniques for numerical data typically transform a dataset to ensure a bound on the likelihood that, given a query, a malicious user could infer information on the original samples. Queries are often solely based on users and their requirements, limiting the design of the perturbation to processes that, while privatizing the results, do not jeopardize their usefulness. In this paper, we propose a privatization technique called Zeal, where perturbator and aggregator are designed as a unit, resulting in a locally differentially private mechanism that, by-design, improves the compressibility of the perturbed dataset compared to the original, saves on transmitted bits for data collection and protects against a privacy vulnerability due to floating point arithmetic that affects other state-of-the-art schemes. We prove that the utility error on querying the average and median is invariant to the bias introduced by Zeal in a wide range of conditions, and that under the same circumstances, Zeal also guarantees protection against the aforementioned vulnerability. Moreover, we show that in many scenarios Zeal can outperform other privatization techniques in terms of utility error, compression and data transmission efficiency. Our experiments show up to 94 % improvements in compression and up to 95 % more efficient data transmissions with respect to the original.

The low-orbit mega-constellation network (LMCN) is an important part of the space-air-ground integrated network system. An effective satellite-ground interconnection design can result in a stable constellation topology for LMCNs. A naïve solution is accessing the satellite with the longest remaining service time (LRST), which is widely used in previous designs. The Coordinated Satellite-Ground Interconnecting (CSGI), the state-of-the-art algorithm, coordinates the establishment of ground-satellite links (GSLs). Compared with existing solutions, it reduces latency by 19% and jitter by 70% on average. However, CSGI only supports the scenario where terminals access only one satellite, and cannot fully utilize the multi-access capabilities of terminals. Additionally, CSGI's high computational complexity poses deployment challenges. To overcome these problems, we propose the Classification-based Longest Remaining Service Time (C-LRST) algorithm. C-LRST supports the actual scenario with multi-access capabilities. It adds optional paths during routing with low computational complexity, improving end-to-end communications quality. We conduct our 1000 s simulation from Brazil to Lithuania on the open-source platform Hypatia. Experiment results show that compared with CSGI, C-LRST reduces the latency and increases the throughput by approximately 60% and 40%, respectively. In addition, C-LRST's GSL switchings number is 14, whereas CSGI is 23. C-LRST has better link stability than CSGI.

This paper studies the computational offloading of CNN inference in dynamic multi-access edge computing (MEC) networks. To address the uncertainties in communication time and edge servers’ available capacity, we propose a novel semantic compression method, autoencoder-based CNN architecture (AECNN), for effective semantic extraction and compression in partial offloading. In the semantic encoder, we introduce a feature compression module based on the channel attention mechanism in CNNs, to compress intermediate data by selecting the most informative features. Additionally, to further reduce communication overhead, we leverage entropy encoding to remove the statistical redundancy in the compressed data. In the semantic decoder, we design a lightweight decoder to reconstruct the intermediate data through learning from the received compressed data to improve accuracy. To effectively trade-off communication, computation, and inference accuracy, we design a reward function and formulate the offloading problem of CNN inference as a maximization problem with the goal of maximizing the average inference accuracy and throughput over the long term. To address this maximization problem, we propose a graph reinforcement learning-based AECNN (GRL-AECNN) method, which outperforms existing works DROO-AECNN, GRL-BottleNet++ and GRL-DeepJSCC under different dynamic scenarios. This highlights the advantages of GRL-AECNN in offloading decision-making for CNN inference tasks in dynamic MEC.

IoT applications often require devices to continuously send huge amounts of sensor data to the cloud in order to detect anomalies. This paper proposes a novel preprocessing stage selecting small portions of the sensor data worth sending for further analysis, resulting in significant savings in transmission costs and processing time in the cloud, down to less than 1% of the complete stream, while achieving comparable detection results.

The increasing prevalence of gigapixel resolutions has presented new challenges for crowd counting. Such resolutions are far beyond the memory and computation limits of current GPUs, and available deep neural network architectures and training procedures are not designed for such massive inputs. Although several methods have been proposed to address these challenges, they are either limited to downsampling the input image to a small size, or borrowing from other gigapixel tasks, which are not tailored for crowd counting. In this paper, we propose a novel method called GigaZoom, which iteratively zooms into the densest areas of the image and refines coarser density maps with finer details. We show that GigaZoom obtains the state-of-the-art for gigapixel crowd counting and improves the accuracy of the next best method by 42%.

The proliferation of virtual reality (VR) content within social networks amplifies the importance of reducing rendering delay, as seamless interactions in shared virtual spaces are crucial for fostering social connections. Integrating VR streaming with social networks requires innovative solutions to address the unique challenges arising from the interaction between immersive experiences and social interactions. In this context, our research focuses on minimizing rendering delay for VR streaming in social networks, leveraging the synergistic benefits of edge computing. However, in scenarios where end-users experience poor channel quality, the rendering delay is prolonged due to lower data rates. To this end, a double reconfigurable intelligent surface (RIS) is employed to assist in improving the channel efficiency for end-user devices located in weak signal reception zones. We formulate an optimization problem related to the rendering resource allocation in edge servers and the distribution of downlink bandwidth for VR content from the edge server as a quadratically constrained quadratic problem. The non-convex optimization problem has been solved by dividing the problem into three sub-problems and solved using the block coordinate descent (BCD) method. In this study, we aim to enhance the overall quality of VR experiences in social settings, paving the way for more compelling and interactive virtual interactions where end-users have low wireless channel reception Quality.

In this paper, we propose a model-mediated tele-operation (MMT) system using an octree-based model (OBM) to spatially map the environment impedance for the emerging use cases in Tactile Internet. Different from the existing just-noticeable-difference (JND) based MMT, our method avoids continuous transmission of environment impedance. Moreover, it allows the local model to generate accurate force feedback and reduces the number of model updates. Furthermore, the OBM can be deployed with or without previous knowledge of the environment. An online estimation of the OBM is proposed using a JND and a rate-of-change threshold. An offline estimation method is also proposed when the geometry and impedance parameters of the remote environment are known. In addition, a point cloud-based force rendering algorithm is tailored to use the OBM, thereby allowing the generating of force feedback for complex environments. An experiment without human-in-the-loop was conducted, showing that for an online estimated OBM, the accuracy of the force feedback was improved by up to 44 percent while using less than half the number of model updates when compared to JND-based MMT. Another experiment with a human operator interacting with a virtual environment showed that using an offline estimated OBM improves the accuracy of the force feedback and is reliable against packet loss and short temporal breakdown of the communication link.

Exponential growth in data collection is creating significant challenges for data storage and analytics latency. Approximate Query Processing (AQP) has long been touted as a solution for accelerating analytics on large datasets, however, there is still room for improvement across all key performance criteria. In this paper, we propose a novel histogram-based data synopsis called PairwiseHist that uses recursive hypothesis testing to ensure accurate histograms and can be built on top of data compressed using Generalized Deduplication (GD). We thus show that GD data compression can contribute to AQP. Compared to state-of-the-art AQP approaches, Pairwise-Hist achieves better performance across all key metrics, including 2.6× higher accuracy, 3.5× lower latency, 24× smaller synopses and 1.5--4× faster construction time.

The exponential growth of data generated by the Internet of Things presents significant challenges for data communication, storage, and analytics. Consequently, organizations often face high costs when attempting to leverage their own data. Novel techniques that holistically optimize data storage and analytics in IoT systems are therefore required. One promising approach is generalized deduplication (GD), which is a lossless compression technique that delivers high compression while also enabling low-cost random access directly on compressed data. In this article, we introduce GreedyGD, a novel GD data compression algorithm that offers reliable, efficient data analytics, along with more compression and faster runtime compared to previous GD compressors. Evaluating GreedyGD on 18 real-world datasets revealed excellent performance: a 11.2× speed-up, 1.6× more compression, and more accurate and reliable analytics while using 4× less data compared to previous GD compressors.

Local differential privacy techniques for numerical data typically transform a dataset to ensure a bound on the likelihood that, given a query, a malicious user could infer information on the original samples. Queries are often solely based on users and their requirements, limiting the design of the perturbation to processes that, while privatizing the results, do not jeopardize their usefulness. In this paper, we propose a privatization technique called Zeal, where perturbator and aggregator are designed as a unit, resulting in a locally differentially private mechanism that, by-design, improves the compressibility of the perturbed dataset compared to the original, saves on transmitted bits for data collection and protects against a privacy vulnerabilities due to floating point arithmetic that affect other state-of-the-art schemes. We prove that the utility error on querying the average is invariant to the bias introduced by Zeal in a wide range of conditions, and that under the same circumstances, Zeal also guarantee protection against the aforementioned vulnerability. Our numerical results show up to 94 % improvements in compression and up to 95 % more efficient data transmissions, while keeping utility errors within 2 %.

Deploying deep neural networks (DNNs) on IoT and mobile devices is a challenging task due to their limited computational resources. Thus, demanding tasks are often entirely offloaded to edge servers which can accelerate inference, however, it also causes communication cost and evokes privacy concerns. In addition, this approach leaves the computational capacity of end devices unused. Split computing is a paradigm where a DNN is split into two sections; the first section is executed on the end device, and the output is transmitted to the edge server where the final section is executed. Here, we introduce dynamic split computing, where the optimal split location is dynamically selected based on the state of the communication channel. By using natural bottlenecks that already exist in modern DNN architectures, dynamic split computing avoids retraining and hyperparameter optimization, and does not have any negative impact on the final accuracy of DNNs. Through extensive experiments, we show that dynamic split computing achieves faster inference in edge computing environments where the data rate and server load vary over time.

This paper studies the computational offloading of CNN inference in device-edge co-inference systems. Inspired by the emerging paradigm semantic communication, we propose a novel autoencoder-based CNN architecture (AECNN), for effective feature extraction at end-device. We design a feature compression module based on the channel attention method in CNN, to compress the intermediate data by selecting the most important features. To further reduce communication overhead, we can use entropy encoding to remove the statistical redundancy in the compressed data. At the receiver, we design a lightweight decoder to reconstruct the intermediate data through learning from the received compressed data to improve accuracy. To fasten the convergence, we use a step-by-step approach to train the neural networks obtained based on ResNet-50 architecture. Experimental results show that AECNN can compress the intermediate data by more than 256 × with only about 4% accuracy loss, which outperforms the state-of-the-art work, BottleNet++. Compared to offloading inference task directly to edge server, AECNN can complete inference task earlier, in particular, under poor wireless channel condition, which highlights the effectiveness of AECNN in guaranteeing higher accuracy within time constraint.

The number of IoT devices is expected to continue its dramatic growth in the coming years and, with it, a growth in the amount of data to be transmitted, processed and stored. Compression techniques that support analytics directly on the compressed data could pave the way for systems to scale efficiently to these growing demands. This paper proposes two novel methods for preprocessing a stream of floating point data to improve the compression capabilities of various IoT data compressors. In particular, these techniques are shown to be helpful with recent compressors that allow for random access and analytics while maintaining good compression. Our techniques improve compression with reductions up to 80% when allowing for at most 1% of recovery error.

Many deep learning tasks require annotations that are too time consuming for human operators, resulting in small dataset sizes. This is especially true for dense regression problems such as crowd counting which requires the location of every person in the image to be annotated. Techniques such as data augmentation and synthetic data generation based on simulations can help in such cases. In this paper, we introduce PromptMix, a method for artificially boosting the size of existing datasets, that can be used to improve the performance of lightweight networks. First, synthetic images are generated in an end-to-end data-driven manner, where text prompts are extracted from existing datasets via an image captioning deep network, and subsequently introduced to text-to-image diffusion models. The generated images are then annotated using one or more high-performing deep networks, and mixed with the real dataset for training the lightweight network. By extensive experiments on five datasets and two tasks, we show that PromptMix can significantly increase the performance of lightweight networks by up to 26%.

This demo presents a working prototype of a live video surveillance IoT system that provides multi-class privacy preservation. The system can use a Raspberry Pi or a laptop to perform real-time face detection, compression and encryption on live video feeds. We design and implement novel methods to accelerate face detection and reduce miss detection for fast moving subjects. The implemented prototype demonstrates the feasibility to provide multi-class privacy-preserving feature and fine-grained access control of video content for real-time IoT applications.

The Internet of Things (IoT) has brought about exponential growth in sensor data. This has led to increasing demands for efficient and novel data transmission, storage, and analytics solutions for sustainable IoT ecosystems. It has been shown that the generalized deduplication (GD) compression algorithm offers not only competitive compression ratio and throughput but also random access properties that enable direct analytics of compressed data. In this article, we thoroughly stress test existing methods for direct analytics of GD compressed data with a diverse collection of 103 data sets, identify the need to optimize GD for analytics, and develop a new version of GD to this end. We also propose the generalized deduplication-enabled approximate edge analytics (GLEAN) framework. This framework applies the aforementioned analytics techniques at the Edge server to deliver end-to-end lossless data compression and high-quality Edge analytics in the IoT, thereby addressing challenges related to data transmission, storage, and analytics. Impressive analytics performance was achieved using this framework, with a median increase in$k$-means clustering error of just 2% relative to analytics performed on uncompressed data, while running$7.5\times $faster and requiring$3.9\times $less storage at the Edge server compared to universal compressors.

End-users under poor wireless network coverage generally suffer from underutilization of bandwidth. This adversely affects the overall performance of task offloading to the edge server. In this work, we study a Reconfigurable Intelligent Surface (RIS)-assisted wireless network that enables end-user's devices under weak signal reception areas to enhance their offloading opportunities for delay minimization. It becomes a challenging task to allocate uploading bandwidth allocation for the offloaded tasks from end-user devices under different signal coverage areas. We formulate the optimization problem of bandwidth allocation for the offloaded tasks in the edge server and the offloading decisions as a quadratically constrained quadratic problem. We exploit a semi-definite relaxation (SDR) method to solve the problem. Moreover, during optimization, we minimize the adverse impact of bandwidth allocation for poor end-users on good end-users performance. From extensive simulation results, we show remarkably elevated improvement in delay reduction with RIS assistance compared to other baselines, increasing the number and ratio of end-user devices under good and poor signal reception areas.

This paper studies the computational offloading of CNN inference in dynamic multi-access edge computing (MEC) networks. To address the uncertainties in communication time and Edge servers' available capacity, we use early-exit mechanism to terminate the computation earlier to meet the deadline of inference tasks. We design a reward function to trade off the communication, computation and inference accuracy, and formu-late the offloading problem of CNN inference as a maximization problem with the goal of maximizing the average inference accuracy and throughput in long term. To solve the maxi-mization problem, we propose a graph reinforcement learning-based early-exit mechanism (GRLE), which outperforms the state-of-the-art work, deep reinforcement learning-based online offloading (DROO) and its enhanced method, DROO with early-exit mechanism (DROOE), under different dynamic scenarios. The experimental results show that G RLE achieves the average accuracy up to 3.41 x over graph reinforcement learning (GRL) and 1.45x over DROOE, which shows the advantages of GRLE for offloading decision-making in dynamic MEC.

Human-computer interaction provides pervasive services advocating several exciting interactive systems, such as remote automation, surgery, and rehabilitation. This paper studies a tight synchronization between human and robot hands to establish near to real-time maneuvering. We leverage the computing resources of the participating units of the master domain to determine the useful data by overserving and predicting the immediate reaction of the human hand movement. Moreover, we consider the unreliable wireless channels that lead to packet error during data transmission from the master domain to the controlled domain. In particular, by bringing the concept of edge computing while utilizing the Raspberry Pis's available yet limited computing resources, we aim to determine the balance between useful data and redundant packets without any significant performance degradation. Finally, we implement the proposed synchronization method of human-robot arm interactions in a real testbed and compare the performance with baselines.

This paper studies inference acceleration using distributed convolutional neural networks (CNNs) in collaborative edge computing. To ensure inference accuracy in inference task partitioning, we consider the receptive-field when performing segment-based partitioning. To maximize the parallelization between the communication and computing processes, thereby minimizing the total inference time of an inference task, we design a novel task collaboration scheme in which the overlapping zone of the sub-tasks on secondary edge servers (ESs) is executed on the host ES, named as HALP. We further extend HALP to the scenario of multiple tasks. Experimental results show that HALP can accelerate CNN inference in VGG-16 by 1.7-2.0x for a single task and 1.7-1.8x for 4 tasks per batch on GTX 1080TI and JETSON AGX Xavier, which outperforms the state-of-the-art work MoDNN. Moreover, we evaluate the service reliability under time-variant channel, which shows that HALP is an effective solution to ensure high service reliability with strict service deadline.

This paper studies inference acceleration using distributed convolutional neural networks (CNNs) in collaborative edge computing network. To avoid inference accuracy loss in inference task partitioning, we propose receptive field-based segmentation (RFS). To reduce the computation time and communication overhead, we propose a novel collaborative edge computing using fused-layer parallelization to partition a CNN model into multiple blocks of convolutional layers. In this scheme, the collaborative edge servers (ESs) only need to exchange small fraction of the sub-outputs after computing each fused block. In addition, to find the optimal solution of partitioning a CNN model into multiple blocks, we use dynamic programming, named as dynamic programming for fused-layer parallelization (DPFP). The experimental results show that DPFP can accelerate inference of VGG-16 up to 73% compared with the pre-trained model, which outperforms the existing work MoDNN in all tested scenarios. Moreover, we evaluate the service reliability of DPFP under time-variant channel, which shows that DPFP is an effective solution to ensure high service reliability with strict service deadline.

This paper studies inference acceleration using distributed CNNs in collaborative edge computing network. To ensure no inference accuracy loss in task partitioning, we propose receptive field-based segmentation. To reduce the computation time and communication overhead, we propose a novel collaborative edge computing using fused-layer parallelization to partition a CNN model into multiple blocks. To find the optimal partition of a CNN model, we use dynamic programming, named as DPFP. To address computation heterogeneity of edge servers (ESs), we design a low-complexity search algorithm which can select the optimal subset of collaborative ESs for inference. The experimental results show that DPFP can accelerate inference up to 71% for ResNet-50 and 73% for VGG-16 compared to running the pre-trained models, which outperforms the existing works MoDNN and DeepSlicing. Moreover, we propose an analytical method to estimate the speedup ratio of different GPU platforms by using FLOPs and effective computing capacity. Furthermore, we evaluate the service failure probability under time-variant channel and variation of image sizes, which shows that DPFP is effective to ensure high service reliability with strict service deadline.

Compressive sensing (CS) can provide joint compression and encryption, which is promising to address the challenges of massive sensor data and data security in the Internet of Things (IoT). However, as IoT devices have constrained memory, computing power, and energy, in practice the CS-based computationally secure scheme is shown to be vulnerable to ciphertext-only attack for short-signal length. Although the CS-based perfectly secure scheme has no such vulnerabilities, its practical realization is challenging. In this article, we propose an energy concealment (EC) encryption scheme, a practical realization of the perfectly secure scheme by concealing energy, thereby removing the requirement of an additional secure channel. We propose three different methods to generate sensing matrix to improve energy efficiency using linear feedback shift registers and lagged Fibonacci sequences. Leveraging the signal’s maximum energy in the EC scheme, we design a new measure to evaluate reconstructed signal quality without the knowledge of the original signal. Furthermore, a new CS decoding algorithm is designed by incorporating the knowledge of maximum energy at the decoder, which improves the signal reconstruction quality while reducing the number of measurements. Additionally, our comprehensive security analysis shows that the EC scheme is secure against various cryptographic attacks. We implement the EC scheme using the three different ways of generating the sensing matrix in the resource-constrained TelosB mote using the Contiki operating system. The experimental results demonstrate that the EC scheme outperforms advanced encryption standard in terms of code memory footprint and total energy consumption.

In this paper, we design the multi-class privacy-preserving cloud computing scheme (MPCC) leveraging compressive sensing for compact sensor data representation and secrecy for data encryption. Three variants of the MPCC scheme are proposed, realizing statistical decryption for smart meters, and data anonymization for images and electrocardiogram signals. The proposed MPCC variants achieve two-class secrecy, one for the superuser who can retrieve the exact sensor data and the other for the semi-authorized user, who can only obtain the statistical data such as mean, variance, etc., or the signals without sensitive part of information, depending on which variant of the MPCC is used. MPCC scheme allows computationally expensive sparse signal recovery to be performed at the cloud without compromising data confidentiality to the cloud service providers. In this way, it mitigates the issues in data transmission energy and storage caused by massive IoT sensor data, as well as the increasing concerns about IoT data privacy in cloud computing. We show that the MPCC scheme has lower computational complexity at the IoT sensor device and data end-users than the state-of-the-art schemes. Experimental results on three datasets, i.e., smart meter, electrocardiogram, and images, demonstrate the MPCC’s performance in statistical decryption and data anonymization.

Deploying deep learning models in time-critical applications with limited computational resources, for instance in edge computing systems and IoT networks, is a challenging task that often relies on dynamic inference methods such as early exiting. In this paper, we introduce a novel architecture for early exiting based on the vision transformer architecture, as well as a fine-tuning strategy that significantly increase the accuracy of early exit branches compared to conventional approaches while introducing less overhead. Through extensive experiments on image and audio classification as well as audiovisual crowd counting, we show that our method works for both classification and regression problems, and in both single- and multi-modal settings. Additionally, we introduce a novel method for integrating audio and visual modalities within early exits in audiovisual data analysis, that can lead to a more fine-grained dynamic inference.

In this paper, we study the deadline-aware task data offloading in edge-cloud computing systems. The hard-deadline tasks strictly demand to be processed within their delay deadline, whereas the deadline can be relaxed for the soft-deadline tasks. Generally, edge computing aims to shorten the transmission delay between the remote cloud and the end-user, however, at the cost of limited computing capability. Therefore, it is challenging to decide where to offload the hard- and soft-deadline tasks based on the average delay and the service price set by the edge and cloud servers. Both edge and cloud servers aim to maximize their revenue by selling the computational resources at the optimal price. Interestingly, a Wardrop equilibrium is reached, considering that each task is considered independently to be offloaded to a suitable location. The numerical results demonstrate that the proposed price- and deadline-sensitive task offloading policy reaches the equilibrium and finds the optimal location for processing while maximizing the revenue of both edge and cloud servers.

Given the ever increasing volume of data generated by the Internet of Things, data compression plays an essential role in reducing the cost of data transmission and storage. However, it also introduces a barrier, namely decompression, between users and the data-driven insights they require. We propose methods for direct analytics of compressed data based on the Generalized Deduplication compression algorithm. When applied to data clustering, the accuracy of the best performing method differs by merely 1-5% when compared to analytics performed upon the uncompressed data. However, it runs four times faster, accesses only 14% as much data and requires significantly less storage since the data is always compressed. These results show that it is possible to simultaneously reap the benefits of compression and accurate, high-speed analytics in many applications.

Deploying deep learning services for time-sensitive and resource-constrained settings such as IoT using edge computing systems is a challenging task that requires dynamic adjustment of inference time. Multi-exit architectures allow deep neural networks to terminate their execution early in order to adhere to tight deadlines at the cost of accuracy. To mitigate this cost, in this paper we introduce a novel method called Multi-Exit Curriculum Learning that utilizes curriculum learning, a training strategy for neural networks that imitates human learning by sorting the training samples based on their difficulty and gradually introducing them to the network. Experiments on CIFAR-10 and CIFAR-100 datasets and various configurations of multi-exit architectures show that our method consistently improves the accuracy of early exits compared to the standard training approach.

We introduce Titchy, which is a compression method for time-series data generated by the Internet of Things. Our proposed method is flexible and has several advantages when applied in the IoT ecosystem: 1) it is able to compress even when only a small amount of memory can be allocated to it; 2) it compresses data in tiny chunks, so it introduces very little latency during online operation and thus, enables frequent and timely updates with compressed data; and 3) it facilitates efficient data storage and retrieval by enabling low-cost random access to the compressed data, eliminating the need to decompress large chunks when only a small amount of data is requested. To evaluate each of these advantages, we have implemented the compressor and conducted extensive experiments with long-term real-world data captured over days or weeks. We also present results for seven state-of-the-art compression methods to act as a baseline. Our evaluation shows that Titchy not only outperforms all seven on random access capability and for frequent transmissions but also provides great compression ratios as well as high compression and decompression speeds.

Recent study has shown that compressive sensing (CS) based computationally secure scheme using Gaussian or Binomial sensing matrix in resource-constrained IoT devices is vulnerable to ciphertext-only attack. Although the CS-based perfectly secure scheme has no such vulnerabilities, the practical realization of the perfectly secure scheme is challenging, because it requires an additional secure channel to transmit the measurement norm. In this paper, we devise a practical realization of a perfectly secure scheme by concealing energy in which the requirement of an additional secure channel is removed. Since the generation of Gaussian sensing matrices is not feasible in resource-constrained IoT devices, approximate Gaussian sensing matrices are generated using linear feedback shift registers. We also demonstrate the implementation feasibility of the proposed perfectly secure scheme in practice without additional complexity. Furthermore, the security analysis of the proposed scheme is performed and compared with the state-of-the-art compressive sensing based energy obfuscation scheme.

In this paper, we design the multi-class privacy-preserving cloud computing scheme (MPCC) leveraging compressive sensing for compact sensor data representation and secrecy for data encryption. The proposed scheme achieves two-class secrecy, one for superuser who can retrieve the exact sensor data, and the other for semi-authorized user who is only able to obtain the statistical data such as mean, variance, etc. MPCC scheme allows computationally expensive sparse signal recovery to be performed at cloud without compromising the confidentiality of data to the cloud service providers. In this way, it mitigates the issues in data transmission, energy and storage caused by massive IoT sensor data as well as the increasing concerns about IoT data privacy in cloud computing. Compared with the state-of-the-art schemes, we show that MPCC scheme not only has lower computational complexity at the IoT sensor device and data consumer, but also is proved to be secure against ciphertext-only attack.

In this paper, we study the priority-aware task data offloading in edge computing-assisted healthcare service provisioning. The edge server aims to provide additional computing resources to the end-users for processing the delay-sensitive tasks. However, at the same time, it becomes a challenging issue when some of the tasks demand lower response time compared to the other tasks. We present a priority-aware task offloading and scheduling strategy that allocates the computing resources to the high-priority tasks. The hard-deadline tasks are processed first. Later, the remaining computing resources are used to tolerate longer average response time for the soft-deadline tasks. Moreover, we derive a lower bound of the average response time for all hard- and soft-deadline tasks. Through extensive simulations, we show that the proposed task scheduling manages to allocate the computing resources of both end-users and edge server to the hard-deadline tasks while scheduling the soft-deadline tasks with low priority.

In cloud storage systems, malicious users may exploit client-side deduplication responses to infer what other users are storing in the cloud. We propose CIDER, a low overhead approach to mitigate this side channel by obfuscating the existence status of data stored in the cloud. We analyze the scheme's ability to obfuscate the side-channel and discuss attacks that a user may still employ and what he could learn from such attacks. Finally, we use simulated and real data to examine the performance under realistic deduplication workloads, revealing that CIDER ends up transmitting less redundant information than similar methods, thus enabling a more efficient utilization of the available bandwidth.

In fog computing, end-users can offload the computation-intensive tasks to the fog node in the proximity. Additionally, the fog nodes also offload these tasks to the cloud and neighboring fog node to seek additional computational resources. In this paper, we propose an offloading strategy in fog computing to minimize the cost that is a weighted sum of energy consumption and total delay for the task processing per end-user. We take the heterogeneous nature of the fog computing nodes that have different CPU frequency to process the tasks. We aim to find an optimal amount of task data to be either locally processed or offloaded to the preferable fog node and the remote cloud under the energy and delay constraints. We then formulate the optimization problem into a non-convex quadratically constrained quadratic program. We further provide an efficient solution to this problem by semidefinite relaxation. Finally, our proposed offloading scheme is evaluated by the simulation to demonstrate the offloading profile and optimal cost of the offloading with a wide range of parameter settings.

We detail a practical compression scheme for lossless compression of time-series data, based on the emerging concept of generalized deduplication. As data is no longer stored for just archival purposes, but needs to be continuously accessed in many applications, the scheme is designed for low-cost random access to its compressed data, avoiding decompression. With this method, an arbitrary bit of the original data can be read by accessing only a few hundred bits in the worst case, several orders of magnitude fewer than state-of-the-art compression schemes. Subsequent retrieval of bits requires visiting at most a few tens of bits. A comprehensive evaluation of the compressor on eight real-life data sets from various domains is provided. The cost of this random access capability is a loss in compression ratio compared with the state-of-the-art compression schemes BZIP2 and 7z, which can be as low as 5% depending on the data set. Compared to GZIP, the proposed scheme has a better compression ratio for most of the data sets. Our method has massive potential for applications requiring frequent random accesses, as the only existing approach with comparable random access cost is to store the data without compression.

Compressive sensing (CS) is regarded as one of the promising solutions for IoT data encryption as it achieves simultaneous sampling, compression, and encryption. Theoretical work in the literature has proved that CS provides computational secrecy. It also provides asymptotic perfect secrecy for Gaussian sensing matrix with constraints on input signal. In this paper, we design an attack decoding algorithm based on block compressed sensing decoding algorithm to perform ciphertext-only attack on real-life time series IoT data. It shows that it is possible to retrieve vital information in the plaintext under some conditions. Furthermore, it is also applied to a State-of-the Art CS-based encryption scheme for smart grid, and the power profile is reconstructed using ciphertext-only attack. Additionally, the statistical analysis of Gaussian and Binomial measurements is conducted to investigate the randomness provided by them.

Mobile edge computing (MEC) is one of the promising solutions to process computational-intensive tasks within short latency for emerging Internet-of-Things (IoT) use cases, e.g., virtual reality (VR), augmented reality (AR), autonomous vehicle. Due to the coexistence of heterogeneous services in MEC system, the task arrival interval and required execution time can vary depending on services. It is challenging to schedule computation resource for the services with stochastic arrivals and runtime at an edge server (ES). In this paper, we propose a flexible computation offloading framework among users and ESs. Based on the framework, we propose a Lyapunov-based algorithm to dynamically allocate computation resource for heterogeneous time-critical services at the ES. The proposed algorithm minimizes the average timeout probability without any prior knowledge on task arrival process and required runtime. The numerical results show that, compared with the standard queuing models used at ES, the proposed algorithm achieves at least 35% reduction of the timeout probability, and approximated utilization efficiency of computation resource to non-cause queuing model under various scenarios.

Mobile edge computing (MEC) is one of the promising solutions to process computational-intensive tasks for the emerging time-critical Internet-of-Things (IoT) use cases, e.g., virtual reality (VR), augmented reality (AR), autonomous vehicle. The latency can be reduced further, when a task is partitioned and computed by multiple edge servers' (ESs) collaboration. However, the state-of-the-art work studies the MEC-enabled offloading based on a static framework, which partitions tasks at either the local user equipment (UE) or the primary ES. The dynamic selection between the two offloading schemes has not been well studied yet. In this paper, we investigate a dynamic offloading framework in a multi-user scenario. Each UE can decide who partitions a task according to the network status, e.g., channel quality and allocated computation resource. Based on the framework, we model the latency to complete a task, and formulate an optimization problem to minimize the average latency among UEs. The problem is solved by jointly optimizing task partitioning and the allocation of the communication and computation resources. The numerical results show that, compared with the static offloading schemes, the proposed algorithm achieves the lower latency in all tested scenarios. Moreover, both mathematical derivation and simulation illustrate that the wireless channel quality difference between a UE and different ESs can be used as an important criterion to determine the right scheme.

The emerging time-critical Internet-of-Things (IoT) use cases, e.g., augmented reality, virtual reality, autonomous vehicle, etc., involve computation-intensive tasks powered by artificial intelligence (AI) techniques. Due to the limited computation resource at IoT devices, it is challenging to fulfill the latency and reliability requirements. Offloading computation tasks using mobile-edge computing (MEC) are a promising solution. The service reliability in AI-powered time-critical services can be modeled by the transmission reliability, timeout probability, and inference accuracy. To improve service reliability, the state-of-the-art work emphasizes on transmission reliability and requires error-free transmission. We show that the AI-powered time-critical services can tolerate small image distortion and still remain the inference accuracy. Therefore, to improve service reliability, it is more important to minimize the timeout probability by shortening transmission latency than perfect error-free transmission. Motivated by this insight, we study the feasibility of user datagram protocol (UDP)-based offloading for such services in the MEC system. A prototype is developed and a series of experiments are conducted to understand how image distortion affects inference accuracy. We measure the latency and transmission reliability of transmission control protocol (TCP)-based and UDP-based offloading in real-life environments and evaluate the service reliability both experimentally and numerically. The evaluation results demonstrate that compared with the TCP-based offloading, the UDP-based offloading can improve the normalized service reliability by up to 70% for time-critical services. In addition, we propose an early termination of image reception (ETR) offloading scheme which can further improve the normalized service reliability by up to 10%, compared with the baseline UDP-based scheme.

Fog computing leverages the computational resources at the network edge to meet the increasing demand for latency-sensitive applications in large-scale industries. In this article, we study the computation offloading in a fog computing network, where the end users, most of the time, offload part of their tasks to a fog node. Nevertheless, limited by the computational and storage resources, the fog node further simultaneously offloads the task data to the neighboring fog nodes and/or the remote cloud server to obtain the additional computing resources. However, meanwhile, the offloaded tasks from the neighboring node incur burden to the fog node. Moreover, the task offloading to the remote cloud server can suffer from limited communication resources. Thus, to jointly optimize the amount of tasks offloaded to the neighboring fog nodes and communication resource allocation for the offloaded tasks to the remote cloud, we formulate a latency-driven task data offloading problem considering the transmission delay from fog to the cloud and service rate that includes the local processing time and waiting time at each fog node. The optimization problem is formulated as a quadratically constraint quadratic programming. We solve the problem by semidefinite relaxation. The simulation results demonstrate that the proposed strategy is effective and scalable under various simulation settings.

The massive increase and assimilation of Internet of Things (IoT) devices and services imposes new challenges in sensing, communication, and reliable storage of data generated by the IoT. We focus on scenarios where the IoT devices may lose connectivity for long periods of time and can only rely on other IoT devices to store data reliably. This constitutes a problem of distributed storage where lost devices cannot be replaced by others in the network due to the fact that there are no additional devices arriving to the system and that each device has a limited storage capability. Thus, state-of-the-art approaches for distributed storage in data centers are not applicable. We show that optimal policies for data repair, in terms of bandwidth and storage usage, for this novel scenario can be implemented efficiently in real-devices using network coding. We provide a translation from the theoretical results in [1] into an implementation using Raspberry Pi devices.

To provide compressed storage for large amounts of time series data, we present a new strategy for data deduplication. Rather than attempting to deduplicate entire data chunks, we employ a generalized approach, where each chunk is split into a part worth deduplicating and a part that must be stored directly. This simple principle enables a greater compression of the often similar, non-identical, chunks of time series data than is the case for classic deduplication, while keeping benefits such as scalability, robustness, and on-the-fly storage, retrieval, and search for chunks. We analyze the method's theoretical performance, and argue that our method can asymptotically approach the entropy limit for some data configurations. To validate the method's practical merits, we finally show that it is competitive when compared to popular universal compression algorithms on the MIT-BIH ECG Compression Test Database.

We study a generalization of deduplication, which enables lossless deduplication of highly similar data and show that classic deduplication with fixed chunk length is a special case. We provide bounds on the expected length of coded sequences for generalized deduplication and show that the coding has asymptotic near-entropy cost under the proposed source model. More importantly, we show that generalized deduplication allows for multiple orders of magnitude faster convergence than classic deduplication. This means that generalized deduplication can provide compression benefits much earlier than classic deduplication, which is key in practical systems. Numerical examples demonstrate our results, showing that our lower bounds are achievable, and illustrating the potential gain of using the generalization over classic deduplication. In fact, we show that even for a simple case of generalized deduplication, the gain in convergence speed is linear with the size of the data chunks.

Computational offloading becomes an important and essential research issue for the delay-sensitive task completion at resource-constraint end-users. Fog computing that extends the computing and storage resources of the cloud computing to the network edge emerges as a potential solution towards low-latency task provisioning via computational offloading. In our offloading scenario, each end-user will first offload the task to its primary fog node. When the primary fog node cannot meet the tolerable latency, it has the possibility to offload to the cloud and/or assisting fog node to obtain extra computing resource to shorten the computing latency at the expense of additional transmission latency. Therefore, a trade-off needs to be carefully made in the offloading decision. At the same time, in addition to the task data from the end-users under its primary coverage, the primary fog node receives the tasks from other end-users via its neighbor fog nodes. Thus, to jointly optimize the computing and communication resources in the fog node, we formulate a delay-sensitive data offloading problem that mainly considers the local task execution delay and transmission delay. An approximate solution is obtained via Quadratically Constraint Quadratic Programming (QCQP). Finally, the extensive simulation results demonstrate the effectiveness of the proposed solution, while guaranteeing minimum end-to-end latency for various task processing densities and traffic intensity levels.

Non-orthogonal multiple access (NOMA) with compressed sensing based multiuser detection (CSMUD) is a promising candidate to enable massive connectivity for the massive machine type communication (mMTC) in 5G. In this paper, we have proposed three approaches to improve the performance and enable differentiation in the quality of service (QoS) among the mMTC devices in NOMA with sequence block CSMUD. First, a repeated-transmission scheme with spreading diversity is proposed which enhances the performance of all sensor nodes. Secondly, a QoS-aware sequence block allocation scheme is proposed to differentiate QoS in mMTC by assigning the low correlated sequence blocks to the high priority nodes. Lastly, the above two approaches are combined to propose a QoS-aware repeated-transmission scheme which not only increases the QoS differentiation but also achieves a significant gain in SNR of both high and low priority nodes. A QoS-aware group orthogonal matching pursuit based algorithm is designed to jointly detect and estimate the data. It is shown that the proposed repeated-transmission with spreading diversity scheme reduces the detection error rate (DER) and the bit error rate (BER) by more than two order of magnitude at SNR = 10 dB. Using the QoS-aware sequence block allocation scheme, it is shown that a gain of more than half a magnitude in DER and BER is achieved at SNR = 13 dB. The combined QoS-aware repeated-transmission scheme achieves a gain of at least 12 dB at BER = 10-4for the high priority nodes as well as a gain of at least 6 dB for the low priority nodes.

A variety of emerging use cases, e.g. virtual reality, autonomous vehicle, etc., require to reliably complete computation-intensive tasks within stringent latency. Mobile edge computing (MEC) is one of the promising solutions to provide computation capacity with low latency. However, it may cause extra communication cost to offload task to MEC via wireless channel. It is challenging to optimize the offloading decision considering the finite resource that is available at edge node for a user. In this paper, we study the code-partitioning offloading strategy where the computational task of the user is modeled by a directed acyclic graph. The reliability and latency of offloading are analyzed comprehensively. An optimization problem is formulated to minimize the offloading failure probability, subject to the latency constraint. Due to the non-convexity, we propose a heuristic algorithm to solve the problem with low complexity. The numerical results show that the proposed algorithm significantly improves the probability to reach the targeted reliability. The heuristic algorithm doubles the performance compared with the naive scheme, in particular, when the latency constraint is stringent. Furthermore, the proposed algorithm is applicable in various network conditions.

Compressed sensing based multiuser detection (CSMUD) is a promising candidate to cope with the massive connectivity requirements of the massive machine type communication (mMTC) in the fifth generation (5G) wireless communication system. It facilitates grant-free non-orthogonal code division multiple access (CDMA) to accommodate massive number of IoT devices. In non-orthogonal CDMA, the users are assigned with non-orthogonal sequences which serve as their signatures. However, the activity detection which is based on the correlation between the spreading sequences, degrades with increase in the number of users, especially in the lower SNR region. In this paper, to improve the performance of the CSMUD, we propose a sequence block based CSMUD, in which block of sequences is used as signature of the user instead of single sequence. A sequence block based group orthogonal matching pursuit algorithm is proposed to jointly detect the activity and data. The proposed scheme reduces the detection error rate (DER) by a magnitude of two at SNR = 10 dB in a system where the number of users are three times more than the number of available resources with each user having activity probability of 0.1. The DER of the proposed scheme is below 10-2even at activity probability of 0.16 for sequence length of 20 and overloading factor of 300%. Furthermore, at SNR = 10 dB, the DER of the proposed scheme outperforms the conventional scheme by a magnitude of one for a system with overloading factor of 500%.

Compressed sensing based multiuser detection is a recently developed scheme to cope with the massive machine-type communication in the fifth generation wireless system. Activity detection is one of the critical steps in compressed sensing based multiuser detection which determines the performance of system. In this paper we address this issue and propose two approaches for efficient detection and recovery of the signal. First, the diversity in the multiuser detection is increased by using a circularly right-shifted version of the sensing matrix along with the original sensing matrix for spreading the user data. Secondly, a deterministic sensing matrix based on Zadoff-Chu sequences is designed and accordingly a low complexity decoding algorithm is proposed which significantly reduces the detection error rate. It is shown that the proposed methods reduce the bit error rate by magnitude of one and two.

Sparse code multiple access (SCMA) is a nonorthogonal codebook based multiple access scheme, proposed to cope with the heterogeneous and challenging performance requirements for mission critical communication and massive machine-type communication in the fifth generation wireless system. In SCMA the codebooks are the main sources of performance improvement, which are generated based on the mother constellation. In this paper, a systematic sub-optimal approach of designing the mother constellation is proposed and a unified metric is derived to obtain the optimum codebooks using a specific mother constellation. It is demonstrated that the new mother constellation outperforms the existing mother constellations in terms of bit error rate (BER) performance. Moreover, it is shown that the optimum codebook obtained based on the unified metric can guarantee the optimum BER performance of the system.

Sparse code multiple access (SCMA) is a non- orthogonal codebook (CB) based multiple access scheme, proposed to cope with the heterogeneous and challenging performance requirements for mission critical communication and massive machine-type communication (MTC) in the fifth Generation (5G) wireless system. In this paper, the performance of SCMA has been studied and analyzed, considering the impact of the energy diversity and minimum Euclidean distance of the mother constellation, overloading of the system and the layer specific operators for codebooks generation. An optimized codebooks generation method for four ring star QAM based signaling constellation is proposed. It is demonstrated that by selecting the optimum design parameters, the bit error rate (BER) can be improved. Moreover, an overloading technique is also proposed to enable higher connectivity at lower decoding complexity.

Compression of electrocardiograms (ECG) in wireless environments, with diagnostic quality, has shown limited potential. This lack of quality preservation, using Wavelet Transform (WT), is due to the fact that the multiple levels of detail that can be achieved in the time domain are not exploited. In the present work, we propose to fully exploit the wavelet capability to operate at different levels of signal detail at different time scales. WT with an appropriate Compressed Sensing (CS) matrix is used in the electrode nodes of body sensor networks to encode and compress the ECG. Then, the signal is reconstructed using a basis pursuit denoise algorithm. Preservation of the diagnostic quality by means of standardized metrics is then tested for multiple wavelet bases and levels. High quality ECGs from 50 healthy patients are used to statistically show that diagnostic quality preservation is possible even at high compression rates. In these cases suitable ECG wavelets are required.

Joint application of polar channel coding combined with vector quantization lossy source coding is considered in this paper. The existing index assignment schemes in the literature cannot be used with polar codes due to their unique crossover probabilities. We elaborate on this problem and locally optimize index assignments. In addition, we propose an algorithm that jointly optimizes the number of quantization levels and the rate of the polar code in order to achieve minimum end-to-end distortion. It finds the optimal tradeoff between the distortion caused by channel errors and the quantization distortion. We also derive estimates for the crossover probabilities of the polar code which are required in the analysis. Simulation results confirm the effectiveness of the proposed algorithms and the accuracy of the crossover probabilities.

In two-phase cooperative multicast communications, the unbalanced outage probabilities of the cell-center and cell-edge users are the main reason that caps the energy-efficiency of the system. In this paper, we consider the unbalanced outage probability and propose a probability-based relay selection and power control method to improve the energy-efficiency, in which each user can autonomously decide whether to participate in the relay transmission. In particular, we obtain the optimal solution that can minimize the user power consumption. In addition, since our method works in a distributed manner, it does not require any feedback either between the BS and users, or among the users. This saves the extra energy consumption caused by the feedback. Simulation results demonstrate that the proposed method can reduce the user energy consumption up to 54%.

Random linear coding (RLC) can improve the performance of multicast transmissions in terms of throughput and energy efficiency. However, RLC and linear codes in general cannot necessarily attain the optimal performance in arbitrary networks. In this regard, partial packet recovery can be considered as a nonlinear strategy to complement such approaches for more general networks. In this paper, we propose a partial packet recovery scheme that benefits from the sparsity of bit errors in partially corrupted RLC packets. As opposed to many previous schemes, it performs without introducing preliminary checksums or preambles, demanding physical layer soft information, or requesting post-redundancy from the transmitter. It relies only on algebraic coding and data processing techniques, the existing knowledge at the receiver, and the conventional acknowledgment messages in RLC. By reconstructing and utilizing the partially corrupted packets that are usually discarded, it can reduce the average number of transmitted RLC packets required for successful decoding by typically 50%, which improves throughput and energy efficiency at the transmitter. We formulate our partial packet recovery in the form of a sparse recovery problem, present its different solutions using compressive sensing theory, discuss their complexity, and present and evaluate a Markov chain model for its performance.

In this paper, we propose a probability-based relay selection and power control method in two-phase cooperative multicast communications to minimize the user power consumption for any given multicast data rate. Our method gives each user the ability to autonomously decide whether to participate in the relay transmission based on an active probability. In particular, we develop an optimal algorithm to calculate the optimal active probability and relay power by balancing the outage reduction efficiency. We also develop a sub-optimal algorithm by balancing the outage probability of the cell-center and the cell-edge users. Simulation results demonstrate that the proposed optimal and sub-optimal algorithms can reduce the user power consumption up to about 54% and 40%, respectively, compared with the conventional algorithm that requires all users to participate in the relay transmission.

To reproduce a band-limited continuous-time signal with optimal fidelity, usually it is sampled at the Nyquist sampling rate and then the sample values are quantized. In nonuniform sampling, the total number of samples are reduced in expense of adding some reconstruction complexity and assuming prior information about the signal. In this paper we propose nonuniform sampling-time with nonuniform bit resolution per sample (bit-depth) to reduce the total required bit budget (i.e. the number of samples timed the bit-depth) even further. This idea is based on the fact that the maximum local variation of band-limited signals is bounded in a given time horizon. Therefore, it is not necessary to allocate a fixed bit-depth proportional to the signal's dynamic range to each sample. Instead, we try to allocate adaptively only the needed number of bits to represent the next sample considering the physical characteristics of the signal. Both sampling and reconstruction entities can generate the next sampling-times and bit-depths locally by observing the current and previous samples only. We propose different techniques in our generalized sampling framework that share a common sampling architecture. Here we only consider ECG signals for evaluation purpose. Based on the simulation results, the total number of required bits to reconstruct the signal with negligible distortion can be reduced up to 88% without imposing any form of transform coding compression.

We propose a novel scheme based on compressive sensing and sparse recovery to boost the performance of cross-packet random linear coding (RLC) by incorporating the partial packets in the decoding algorithm. In conventional RLC schemes, to successfully decode the packets the receiver needs to collect a certain number of correct innovative encoded packets. During this process, there are usually a lot of partially correct packets that are discarded. Our objective is to recover the errors in the partial packets to decrease the total transmitted packets to improve the performance in terms of throughput and energy efficiency. Assuming a systematic RLC, we first formulate this problem in form of a standard sparse recovery problem where the channel errors are sparsely distributed within the packets. Then we show that to correct a certain number of errors at the receiver, the minimum required number of transmitted packets is lower-bounded by the number of partial packets. We show that by correcting and exploiting the partial packets, the required number of RLC transmit packets to successfully deliver a given generation is reduced by typically 57% in comparison with the conventional scheme.

A novel joint source-channel (JSC) coding scheme that combines transform source coding with polar channel codes is proposed with performance gains at low SNRs in terms of bandwidth and energy efficiency. Binary wavelet transform source coding is first applied to decompose the binary source data into a set of components with uneven distributions, particularly, components that are zero with high probability. Normally, a polar code of rate K/N appends N − K redundant bits to the message which can decrease the throughput and introduce an extra energy consumption. By exploiting the binary wavelet transform, the proposed scheme transforms the original signal to the polar code without adding redundant zeros. Hence, the resulting JSC code is of rate one. Since no floating-point arithmetic is required and all operations are performed in GF(2), complexity is vastly reduced which can eventually alleviate the total hardware cost as well as the energy consumption.

A precise prediction of domestic demand is very important for establishing home energy management system and preventing the damage caused by overloading. In this work, active and reactive power consumption prediction model based on historical power usage data and external environment parameter data (temperature and solar radiation) is presented for a typical Southern Norwegian house. In the presented model, a neural network is adopted as a main prediction technique and historical domestic load data of around 2 years are utilized for training and testing purpose. Temperature and global irradiation (which illustrates the solar radiation level quantitatively) are employed as external parameters. From the results, the efficiency of predictions are evaluated and compared. It can be observed from the numerical results that predictions using historical power data together with external data perform better than the case where only power usage data are adopted.

In this paper we address the problem of rate scheduling in the Impulse Radio (IR) ultra-wideband (UWB) wireless body area networks (WBANs) and the minimum energy required to stabilize the queuing system. Targeting low complexity WBAN applications, we assume noncoherent receivers based on energy detection and autocorrelation for all nodes. The coordinating node can minimize the average energy consumption of the system and achieve the queue backlog stability of the sensor nodes by controlling the number of pulses per symbol. We first illustrate the necessary and sufficient conditions of network stability for a multi-mode UWB system and then propose a feasible rate scheduling algorithm based on the Lyapunov optimization theory. The scheduling algorithm uses the instantaneous channel state information and the length of the local queue of all sensor nodes and can approach the optimal energy-delay tradeoff of the network. We apply our theoretical framework to the IR-UWB physical layer of the IEEE 802.15.6 standard and extract the optimal physical layer modes that can achieve the desired energy-delay tradeoff.

With the volume of mobile data traffic almost doubling each year, the mobile data offloading challenge has become ever relevant. This paper takes on a practical approach to offload the cellular network. A solution called Cool-SHARE is proposed and implemented on Android. It is an app for seamlessly sharing apps over short-range links with limited cellular control and can be extended for multimedia data sharing. Two types of sharing scenarios are defined, i.e., social sharing and opportunistic sharing. The challenges of sharing are identified and analysed. The discovery, connection, and data transfer delays are measured and analysed on a Samsung Galaxy S III using Bluetooth and Wi-Fi Direct. The challenges including seamless discovery and connection, authenticity and payment are solved. Furthermore, suggestions for reducing delays of the low-level Wi-Fi Direct implementation are proposed.

The wireless channels in Body Area Networks (BANs) have significant temporal variations due to body movements. Therefore there is a potential to save energy by exploiting adaptive schemes, such as transmit power control (PC) and link adaptation (LA). This paper investigates the energy saving efficiency of PC and LA in BANs. The theoretical bounds of energy saving efficiency of PC and LA, and the condition that LA outperforms PC are derived. Generally speaking, LA is more efficient in saving energy than PC in BANs. The energy saving efficiency and packet erasure rate (PER) of PC and LA are evaluated through simulations with a large measured dataset from a BAN testbed. The simulation results show that LA can save 80-85% more energy than PC during 80% time. The increased PER due to the adaptive schemes is below 1% during 80-95% time at the different links.

This work studies how to select optimal code parameters of Random Linear Network Coding (RLNC) in Wireless Sensor Networks (WSNs). With Rateless Deluge [?] the authors proposed to apply Network Coding (NC) for Over-the-Air Programming (OAP) in WSNs, and demonstrated that with NC a significant reduction in the number of transmitted packets can be achieved. However, NC introduces additional computations and potentially a non-negligible transmission overhead, both of which depend on the chosen coding parameters. Therefore it is necessary to consider the trade-off that these coding parameters present in order to obtain the lowest energy consumption per transmitted bit. This problem is analyzed and suitable coding parameters are determined for the popular Tmote Sky platform. Compared to the use of traditional RLNC, these parameters enable a reduction in the energy spent per bit which grows as the generation size grows. These results also indicate that the use of high field sizes could be problematic from an energy point of view due to the additional complexity.

To enhance network scalability and increase network lifetime in large-scale wireless sensor networks (WSNs), clustering has been recognized as an effective solution for hierarchical routing, topology control and data aggregation. Inspired by the collective behavior of flocks and schools, we propose a Bio-inspired self-organizing Low-Complexity Clustering (B-LCC) algorithm for large-scale dense WSNs. The B-LCC algorithm does not require sensor locations, time synchronization nor any priori knowledge of the network. It is completely distributed and can achieve a well-distributed cluster heads. The processing time complexity of the B-LCC algorithm is O(1) per cluster, which outperforms most of the existing clustering algorithms as they have processing time complexity of O(n) per node in the worst case. Additionally, the B-LCC algorithm has a stable performance in topology control and the formed topology is robust to node failure.

This paper investigates the problem of efficient data dissemination in Mobile Ad hoc NETworks (MANETs) with high mobility. A testbed is presented; which provides a high degree of mobility in experiments. The testbed consists of 10 autonomous robots with mobile phones mounted on them. The mobile phones form an IEEE 802.11g ad hoc network to communicate with each other. A dynamic network topology is assumed, where the mobile devices form a cooperative cluster in order to exchange data packets among each other. In our multimedia exchange scenario, the initial state is that one device carries all information, and the goal is to convey that information to all devices. A strategy is proposed that uses UDP broadcast transmissions and random linear network coding to facilitate the efficient exchange of information in the network. An application is introduced that implements this strategy on Nokia phones. The measurement results collected from the testbed are presented, and the performance of the proposed strategy is compared to a reference strategy that is based on TCP unicast connections.

To support reliable real-time applications in Body Sensor Networks (BSNs), it is necessary to develop an efficient and robust routing protocol. However, it is challenging due to the specific radio propagation characteristics, dynamic network topology, variable link quality caused by body movements or environments, extremely low transmission power and limited battery among many others. In this paper, we propose Reactive Virtual Coordinate Routing protocol (RVCR) to tackle the challenges in BSNs by making a good use of the specific radio propagation and link characteristics on the human body. The basic idea of RVCR is that it efficiently exploits the temporary high-quality link opportunities resulting from the body movements and the environments to effectively forward packets towards the destination. RVCR achieves a stable packet delivery ratio (PDR) of more than 99% independent of the body movements and the environments at the lowest possible transmission power. Besides high PDR, RVCR has a stable and small average number of hops per packet, which makes it energy efficient. Additionally, RVCR can keep the same performances even in case of severe link failures.

In group-oriented applications for wireless networks, reliable multicast strategies are important in order to efficiently distribute data, e.g. in Wireless Mesh Networks (WMNs) and Mobile Ad-hoc NETworks (MANETs). To ensure that developed protocols and systems will operate as expected when deployed in the wild, a good understanding of several factors such as packet loss characteristics is necessary. In this paper the correlation of erasures in a cluster of receiving mobile devices is measured and analyzed. In the considered scenario, a source node broadcasts packets to a cluster of receivers located relatively far away. To ensure that the obtained data can easily be applied in analysis, we introduce the \textit{cluster erasure transition matrix}. We then analyze a simple broadcast and cooperative scheme, and show that the assumption of independent packet erasures unfairly favors the cooperative scheme according to the obtained measurements.

In this paper user cooperation with network coding is applied to MBMS (Multimedia broadcast/multicast service) where Raptor codes are currently used. User cooperation together with network coding is used to save bandwidth and improve user perceived QoS for broadcast/multicast services in the future mobile communication networks. The proposed approach is tailored to LTE networks and extended with local cooperation. The simulation results show that local retransmissions can reduce the amount of redundant information on the cellular link with up to 80% as long as there are at least two cooperative mobile devices. The results also show that network coding can save more than half of the traffic on the short-range link as long as there are four devices in the cooperation cluster.

This paper investigated the possibility of intrinsic information conveying in network coding systems. The information is embedded into the coding vector by constructing the vector based on a set of predefined rules. This information can subsequently be retrieved by any receiver. The starting point is Random Linear Network Coding (RLNC) and the goal is to reduce the amount of coding operations both at the coding and decoding node, and at the same time remove the need for dedicated signaling messages. In a traditional RLNC system, coding operation takes up significant computational resources and adds to the overall energy consumption, which is particular problematic for mobile battery-driven devices. In RLNC coding is performed over a Finite Field. We propose to divide this field into sub fields, and let each sub field signify some information or state. In order to embed the information correctly the coding operations must be performed in a particular way, which we introduce. Finally we evaluate the suggested system and find that the amount of coding can be significantly reduced both at nodes that recode and decode.

In this paper we present our WiMAX (worldwide interoperability for microwave access) network performance monitoring and optimization solution. As a new and small WiMAX network operator, there are many demanding issues that we have to deal with, such as limited available frequency resource, tight frequency reuse, capacity planning, proper network dimensioning, multi-class data services and so on. Furthermore, as a small operator we also want to reduce the demand for sophisticated technicians and man labour hours. To meet these critical demands, we design a generic integrated network performance monitoring and optimization solution which includes traffic monitor and analyzer, dynamic quality of service control tool, radio signal quality monitor, CPE mobility tracking, interference monitor & analyzer and optimization tool, report generator and alarm management tool. We develop and implement this integrated network performance monitoring and optimization system in our WiMAX networks. This integrated monitoring and optimization system has such good flexibility and scalability that individual function component can be used by other operators with special needs and more advanced function components can be developed in the future. The usage of system has proven that it can reduce both CAPEX (capital expense) and OPEX (operational expense), meanwhile it can improve end-to-end quality of service.

To solve the performance degradation issue in current WLAN caused by the crowded unlicensed spectrum, we propose a cognitive radio (CR) media access protocol, C-CSMA/CA. The basic idea is that with cognitive radio techniques the WLAN devices can not only access the legacy WLAN unlicensed spectrum but opportunistically access any other under-utilized licensed spectrum without a license. The application scenario of C-CSMA/CA is infrastructure BSS (Basic Service Set) WLAN. C-CSMA/CA efficiently exploits the inherent characteristics of CSMA/CA to design distributed cooperative outband sensing to explore spectrum hole; moreover, it designs dual inband sensing scheme to detect primary user appearance. Additionally, C-CSMA/CA has the advantage to effectively solve the cognitive radio self-coexistence issues in the overlapping CR BSSs scenario. It also realizes station-based dynamic resource selection and utilization. It is compatible with any legacy WLAN (BSS) system. We develop and implement the simulation of C-CSMA/CA by OPNET. The simulation results show that C-CSMA/CA highly enhances throughput and reduces the queuing delay and media access delay.

In this paper we propose the one4all cooperative access strategy to introduce a more efficient media access strategy for wireless networks. The one4all scheme is designed for the infrastructure based distributed wireless network architecture. The basic idea is that mobile devices can form a cooperative cluster using their short-range air interface and one device contends the channel for all the devices within the cluster. This strategy reduces the number of mobile devices involved in the collision process for the wireless medium resulting in larger throughput, smaller access delay, and less energy consumption. Based on an analytical model, the proposed strategy is compared with the two existing strategies RTS/CTS (request to send/ clear to send) and packet aggregation. The results show that the proposed cooperative scheme has similar throughput performance as packet aggregation and it has much higher throughput than the conventional RTS/CTS scheme. Furthermore, the newly introduced cooperative scheme outperforms packet aggregation in terms of channel access delay and energy consumption.

Packet reservation multiple access (PRMA) protocol is an implicit reservation MAC protocol. It is initially designed for voice packets in the cellular networks, but it is currently also used for data packets in OFDM based fixed wireless access networks. When it is applied for data packets, the system throughput depends on the size of packets and the number of consecutive packets. From the statistics of existent wireless data networks using PRMA protocol, it shows that the system throughput is quite low because of the inconsecutive small packets. In order to improve the throughput, packet aggregation scheme is considered to be applied in PRMA. Before designing packet aggregation algorithm, it is worth investigating the effect of packet aggregation scheme on the performance of throughput and delay. In this paper we develop a generic Markov chain model for PRMA with packet aggregation. Based on this model the throughput and delay are derived and analyzed. A numerical example is calculated, which illustrates the effect of packet aggregation on the throughput and delay with varying packet arrival rate. The results of the paper are valuable inputs for designing optimal packet aggregation algorithm, considering the tradeoff between throughput and delay.

In this paper we advocate to exploit terminal cooperation to stabilize IP communication using header compression. The terminal cooperation is based on direct communication between terminals using short range communication and simultaneously being connected to the cellular service access point. The short range link is than used to provide first aid information to heal the decompressor state of the neighboring node in case of a packet loss on the cellular link. IP header compression schemes are used to increase the spectral and power efficiency loosing robustness of the communication compared to the uncompressed version. By introducing the terminal cooperation supporting header compression the robustness is increased. Within this article we will show that header compression should be applied to reduce the energy consumption of the terminals and moreover the header compression should be supported by cooperation to increase the robustness in terms of a decreased packet loss rate.

Reliable multicast applications such as software distribution, data distribution and replication and mailing list delivery, etc. [1] are getting more and more interests from network and service providers. The conventional error/loss recovery schemes are not efficient when they are applied to multicast scenarios in wireless networks. The reason lies in the unreliable wireless channel, the limited wireless bandwidth and resource, the battery powered wireless devices, and others. To have an effective error/loss recovery scheme for reliable multicast in wireless networks, we advocate a new communication architecture. It is referred to as cellular controlled peer-to- peer network, where the mobile devices communicate directly with each other to perform cooperative retransmissions using their short-range communication capabilities in addition to their cellular links. Based on the cooperative architecture a novel retransmission scheme is proposed exploiting the short-range retransmission in this paper. The state of the art, the non-cooperative error recovery schemes (e.g., ARQ, Layered FEC and Integrated FEC II) and the proposed scheme are compared with each other in terms of energy consumption to show the benefit of the newly introduced scheme.

Packet reservation multiple access (PRMA) protocol is an implicit reservation MAC protocol. It is initially designed for voice packets in the cellular networks [2, 3] but it is currently also used for data packets in OFDM based fixed wireless access networks [8, 9], When it is applied for data packets, the system throughput depends on the size of packets and the number of consecutive packets. From the statistics of existent wireless data networks using PRMA protocol, it shows that the system throughput is quite low because of the inconsecutive small packets. In order to improve the throughput, packet aggregation scheme is considered to be applied in PRMA. Before designing packet aggregation algorithm, it is worth investigating the effect of packet aggregation scheme on the performance of throughput and delay. In this paper we develop a generic Markov chain model for PRMA with packet aggregation. Based on this model the throughput and delay are derived and analyzed. A numerical example is calculated, which illustrates the effect of packet aggregation on the throughput and delay with varying packet arrival rate. The results of the paper are valuable inputs for designing optimal packet aggregation algorithm, considering the tradeoff between throughput and delay.

This paper introduces power saving strategies for cooperative wireless communication systems. The described scenario focuses on IP-services over DVB-H networks showing the strength of non-altruistic cooperation between mobile devices. The envisioned cooperation is based on cellular reception of data, which is then shared among mobile devices within each others' proximity over short-range links. As the state-of-the-art, we use Bluetooth technology for the short-range link communication in this cooperative scheme. In this paper, three topology based cooperative algorithms for the short-range link communication are designed. Then numerical results show that a power saving gain of over 50% can be achieved by cooperative networking of three mobile terminals in fully cooperating mode.