Edge-based visual simultaneous localization and mapping (Visual-SLAM) enables real-time indoor localization of robots by using substantial computing resources of an edge server. However, this system can place a burden on resource-constrained wireless networks because robots transmit their video data to the edge server via wireless networks. Given the sharing of the limited network capacity among multiple devices, the video bitrate needs to be reduced so as not to overshoot the available bandwidth. Although lowering the video quality can reduce the video bitrate, it also results in a loss of detailed image features and can negatively impact localization accuracy. In this paper, we propose a Vision Transformer-based patch quality controller (ViT-PQC) that reduces the video bitrate below the available bandwidth while preserving localization accuracy. ViT-PQC segments the video frame into patches and assigns an optimal image quality to each patch by taking into account both the image features of the recent frame and time shift features between the two most recent frames. Evaluation results demonstrate that ViT-PQC reduces the number of bitrate overshoots by 94% while preserving localization accuracy.

Wireless multi-connectivity solutions are essential for reliable low-latency wireless communication services enabling delay-sensitive applications such as industrial robotics. However, non-expert industry vertical players in networking seek low-installation-cost and high-quality solutions to use wireless multi-connectivity. This paper proposes a wireless multi-connectivity management gateway (WMC-GW) to effectively utilize multiple wireless networks without modifying applications and wireless network systems. We install a WMC-GW at each mobile robot and another WMC-GW at the application server. There are two features. The first is real-time radio access technology (RAT) selection, where the WMC-GW select an appropriate RAT based on predicted delay trends using IP packet-level delay gradients to follow sensitive delay variations. The second is the flexibility of flow-level policy control, where we develop multiple RAT selection policies based on the delay gradients. Through performance evaluation, our approach effectively works in low-latency RAT selection.

Communication traffic patterns, representing network and application behavior, are beneficial for the transport and application layers to estimate the states of black-box network systems. Masking noisy traffic features helps high-quality communication by reducing the misestimation of network states due to unstable delay behavior for delay-sensitive applications. Motivated by the effects, we propose TrafficArranger, a traffic pattern re-arrangement system using packet scheduling to provide a target quality of service (QoS), including throughput, delay, and packet interval, as required by each flow. The main component of TrafficArranger is Hierarchical Calendar Queueing (HCQ) for delay-based packet scheduling. We introduce a number of techniques: (i) stepped dequeue to control the dequeue level for flexible jitter control for improving robustness to traffic load while keeping control accuracy and (ii) rush dequeue and reenqueue skipping to address out-of-order issues in HCQ. Through performance comparison with some leaves of Linux qdisc and state-of-the-art HCQ method, i.e., Gearbox, only TrafficArranger controls QoS as required for all QoS classes (i.e., throughput, delay, and packet interval). Also, we demonstrate that TrafficArranger effectively works for performance stabilization in a cooperative adaptive cruise control system.

Accurately estimating congestion states of the bottleneck link in mobile networks (e.g., LTE and 5G) from round-trip times (RTTs) is an important task for congestion control algorithms (CCAs). In mobile networks, RTTs fluctuate easily because they are sensitive to stochastic behavior, e.g., congestion and radio quality, and deterministic behavior, e.g., scheduling, making it difficult to estimate the congestion state accurately. In this paper, we propose an RTT-gradient analysis method for accurately estimating the congestion state (i.e., overuse/normal/underuse) by filtering stochastic noise while reducing misestimation caused by deterministic delay behavior in mobile networks. The proposed method consists of two features. The first is gradual RTT smoothing, a two-stage Kalman filter designed in series for filtering noisy RTT variations. The second is adaptive threshold clipping, which eliminates estimator instability caused by deterministic delay variations in the mobile network. Our experimental results in a commercial 5G network show that our method is more robust than a conventional method (the state estimator of Google Congestion Control (GCC)). Furthermore, simulation results using an open dataset for mobile networks show that our method improves throughput by 1.6% on average for 13 scenarios out of 16 scenarios total from GCC.

Private mobile networks, such as local 5G, have attracted the attention of industry players who expect flexible radio resource allocation by methods such as time-division duplex (TDD) scheduling based on the uplink and downlink traffic demand of their solutions. However, there are two challenges when communicating using TCP congestion control algorithms (CCAs) over the TDD link: TDD-induced ACK-waiting time and misestimating congestion states due to deterministic delay variation caused by TDD scheduling. In this paper, we propose a TDD-aware TCP pacing method for improving TCP throughput by pacing the sending time between two consecutive segments within the ACK-waiting time. We determine the pacing rate on the basis of the TDD-induced delay variation for sending TCP segments within allocated TDD slots while reducing round-trip time (RTT). We evaluate the performance of our method by using a network simulator (ns-3). TCP pacing improves throughput by about 10–70% compared with when there is no pacing, especially for TCP Illinois. We also verify that our TDD-aware pacing improves throughput by about 10% compared to the default pacing rate on the Linux kernel.

This paper proposes a system to improve the visibility of real-time Light-Detection-and-Ranging (LIDAR) data by enabling calibration with pre-created three-dimensional (3D) images. The proposed system automatically selects suitable frames to periodically maintain calibration, and we demonstrate its effectiveness through experiments.

Understanding packet-level communication delay behavior in mobile networks is becoming increasingly important with the rise of delay-sensitive applications such as industrial mobile robots. Although attention has traditionally focused on queueing delays due to congestion, in mobile networks, variable communication delays due to multiple delay factors degrade the performance of delay-sensitive applications that send packets with a high packet rate. This paper contributes to identifying the relationship between packet rate and delay components. We first categorize delays in end-to-end communication into four categories: transmission, propagation, queueing, and processing delays. We then formulate the relationship between the packet transmission interval and the delay factors, which shows an interesting trend that the ratio of delay components differs with transmission interval time. Also, when the packet transmission interval is short, the impact of deterministic delay jitter due to processing delay is more significant than that of stochastic queueing delay. We examine the trend of delay component ratio through experiments in an operational 5G network in Japan.

This demonstration proposes an edge system that enables sensing and computing capacities to be offloaded from vehicles to the system. We developed a prototype of the proposed system, in which a sensor network using multiple light-detection-and-ranging (LIDAR) units works for sensing and computing, while autonomous vehicles are equipped only with lightweight sensing and computation capabilities. This demon-stration presents the performance of the prototype in terms of movement accuracy to verify the feasibility of the proposed system.

Autonomous driving is attracting attention. Though related technologies are evolving, it is forecasted that Level 5 autonomous driving (full driving automation) would be ordinary in the 2040s to 2060s. Until the epoch, remote human support is necessary to deal with the cases in which an autonomous car cannot decide by itself. For the remote human support, live video streaming and user interface for effective monitoring are significant challenges. This paper proposes a live video streaming method that minimizes video bitrate while keeping the required video quality (video analytics accuracy) and AI-assisted monitoring GUI (graphical user interface) to monitor multiple cars effectively. First, the authors evaluated the GUI through simulation. The result shows that it improves efficiency and reduces mental and physical loads. Then, the authors conducted field tests in which an operator monitored two cars. The operators said that the proposed system is much better than the conventional systems they usually use and can monitor and control at least two cars simultaneously employing the system.

Next generation wireless networks will require the flexibility to accommodate an extremely diverse set of service types. Increasing emphasis on quality of service (QoS) and limited radio necessitates the use of mixed-numerologies to accommodate diverse service requirements. In this paper, we present a mixed-numerology profile selection method that adapts to the context of wireless networks to maximize the QoS of end users. The idea is to take service requirements, channel conditions and traffic patterns into account simultaneously for optimizing a mixed-numerology profile. We extract statistical features from these aspects to train a Mondrian forest model, designed to estimate the QoS in different mixed-numerology profiles. This enables the optimization of mixed-numerologies under any wireless communication scenario, with specific service requirements, and under any traffic scenario. Comprehensive simulations were conducted to evaluate the performance of the proposed mixed-numerology method. Results show that the proposed method is able to provide QoS satisfaction levels of 80–95%, whereas a non-optimized approach provides 60–85%, while minimizing the total number of numerology indexes in operation.

Understanding the delay jitter of mobile communications becomes important because of widely spreading delay-sensitive applications such as remote control of mobile robots with high-frequency communications via wireless networks. Prior studies on delay jitter modeling have proposed using a single probability distribution (e.g., Gamma and Laplace distributions). However, mobility-induced wireless quality fluctuations form a mixture of probability patterns, e.g., several peaks and a heavy tail. This paper proposes a method to estimate delay jitter accurately in high-frequency and mobile communications. Our method has two features. The first is to model the delay jitter by a mixture of multiple Laplace distributions by taking into account the probability patterns. For quick convergence of model training, the model is trained with access manner-aware initialization in each Wi-Fi and mobile network. The second is to construct a likelihood-based observation segmentation for estimating model parameters accurately against mobility. Performance evaluation through experiments in indoor Wi-Fi and outdoor 5G scenarios shows that our proposed method improves modeling accuracy by 28.7% compared with the case of the prior studies.

Understanding communication delay in mobile networks is becoming more important as delay-sensitive scenarios become more prevalent. Round-trip time measurement, the conventional technique to measure communication delay, e.g., ping, outputs network-induced delay for each packet but is insufficient in identifying specific delay factors. We propose a model called Delay Component Model (DCM) to aid in clearly visualizing communication delay. We construct the DCM on the basis of our measurement study on packet-receipt intervals with packet transmission at a constant interval via commercial mobile networks in Japan. We find that the measured receipt-time intervals form striped patterns due to the combination of two components: a constant scheduler (ConstSched) and probability scheduler (ProbSched). We use the principle of forming striped patterns and develop a method of estimating the DCM structure. Finally, we evaluate our method by analyzing delay patterns measured in Long Term Evolution (LTE) and fifth-generation mobile (5G) networks. The results indicate that delay patterns in these networks are due to the combination of three components, i.e., a ConstSched with 20-ms intervals, ProbSched with 8-ms delay for LTE and 6-ms delay for 5G, and ProbSched with 1-ms delay.

Cloud offloading enables smaller edge devices which contain fewer computing resources to be used for computer vision. This contributes to mobile applications of computer vision such as robotics and augmented reality. However, cloud offloading is impacted by bandwidth fluctuations in wireless networks. When the available bandwidth is restricted, it is difficult to offload workloads (i.e. frame) to cloud instances. Instead of offloading workloads to the cloud, existing approaches send a result of edge prediction or use a past cloud prediction to cover non-offloaded frames, which result in low accuracy. In this paper, we propose an Edge Cloud Ensemble method for object detection to improve accuracy in low bandwidth environments. In our method, the edge sends an inaccurate prediction and a motion vector of a detected object’s region to the cloud while maintaining low transmission overhead. The cloud corrects the inaccurate prediction by using the motion vectors to shift a past, accurate cloud prediction. The results of our experiments demonstrate that our approach can improve accuracy in low bandwidth environments compared with existing methods, especially in moving cameras.

Understanding the required quality of service (QoS) for guaranteeing the necessary video quality for Internet video streaming services is important for enabling network operators to provide high-quality networks. Recent trace-based approaches that infer video quality through machine-learning based traffic-feature analysis have a drawback in terms of the maintenance cost for updating their analysis model due to the massive number of videos uploaded daily. In this paper, we construct a QoS analysis model that calculates the required throughput for delivering video with a certain resolution by using encoding information instead of traffic tracing. For accurate modeling, we consider and formulate communication overhead, i.e. streaming behavior, retransmission, and headers. Through experiments, we demonstrate that our model can perform in two use cases: (1) estimating effective resolution for arbitrary QoS, and (2) calculating the required QoS for guaranteeing the necessary video quality.

Traffic reduction for bandwidth-hungry video streaming services, such as YouTube, benefits not only subscribers struggling to avoid going over their contracted data limit, but also service providers when the number of people who use video streaming services increase. Because not all stakeholders who want to reduce traffic usage are willing to conduct cumbersome operations, e.g., manually setting lower resolution, we argue here that network operators should introduce a traffic pacer for providing traffic reduction services as an optional plan for subscribers. This paper proposes NetPacer, an in-network traffic pacing system for reducing traffic usage by degrading the video quality. NetPacer has two features. The first is relative pacing, which degrades the video quality relative to the initial quality by traffic shaping, thus enabling flexible quality control. The second is in-network timely video quality identification via encrypted traffic analysis by using machine learning. Through experiments, we demonstrate that NetPacer successfully reduces traffic by 30.8% by degrading the resolution by one level while keeping the QoE (i.e., Mean Opinion Score (MOS)) degradation below 0.268 points on average for 50 YouTube videos.

Zero-rating service provided by Mobile Virtual Network Operators (MVNOs) has been attracting smartphone users who frequently watch web videos that are delivered by heavily bandwidth-consuming applications. With the increase of encrypted traffic, MVNOs need to identify video hosting sites accessed by smartphone users via encrypted traffic analysis for enabling such services. If traffic from permitted sites is identified as coming from non-permitted sites due to mistaken identification of video hosting sites, unreasonable payments are inevitable for MVNOs or subscribers, and vice versa. In this paper we propose two feature sets considering multiple flow transmission and analyze the feature sets by supervised machine learning for identifying video hosting sites accurately. The first set is traffic features extracted from flows for only transmitting video contents. The second is 4-tuple distribution of established flows for transmitting various contents in a single video web page. These feature sets are based on our investigation of the characteristics of four of the most popular video hosting sites in Japan. Through video hosting site identification experiments, the identification accuracy of single flow analysis reaches 85.9%, and the accuracy of the proposed method reaches 92.0%.

Private mobile networks such as private LTE/local 5G, which support flexibly configured and empowered innovative technologies that are not feasible in closed public mobile networks recently, have been catching much attention both in academia and in industries. In this paper, we design and implement the automatic check-in service as an example of value-added services of private mobile networks utilizing the flexibility of softwarization. To alleviate the inherent coverage problem of a private mobile network, we integrate our private mobile network with a public LTE by sharing the subscriber database so that a user can use the automatic check-in services deployed in various private mobile networks with only one SIM issued by a public network. We perform field tests in a private mobile network and also a private-public hybrid mobile network and disclose that the users' check-out behavior is predictable through numerical analyses. Based on the finding, we introduce two machine learning-based inference mechanisms that can predict a user's check-out behavior at an inference accuracy of 83% and 93% in a private network and a hybrid one separately. We believe this paper can provide valuable experience for those who are developing their private mobile networks.

Online distillation can dynamically adapt to changes of distribution in a target domain by continuously updating a smaller student model from a live video stream. This ensures the accuracy of the student model even if distribution changes occur due to a change of content. However, online distillation degrades the overall accuracy because it causes overfitting to “current” distribution not to “recent” distribution. The student model is trained on sequential incoming data, and its model parameters are overwritten with the “current” distribution. As a result, the student model forgets the “recent” distribution. To overcome this problem, we propose a new training framework using cache. Our framework temporarily stores incoming frames and teacher model outputs in a cache and trains a student model with data selected from the cache. Since our approach trains the student model with not only incoming data but also past data, it can improve the overall accuracy while adapting to changes of distribution without overfitting. To use limited cache size efficiently, we also propose a loss-aware cache algorithm that chooses training data prioritized by its loss value. Our experiments show that training with cache improves the accuracy compared with online distillation, and the loss-aware cache algorithm outperforms a cache algorithm modeled on traditional offline training.

Current video compression algorithms are designed to achieve a smaller data size and enable higher human perception for the real time streaming of video applications. However, with the recent explosive technical progress of deep neural network (DNNs), video surveillance is increasingly being performed not by humans but by computer vision systems. In this work, we propose a video compression method for object detection by computer vision algorithms. Our method detects the ROI in an image and differentiates the image quality between the ROI and other areas. It also optimizes the image quality in the ROI by estimating the recognition accuracy of the object detection model. The results of an experimental evaluation demonstrate that our proposed method can achieve high-quality encoding in terms of data size and successfully estimate the recognition accuracy of the object detection model.

We observe two major revolutionary trends in net-work operations: democratization of cost-effective and flexible communication means for vertical players, such as public safety, by private mobile networking combined with edge computing, and automatic and autonomic network operations empowered by Artificial Intelligence (AI). Further innovations are required for making private networking readily available for vertical players that are reluctant to acquire expertise in complex network operations. We propose Edge Concierge, of which concept is to democratize cost-effective and flexible network operations using network layer AI at private network edges. Edge Concierge assists smart network operations for private mobile network operators and energy saving by changing working state of AI-empowered anomaly detection applications by network layer AI. We also employ unsupervised machine learning using Hidden Markov Model (HMM) for estimating contexts by solely observing net-work traffic at mobile edge computing (MEC) middle boxes. In detail, we design a system of real-time and self-learning context estimation by a multi-level probabilistic state transition model trained by unsupervised learning, which is implemented in a commodity PC. In order to evaluate our proposed system, we take public safety context of smart cities as an example use case and show the benefits.

Automated guided vehicles (AGVs) have recently been introduced in the production and manufacturing industries. The efficiency, safety, and cost of AGV systems are issues to be addressed, and wireless communication will play an important role in addressing them. By using wireless communication, AGVs can avoid sudden breakings and collisions at intersections by sharing their positions beforehand even if each AGV does not have additional sensors. However, the lack of radio resources can cause transmission delay, which can cause AGV collisions or traffic jams. We propose a physical context-aware communication control method for collision-free and smooth AGV operation. The proposed method predicts possible AGV collisions and optimizes status report intervals for AGVs. The evaluation results indicate that a 1.5 times increase in work efficiency and 1.6 times larger AGV capacity can be achieved with the proposed method than with a conventional periodical reporting method.

The requirements of smartphone users have shifted from the quality of service (i.e., throughput) to the quality of experience. Also, the amount of encrypted traffic has increased to protect personal information. Therefore, to provide a quality mobile network experience for smartphone users, network operators need to identify applications from the encrypted traffic and control their traffic. In this paper, we propose a method of identifying applications running on a specific smartphone by analyzing only the time series patterns in IP traffic without inspecting the encrypted traffic. The proposed method estimates application flow with a two-level probabilistic state transition model and identifies applications on the basis of the statistics per estimated flow. Through experiments identifying applications running on a smartphone, we evaluated the estimation accuracy of proposed method.

Real-time prediction of spatial information has attracted a lot of attention. Machine learning enables us to provide real-time prediction of spatial information such as road traffic by using aggregated sensor data. The amount of mobile traffic is forecasted to increase exponentially, thereby causing serious transmission delays when traffic loads are heavy. If a part of the data used for predicting spatial information in real time does not arrive on time, the prediction accuracy degrades because the prediction is done without the missing data. A utility-based scheduling technique has been suggested as a way of prioritizing such delay-sensitive data. However, no study has not addressed the utility-based scheduling for the real-time prediction of spatial information. Therefore, this paper proposes a scheme that enables modeling the utility function for real- time prediction of spatial information. The scheme is roughly composed of two steps: the first creates training data from original time-series data and a machine learning model using the data, while the second models the utility function using the feature selection method in the learning model. Feature selection method enables extracting the importance of data in terms of how much the data contributes to the prediction accuracy. This paper assumes the road traffic prediction as a scenario and shows the utility function modeled by the proposed scheme using real spatial datasets. A numerical study demonstrates how the model of the utility function works effectively in prioritizing data for real-time prediction in terms of accuracy.

Machine Type Communications (MTC) has been paid much attention as a new communication paradigm to increase mobile network traffic. Most of MTC terminals are time-driven, that is, they send and receive data periodically. Therefore, network access requests on mobile core networks are concentrated at a specific timing, which results in instantaneous increase in network load. Considering the fact that such time-driven MTC would accept a certain amount of latency in their cyclic communication, in this paper, we propose a scheduling method of communication timings of time-driven MTC terminals to mitigate traffic concentration. We extend the standardized back-off mechanism of 3GPP to configure the back-off time length for each terminal to decrease the number of concurrent bearers in the network, while satisfying requirements on communication latency. We compare proposed methods by simulation experiments and reveal that we can achieve almost zero access rejections at reasonable communication quality by a simple timeslot selection algorithm when the core network maintain the timeslot assignment status for accommodated User Equipments. To the best of our knowledge, this is the first proposal to alleviate short-term congestion of mobile core networks by MTC with TDMA-like network control.