Latency Optimization Research Articles

Efficiency ratio optimization on uplink transmission power for Cloud-Based Radio Access Network

With the adoption of fifth generation (5G) wireless network systems into the fourth Industrial Revolution, the data transfer efficiency and speed will see a significantly improvement, which will lead to systems having the ability to provide a real-time information, and a solution to the capacity problem. Cloud-Based Radio Access Network (C-RAN) is the groundwork on which future 5G networks are based on, and it can increase the spectral efficiency and lower the transmission latency. Moreover, the used of C-RAN in future is expected to decrease the capital, operational and maintenance expenses of mobile operators. This study aims to investigate and validate the joint measures of spectral efficiency and latency optimization by using a quantization process of uplink signal at the remote radio head (RRH) of the C-RAN network. The quantization process is used to convert the received radio frequency signal from mobile user equipment into in-phase and quadrature baseband samples. The uplink transmission power and the number of quantization bits are computed in a way that the performance of spectral efficiency and latency of the RRHs are simultaneously optimized. Results shows that the proposed scheme is capable working with the different types of RRH.

AI-Assisted Low Information Latency Wireless Networking

The 5G Phase-2 and beyond wireless systems will focus more on vertical applications such as autonomous driving and industrial Internet-of-things, many of which are categorized as ultra-Reliable Low-Latency Communications (uRLLC). In this article, an alternative view on uRLLC is presented, that information latency, which measures the distortion of information resulted from time lag of its acquisition process, is more relevant than conventional communication latency of uRLLC in wireless networked control systems. An AI-assisted Situationally-aware Multi-Agent Reinforcement learning framework for wireless neTworks (SMART) is presented to address the information latency optimization challenge. Case studies of typical applications in Autonomous Driving (AD) are demonstrated, i.e., dense platooning and intersection management, which show that SMART can effectively optimize information latency, and more importantly, information latency-optimized systems outperform conventional uRLLC-oriented systems significantly in terms of AD performance such as traffic efficiency, thus pointing out a new research and system design paradigm.

Research on the optimization of IIoT data processing latency

As for data processing, data should be sent to cloud, stored and computed. Usually, the amount of data is huge enough for higher latency of processing. In addition, the mobility of cloud service would be much slower. In this paper, an Industrial Internet of Things cloud–fog hybrid network (ITCFN) framework is proposed to solve high latency upon processing industrial data on cloud. In the production equipment area, edge devices such as routers and switches are utilized by framework to construct a fog computing layer between cloud server and production equipment. Since computing power of the edge devices in fog computing is much poor, a distributed computing method for multi-devices is proposed. The aim of minimum task processing delay is achieved by using the constrained particle swarm optimization load balancing algorithm based on simulated annealing method (SAPSO-LB). The experimental results show that the ITCFN based on SAPSO-LB algorithm can effectively reduce the industrial data processing delay. When ten fog computing devices are used and the collected data is between 4GB and 12GB, compared with cloud computing, the latency is improved by 84.1%-29.9%.

360 Panorama Synthesis from a Sparse Set of Images on a Low-Power Device

A full 360 $^\circ$ × 180 $^\circ$ image provides an unlimited field of view (FOV) and an immersive experience for the users without any loss of information of the surrounding. In this study, a deep learning based approach is proposed to synthesize a 360 $^\circ$ image from a sparse set of images captured with a limited FOV. The proposed network consists of a cascade of the FOV estimation network and the panorama synthesis network. We propose a hierarchical generative network to synthesize high quality 360 $^\circ$ panorama images. The design of progressive multi-scale generator and multiple discriminator reduces the high frequency artifact which is commonly observed in image synthesis using generative networks. The network is further compressed to run on a low-power device such as a smartphone using our proposed size and latency optimization. Experimental result demonstrates that the proposed method produces 360 $^\circ$ panorama with satisfactory image quality of up to 512 × 1024 resolution. It is also shown that the proposed method outperforms the alternative method and can be generalized for non-panoramic scenes and real images captured by a smartphone camera.

Towards self-optimisation in fog computing environments

In recent years, the number of smart devices has grown exponentially. However, the computational demand for latency-sensitive applications has grown and the traditional model of cloud computing is no longer able to meet alone all the needs of this type of application. In this direction, a new paradigm of computation was introduced, it is called Fog Computing. Many challenges need to be overcome, especially those regarding issues such as security, power consumption, high latency in communication with critical IoT applications, and need for QoS. We have presented a container migration mechanism to Fog and the cloud computing that supports the implementation of optimisation strategies to achieve different objectives and solutions to problems of resources allocation. In addition, our work emphasises the performance and latency optimisation, through an autonomic architecture based on the MAPE-K control loop, and provides a foundation for the analysis and optimisation architectures design to support IoT applications.

Latency and Energy Optimization Using MAC-Aware Routing for WSNs

Protocols for wireless sensor networks are generally designed following the layered protocol stack where layers are independent. Uncorrelated decisions coming from different layers may affect certain metrics such as the latency of communications, the energy consumption, etc. Cross-layer approaches overcome this problem by exploiting the dependencies between the layers. In this article, the authors propose latency and energy mac-aware routing for wireless sensor networks (LEMAR-WSN), a new cross-layer routing approach using information of the TDMA schedule and exploiting the information of the energy consumed by each node in order to optimize the latency of communications and the energy consumption when relaying information to the sink in a wireless sensor network. Simulation results show that the proposed approach improves the average latency of communications up to 20% and the average.

Joint Optimization of Latency and Deployment Cost Over TDM-PON Based MEC-Enabled Cloud Radio Access Networks

The convergence of optical and wireless networks, and the marginalization of centralized data centers are required for next-generation telecommunications. According to the operators' concern of cost-efficient deployment combined with latency-sensitive requirements, network planning will lay a foundation for the development of next-generation networks. In particular, the network planning of mobile edge computing (MEC)-enabled cloud radio access networks (C-RAN) in a cost-effective and low-latency manner becomes exceedingly challenging. However, most research studies focus on the cost-effective deployment of C-RAN without considering the optimal placement of MEC servers simultaneously. Moreover, deployment costs and access latency will be directly affected by the optimal deployment of both the MEC servers and access networks. Therefore, we investigate how to coordinate the optimal deployment of MEC servers and C-RAN for latency-sensitive services. In detail, a MILP model is proposed to minimize the joint cost of latency and network deployment of time division multiplexing-passive optical network (TDM-PON) based MEC-enabled C-RAN. Given the model's complexity, a heuristic algorithm is also proposed to solve the mixed integer linear programming (MILP) model. Simulations are conducted to evaluate the performance comparisons of different approaches under different network scenarios. The impacts of some key parameters on comparisons of different approaches are also analyzed.

Latency Optimization for Coded Computation Straggled by Wireless Transmission

Although distributed computing is an efficient solution to large-scale computational tasks, its performance is determined by the slowest computing nodes, i.e., the computation stragglers among distributed nodes. To mitigate their impact, coded computation has emerged as a promising technique by introducing clever computational redundancy. Most existing works only considered the computation latency, however, transmission latency is a serious concern in wireless computing networks. In such a case, the performance is straggled by not only local computation but also wireless transmission. In this letter, we consider the wireless distributed computing network where nodes have different transmission rates and different computation capabilities. Unlike computation stragglers always existing, we derive a straggling factor for each node to indicate whether the network is straggled by transmission. To deal with both stragglers, the wireless coded computation (WCC) algorithm is proposed, which generalizes existing methods. By comparing WCC with four benchmark schemes, Uniform Uncoded Allocation, Non-Uniform Uncoded Allocation, Uniform Coded Allocation, and Heterogeneous Coded Matrix Multiplication, it shows that WCC results in significant speedups of up to 72%, 66%, 64% and 47% over the four aforementioned benchmark schemes, respectively.

The Decision Latency Optimization Problem in SDN With Multi-Controller

This letter studies the problem of optimizing the decision latency for software-defined networking (SDN) structured by multiple controllers. According to the mode of processing non-local tasks in SDN, we propose a novel model for this multi-controller structure, which generalizes the traditional flat and hierarchical structures. We formulate the decision location problem (DLP) to optimize the decision latency based on this model. It is shown that the DLP in the flow initialization can be reformulated as a 0-1 integer optimization problem, which we use a heuristic genetic algorithm to solve. The numerical evaluation verifies the effectiveness of the proposed method.

Step by Step Towards Effective Human Activity Recognition: A Balance between Energy Consumption and Latency in Health and Wellbeing Applications.

Human activity recognition (HAR) is a classification process that is used for recognizing human motions. A comprehensive review of currently considered approaches in each stage of HAR, as well as the influence of each HAR stage on energy consumption and latency is presented in this paper. It highlights various methods for the optimization of energy consumption and latency in each stage of HAR that has been used in literature and was analyzed in order to provide direction for the implementation of HAR in health and wellbeing applications. This paper analyses if and how each stage of the HAR process affects energy consumption and latency. It shows that data collection and filtering and data segmentation and classification stand out as key stages in achieving a balance between energy consumption and latency. Since latency is only critical for real-time HAR applications, the energy consumption of sensors and devices stands out as a key challenge for HAR implementation in health and wellbeing applications. Most of the approaches in overcoming challenges related to HAR implementation take place in the data collection, filtering and classification stages, while the data segmentation stage needs further exploration. Finally, this paper recommends a balance between energy consumption and latency for HAR in health and wellbeing applications, which takes into account the context and health of the target population.

Collision-Free Sequential Task Offloading for Mobile Edge Computing

In this letter, a collision-free sequential task offloading scheme to multiple mobile-edge computing servers is proposed. The problem is formulated as a multi-objective optimization of latency and offloading failure probability. An exact solution technique is developed to obtain a benchmark optimal solution for the problem. A more computationally efficient heuristic algorithm is additionally developed, whose sub-optimal solution yields a performance close to optimal. Simulation results illustrate that the proposed offloading scheme can effectively reduce both latency and offloading failure probability.

QoS-aware traffic scheduling framework in cognitive radio based smart grids using multi-objective optimization of latency and throughput

To meet the diverse QoS requirements of future Smart Grid (SG) communication network, efficient traffic scheduling and optimization techniques to prioritize data from various SG applications are required. Recently, focus has only been on the latency and criticality based priority-aware policies for same channel bandwidth. Traffic scheduling and subsequent channel allocation are required to support both the differential throughput and latency requirements simultaneously for channels having different bandwidths and SNRs. In this paper, a QoS-aware framework for data traffic scheduling in cognitive radio based SG communication network is proposed. The channels available to smart grid Communication Nodes (SCNs) are categorized as low and high bandwidths. For each bandwidth, all SG applications are categorized into several priority-classes comprised of latency and throughput sub-classes. For both channel bandwidths, the scheduler maintains and updates two sets of priority queues based on weights associated with each class and its data type. The complete scheduling framework is formulated as a multi-objective optimization problem. The overall objective function is the weighted sum of individual utility functions of latency and throughput. A novel usage of Adam optimizer is proposed to minimize the latency and maximize the throughput by obtaining optimal system cost, resulting in optimal decision policy. Simulation results show that the proposed algorithm achieves desired QoS requirements in the presence of heavy PU traffic, whereas in case of no priority assignment, QoS requirements of lower priority applications are met by compromising the QoS of higher priority data.

Latency performance analysis for safety‐related information broadcasting in VeMAC

AbstractThe performance analysis is a powerful tool to reveal the features of protocols. We can find hints to optimize protocols through performance analysis. Ultralow latency transmission is one of stringent requirements for event‐driven safety‐related message transmission in vehicular medium access control (VeMAC) protocol. In this paper, we focus on analyzing the latency performance for event‐driven safety‐related application in VeMAC. As analyzed, the optimization of MAC layer latency depends on minimization of the slot acquisition latency. Then, we derive the mean slot acquisition latency according to the behavior of nodes. There are two typical protocol versions, with the split parameter equals to zero and infinity, respectively. From the analytical model, although the protocol version with split parameter equal to infinity can reduce the merge collision through split slots with in a frame, it prolongs the mean slot acquisition latency since the block‐wised slots placement. Then, we develop an interleaved frame structure that can reduce the mean slot acquisition latency and merge collision simultaneously. The interleaved frame structure can adapt to various vehicular densities through the proposed slots allocation algorithm. The derived theoretical mean latency shows the advantages of the proposed frame structure. At last, we present simulation results to verify the analytical model and show that the proposed scheme can provide desired performance.

Minimizing data collection latency with unmanned aerial vehicle in wireless sensor networks

The benefits of using Unmanned Aerial Vehicles (UAVs) as mobile sinks for data collection have attracted great attention in Wireless Sensor Networks (WSNs). The problem that computes the optimal trajectories for UAVs to collect data from WSN is generally NP-Hard. However, the existing works focus on the optimal trajectories of UAVs while considering the data transmission based on either predefined path sets or paths with predefined hovering points, or they focus on seeking the optimal paths while ignoring the data transmission latency between UAVs and sensors. In this paper, we focus on the Transportation and Communication Latency Optimization (TCLO) problem which is to find the optimal trajectory of UAV in a continuous space to collect all data from sensors in a WSN, while minimizing the sum of travelling time and data transmission time without predefined paths or hovering points. To solve the TCLO problem, we first study a special case of the TCLO problem, which is called the TCLO-disjoint problem, in which the sensor neighborhoods are disjoint. An approximation algorithm is proposed for the TCLO-disjoint problem. Based on the TCLO-disjoint problem, we propose an approximation algorithm for the TCLO problem. The proposed algorithm is verified by extensive simulations, which shows its effectiveness to minimize the data collection latency of UAV in WSNs.

Simultaneous Area and Latency Optimization for Stochastic Circuits by D Flip-Flop Insertion

Stochastic computing (SC) is an unconventional computing technique using digital circuits. It performs arithmetic computation on stochastic bit streams (SBSs), which encode real values through the ratios of ones in the streams. Despite its advantages such as simple arithmetic units and strong error tolerance, SC faces two big challenges: 1) long computation latency and 2) large hardware overhead to generate independent SBSs. A recent work proposes to insert D flip-flops (DFFs) into the stochastic circuit to reduce the overhead to generate SBSs. In this paper, observing that DFFs can also be exploited to reduce circuit delay, we propose a novel method to insert DFFs into a stochastic circuit to simultaneously reduce the computation latency of the circuit and the overhead of generating SBSs, thus addressing both challenges at the same time. Experimental results showed that compared to the state-of-the-art method in optimizing stochastic circuits with DFF insertion, our method can reduce the computation latency by 14.3% and the number of DFFs by 48.1%.

Joint Optimization of Energy Consumption and Latency in Mobile Edge Computing for Internet of Things

With wide adoption of Internet of Things (IoT) across the world, the IoT devices are facing more and more intensive computation task nowadays. However, the IoT devices are usually limited by their computing capability and battery lifetime. Mobile edge computing provides new opportunities for developments of IoT, since edge computing servers which are close to devices can provide more powerful computing resources. The IoT devices can offload the intensive computing tasks to edge computing servers, while saving their own computing resources and reducing energy consumption. However, the benefits come at the cost of higher latency, mainly due to additional transmission time, and it may be unacceptable for many IoT applications. In this paper, we try to find a tradeoff between the energy consumption and latency, in order to satisfy user demands of various IoT applications. We formalize the problem into a constrained multiobjective optimization problem and find the optimal solutions by a modified fast and elitist nondominated sorting genetic algorithm (NSGA-II). To improve the performance of the algorithm, we propose a novel problem-specific encoding scheme and genetic operators in the proposed modified NSGA-II. We also conduct extensive simulation experiments to evaluate the proposed algorithm and its sensitivity under certain major parameters. The experimental results show that the proposed algorithm can find a large number of optimal solutions to adjust the corresponding offloading decision according to the real-world situation.

Modified X–Y routing for mesh topology based NoC router on field programmable gate array

Network on chip (NoC) has been proposed as an enormously scalable solution to address communication problems in system on chip (SoC). The interconnections among multiple cores/multiple Intellectual Property modules on a chip have a major impact on communication and performance of the chip design in terms of area, throughput, latency and power. Hence, an efficient design of the NoC interconnect is of paramount importance. In this study, a novel idea of the NoC router using a single-side buffer in the input block and a programmable priority encoder in the scheduler is discussed and a 4 × 4 mesh topology based router design is implemented on a field programmable gate array device using these blocks. This gives an area optimisation of the router due to the use of a small buffer. Furthermore, a modified X–Y routing algorithm is discussed and implemented for the 8 × 8 mesh topology-based router. The result shows that the modified X–Y routing algorithm requires only two hops to reach any source destination pair resulting in latency optimisation of the NoC router. Thus, the router design with a single buffer and a modified X–Y routing algorithm provides the best solution for area and latency optimisation.

Differentiated Latency in Data Center Networks with Erasure Coded Files Through Traffic Engineering

This paper proposes an algorithm to minimize weighted service latency for different classes of tenants (or service classes) in a data center network where erasure-coded files are stored on distributed disks/racks and access requests are scattered across the network. Due to limited bandwidth available at both top-of-the-rack and aggregation switches and tenants in different service classes need differentiated services, network bandwidth must be apportioned among different intra- and inter-rack data flows for different service classes in line with their traffic statistics. We formulate this problem as weighted queuing and employ a class of probabilistic request scheduling policies to derive a closed-form upper-bound of service latency for erasure-coded storage with arbitrary file access patterns and service time distributions. The result enables us to propose a joint weighted latency (over different service classes) optimization over three entangled "control knobs": the bandwidth allocation at top-of-the-rack and aggregation switches for different service classes, dynamic scheduling of file requests, and the placement of encoded file chunks (i.e., data locality). The joint optimization is shown to be a mixed-integer problem. We develop an iterative algorithm which decouples and solves the joint optimization as 3 sub-problems, which are either convex or solvable via bipartite matching in polynomial time. The proposed algorithm is prototyped in an open-source, distributed file system, {\em Tahoe}, and evaluated on a cloud testbed with 16 separate physical hosts in an Openstack cluster using 48-port Cisco Catalyst switches. Experiments validate our theoretical latency analysis and show significant latency reduction for diverse file access patterns. The results provide valuable insights on designing low-latency data center networks with erasure coded storage.

Latency-Optimal Virtual Network Functions Resource Allocation for 5G Backhaul Transport Network Slicing

The concept of network slicing (NS) has been proposed for flexible resource provisioning where a physical resource is partitioned into logically independent networks on demand. The NS resource allocation implies the definition of a feasible path in the infrastructure network with adequate resource availability. However, due to complex structural characteristics of the backhaul transport network, a number of issues arise when fast deploying the end-to-end (E2E) slices onto network infrastructures. In this paper, a pair-decision resource allocation model is firstly formulated to construct the mapping relationship between logical networks and substrate networks in a coordinated way. In order to improve extreme quality of service (QoS) and user experiment, latency-optimal virtual resource allocation problem is defined, subject to the backhaul capacity and bandwidth constraints. The problem is formulated as an integer linear programming (ILP) and solved with the branch-and-bound scheme, whose resolution yields an optimal virtual network function (VNF) placement and traffic routing policy. Numerical results reveal that the proposed scheme can enable the transport network latency optimization with a reduction of up to 30% and 41.6% compared to the Network Slice Design Problem (NSDP) and Random Fit Placement Algorithm (RFPA) schemes respectively. In the meanwhile, the network load balance and serviceability have been improved efficiently with better resource utilization as well.

Memory latency optimizations for the elementary functions on the Sunway architecture

As fundamental software of high-performance computers, elementary functions have a significant impact on the performance of the high-level applications. Benefiting from the Chinese-designed manycore system consisting of processing cores and auxiliary cores, the Sunway TaihuLight supercomputer is considered as one of the fastest supercomputers in the world, having ranked on the top of the TOP500 supercomputer list several times. The processing cores of the Sunway architecture are coupled using a shared memory strategy, leading to high latency of memory accesses and performance degradation of the elementary functions where a variety of memory accesses exist. To address this issue, we propose a set of optimizations for memory latency of the Sunway processing cores. Firstly, we obtain a reduced data table in the context of guaranteed accuracy by optimizing underlying algorithms, grouping and mapping, removing error compensations, etc. Secondly, we perform data movement from the global memory shared by all processing cores to the scratchpad memory of individual processing cores, significantly reducing the memory latency. Finally, we convert the memory accesses that cannot be localized due to the limited space of the scratchpad memory into equivalent immediate loads and/or shift operators, further improving the performance. In addition, we automate the algorithm by carefully selecting the most suitable data conversion approach and table-lookup algorithm, mitigating the code explosion issue effectively. We implement our method and evaluate the effectiveness of the optimizations by conducting experiments on the Sunway architecture. The experimental results show that exponential functions can achieve performance improvements by 91 and 86.2% from the data movement and data conversion strategies.