Latency (engineering)

About: Latency (engineering) is a research topic. Over the lifetime, 3729 publications have been published within this topic receiving 39210 citations. The topic is also known as: lag.

Low-latency heading feedback control with neuromorphic vision sensors using efficient approximated incremental inference

Abstract: Asynchronous neuromorphic vision sensors have unique properties that make them ideal for high speed control applications. We consider a one dimensional simplification of a more general six dimensional trajectory tracking problem for mobile platforms, and present a computationally efficient method for feedback control that takes advantage of the asynchronous, event-based nature of these sensors to provide very high bandwidth and low latency feedback. This is an important step toward application of these incredible sensors to mobile robotic systems and useful in its own right. Through experimental tests we compare sensors and show that neuromorphic vision sensors can provide good closed loop performance in terms of computation, data rate, frequency and latency, and tracking error.

TailCutter: Wisely cutting tail latency in cloud CDN under cost constraints

Abstract: Cloud computing platforms enable applications to offer low latency access to user data by offering storage services in several geographically distributed data centers. In this paper, we identify the high tail latency problem in cloud CDN via analyzing a large-scale dataset collected from 783,944 users in a major cloud CDN. We find that the data downloading latency in cloud CDN is highly variable, which may significantly degrade the user experience of applications. To address the problem, we present TailCutter, a workload scheduling mechanism that aims at optimizing the tail latency while meeting the cost constraint given by application providers. We further design the Maximum Tail Minimization Algorithm (MTMA) working in TailCutter mechanism to optimally solve the Tail Latency Minimization (TLM) problem in polynomial time. We implement TailCutter across data centers of Amazon S3 and Microsoft Azure. Our extensive evaluation using large-scale real world data traces shows that TailCutter can reduce up to 68% 99th percentile user-perceived latency in comparison with alternative solutions under cost constraints.

A 240x180 120dB 10mW 12us-latency sparse output vision sensor for mobile applications

Abstract: This paper proposes a CMOS vision sensor that combines event-driven asychronous readout of temporal contrast with synchronous frams-based active pixel sensor (APS) readout of intensity. The image frames can be used for scene content analysis and the temporal constrast events can be used to track fast moving objects, to adjust the frame rate, or to guide a region of interest readout Therefore the sensor is suitable for mobile applications because it allows low latency at low data rate and low system-level power consumption. Sharing the photodiode for both readout types allows a compact pixel design that is 60% smaller than a comparable design. The 240x180 sensor has a power consumption of 10mW. It is built in 0.18um technology with 18.5um pixels. The temporal contrast pathway has a minimum latency of 12us, a dynamic range of 120dB, 12% contrast detection threshold and 3.5% contrast matching. The APS readout has 55dB dynamic range with 1% FPN.

Design and implementation of a cloud-based event-driven architecture for real-time data processing in wireless sensor networks

Abstract: The growth of the Internet of Things (IoTs) and the number of connected devices is driven by emerging applications and business models. One common aim is to provide systems able to synchronize these devices, handle the big amount of daily generated data and meet business demands. This paper proposes a cost-effective cloud-based architecture using an event-driven backbone to process many applications’ data in real-time, called REDA. It supports the Amazon Web Service (AWS) IoT core, and it opens the door as a free software-based implementation. Measured data from several wireless sensor nodes are transmitted to the cloud running application through the lightweight publisher/subscriber messaging transport protocol, MQTT. The real-time stream processing platform, Apache Kafka, is used as a message broker to receive data from the producer and forward it to the correspondent consumer. Micro-services design patterns, as an event consumer, are implemented with Java spring and managed with Apache Maven to avoid the monolithic applications’ problem. The Apache Kafka cluster co-located with Zookeeper is deployed over three availability zones and optimized for high throughput and low latency. To guarantee no message loss and to simulate the system performances, different load tests are carried out. The proposed architecture is reliable in stress cases and can handle records goes to 8000 messages in a second with low latency in a cheap hosted and configured architecture.

Resource Allocation for Open-Loop Ultra-Reliable and Low-Latency Uplink Communications in Vehicular Networks

Abstract: To support ultra-reliable and low-latency communication (URLLC) in vehicular networks, the virtual cells, where multiple access points (APs) cooperatively serve one mobile node, have been proposed to reduce the end-to-end latency in the downlink. The latency can be further reduced by eliminating the need for retransmission and feedback control, i.e., open-loop communications. However, it is difficult to achieve a high reliability of the uplink via virtual cells, because of multiple access interference and collisions from other virtual cells nearby. In this paper, we formulate the proactive radio resource allocation in the open-loop uplink of vehicular networks as a stochastic optimization problem with the objective to maximize the uplink reliability while ensuring the network stability. The optimal resource allocation policies are obtained solving the optimization problem using the Lyapunov optimization technique in a distributed manner. To reduce the computational and to improve the challenges of the Lyapunov optimization, we propose a virtual resource slicing algorithm that maps radio resource units to virtual resource blocks. Simulation results exhibit that both ultra-low latency and ultra high reliability are guaranteed in the open-loop uplink of vehicular networks. Based on the theoretical performance analysis and simulations, the uplink radio access procedure is summarized for the URLLC in vehicular networks.